Electoral system changes and party factionalism: the...

25
1 Electoral system changes and party factionalism: the consequences of electoral reforms on intra-party dynamics. Paper submitted at the IPSA Conference, Madrid 2012, RC06 ‘Political Sociology’, Panel “The consequences of electoral reforms: voters, parliaments, and party systems”. Caroline Close Research Fellow at F.R.S. FNRS CEVIPOL - Université Libre de Bruxelles Paper presented with the support of / Communication présentée avec le soutien de / Association belge de science politique-Communauté française (ABSP-CF) Abstract 1 Electoral systems have been for a long time a core topic in the field of comparative politics. Causes and consequences of electoral system have been analyzed in different areas of the polity, both at the intra and the inter-party dimensions. During the last two decades, many countries have experienced electoral system change (New Zealand, Italy, Japan…) - increasing scholars’ attention towards electoral reform. The paper focuses on the consequences of electoral system on party factionalism, and asks how recent electoral reforms can affect this intra-party dimension. While few systematic cross-sectional studies have been led on the topic, the paper attempts to fill this gap by investigating the effect of national legislative election rules (IV) on party factionalism (DV) in fifteen European parliamentary systems. The paper focuses exclusively on factionalism within the parliamentary party group (PGG), operationalized as a propensity to dissent, and measured through MPs’ dissenting preferences, attitudes and behaviors collected via cross-national MP surveys. The results suggest that recent reforms increasing the personalization of the vote would promote candidate-centered behavior in the parliamentary arena, thus increasing the risk of factional politics. This is a work in progress. Please cite at your own risk! 1 This paper constitutes a small part of a broader PhD dissertation, entitled ‘Party factionalism. A comparative study of the impact of institutional and organizational factors on the degree of factionalism inside political parties’. This PhD thesis concentrates on the institutional causes of party factionalism (especially on PPG factionalism), through a comparative cross-sectional analysis. At the system level, the research examines the impact of the electoral system on factionalism. At the party level, the research focuses on the consequences of candidate selection methods on factionalism.

Transcript of Electoral system changes and party factionalism: the...

1

Electoral system changes and party factionalism: the consequences of

electoral reforms on intra-party dynamics.

Paper submitted at the IPSA Conference, Madrid 2012, RC06 ‘Political Sociology’, Panel

“The consequences of electoral reforms: voters, parliaments, and party systems”.

Caroline Close

Research Fellow at F.R.S. – FNRS

CEVIPOL - Université Libre de Bruxelles

Paper presented with the support of / Communication présentée avec le soutien de /

Association belge de science politique-Communauté française (ABSP-CF)

Abstract1

Electoral systems have been for a long time a core topic in the field of comparative politics.

Causes and consequences of electoral system have been analyzed in different areas of the

polity, both at the intra and the inter-party dimensions. During the last two decades, many

countries have experienced electoral system change (New Zealand, Italy, Japan…) -

increasing scholars’ attention towards electoral reform. The paper focuses on the

consequences of electoral system on party factionalism, and asks how recent electoral reforms

can affect this intra-party dimension. While few systematic cross-sectional studies have been

led on the topic, the paper attempts to fill this gap by investigating the effect of national

legislative election rules (IV) on party factionalism (DV) in fifteen European parliamentary

systems. The paper focuses exclusively on factionalism within the parliamentary party group

(PGG), operationalized as a propensity to dissent, and measured through MPs’ dissenting

preferences, attitudes and behaviors collected via cross-national MP surveys. The results

suggest that recent reforms increasing the personalization of the vote would promote

candidate-centered behavior in the parliamentary arena, thus increasing the risk of factional

politics.

This is a work in progress. Please cite at your own risk!

1 This paper constitutes a small part of a broader PhD dissertation, entitled ‘Party factionalism. A comparative

study of the impact of institutional and organizational factors on the degree of factionalism inside political

parties’. This PhD thesis concentrates on the institutional causes of party factionalism (especially on PPG

factionalism), through a comparative cross-sectional analysis. At the system level, the research examines the

impact of the electoral system on factionalism. At the party level, the research focuses on the consequences of

candidate selection methods on factionalism.

2

Introduction

The study of electoral systems represent nowadays one of the most prominent areas of

comparative politics. Traditionally, scholars have treated electoral systems as independent

variables (Duverger 1951). They have classified electoral systems, and analyzed their effects

on politics (Leyenaar and Hazan 2011:437-438). Electoral systems have been viewed as

‘fixed’ institutions, resulting from particular cultural and historical features (Taagepera and

Shugart 1989). Then, as the 1990s saw a growing number of stable democracies adopting

electoral reforms (Italy, Japan, Israël, New Zealand2), electoral systems have been

increasingly observed as dependent variables (Benoit 2007; Leyenaar and Hazan 2011). As

Leyenaar and Hazan schematize (2011), scholars have begun to examine what is an electoral

reform, why does a change occur and how, which actors are involved (who) and finally, in

which countries do changes occur (where). Though we acknowledge the need to focus on the

inputs of electoral reforms rather than on the outcomes (Leyenaar and Hazan 2011:450), we

argue that work has still to be done on the intra-party consequences of electoral system and

electoral reform. Therefore, this paper takes the ‘traditional’ conception of electoral system

and examines electoral rules as independent variables impacting on the intra-party dimension.

Analyzing electoral system as an independent variable, the political discipline has particularly

questioned how voting systems impact on political and party systems (Duverger 1951;

Lijphart 1994; Rae 1967; Sartori 1976; Taagepera and Shugart 1989). At the individual level,

some scholars have investigated the impact of electoral system design on political behavior,

through institutional, rational choice and socio-psychological approaches. Some have focused

on voters’ behavior (Karp 2006; Klingemann 2009; Pilet 2008), others on individual

politicians’ strategic behavior (Blais and Indridason 2007).

Focusing on elected representatives’ behavior, legislative studies scholars have hypothesized

how electoral system design would impact on parliamentary party unity or cohesion (Bowler

et al. 1999; Carey 2007; Hix 2004) - which they operationalize through MP voting behavior

and measure through roll-call voting data3. These works generally share a basic rational

assumption: securing a legislative seat in the next election constitutes the first aim of any

incumbent politician. Looking at the impact of electoral system design on politicians’

behavior, scholars observe that elected representatives adopt different voting behaviors,

depending on whether electoral rules make them more accountable to party leaders or to

voters (see Carey’s theory of competing principals, 2007): voting with their own parties -

party-centered behavior; or against their own party - candidate-centered behavior or personal

vote-seeking behavior (Carey and Shugart 1995), a phenomenon linked to the personalization

of the vote (Karvonen 2010; McAllister 2007; Pilet and Renwick 2011). These legislative

2 See the special issue of the International Political Science Review / Revue internationale de science politique

(1995), Vol.16, No. 1,‘The Politics of Electoral Reform. La politique de la réforme électorale’. 3 In this paper, roll-call voting data are not used. Although measures such as the Rice Cohesion Score and the

Agreement Index (Hix et al. 2005) have proved to be efficient and useful for comparing legislatures, using

legislative voting data as a measure of party cohesion or discipline can be quite problematic (see Carubba et al.

2008; Hug 2010; Sieberer 2006). Among other problems, cross sectional studies must face the diversity of ways

“in which roll call votes occur in parliaments” (Hug 2010: 226). Moreover, Hug (2010) has put into light the

way ‘selection effects’ might bias the measure of party cohesion based on published or requested roll calls votes:

roll call voting would indicate party (enforced) discipline rather than cohesion. But most of all, a huge problem

arises when using indexes such as Rice Cohesion Score for comparing causes of party unity in terms of

institutional differences in European legislatures: as Sieberer (2006) shows, almost all European parliaments

included in his study score above 95 points on average on the Rice Index, thus few differences might be

observed and explained.

3

studies constitute one of the starting points of this paper. In the same way, the paper examines

the impact of electoral rules on MPs’ behaviors, which in turn partly determine PPG unity,

fragmentation or factionalism.

The literature on party factionalism constitutes another starting point of this work. During the

last decades, researchers have increasingly acknowledged that political parties are not unitary

actors, but rather “collective entities in which competition, divided opinions and dissent create

internal pressures” (Boucek 2009: 55). Though not directly observable, factions are in most

parties formed as consequences of these pressures and are therefore an important aspect of the

political arena. Although causes of factionalism remain unknown, case studies on

factionalized parties (mainly the Italian Democrazia Cristiana and the Japanese Liberal

Democratic Party) suggest that political institutions, particularly the electoral system, can

effectively impact on party factionalism (Boucek 2002; Cox and Rosenbluth 1995; Katz 1980;

Kohno 1992). Moreover, Japan and Italy, both known for their exceptional cases of

institutionalized factions (in the LDP and in the DC), implemented electoral reforms at the

beginning of the 1990s, which, according to some experts, were aimed at curing the diseases

produced by too much factional politics (clientelism, corruption, extreme fragmentation…)

(Bettcher 2005; Cox et al. 1999; Di Virgilio and Kato 2001). However, very few systematic

cross-sectional analyses have been led on the consequences of such electoral system changes

on factionalism, mainly because comparing factions in different contexts has been extremely

challenging.

Using micro-level attitudinal data generated by the PARTIREP4 cross-national MP survey,

the paper offers a new way of conceptualizing a parliamentary party’s degree of factionalism

(PPG factionalism) - i.e. in terms of MPs’ propensity to express internal dissent. We argue

that MPs from the same party share a common ideological orientation, but can have interests

and personal views on political, societal and economic issues that diverge from those of their

party. These diverging preferences and the extent to which MPs actually express their

disagreements within their PPG (from remaining silent to organizing collective sub-party

dissension) can be understood in terms of degree of factionalism. To what extent electoral

rules can impact on MP’s propensity to dissent? This is the core question of the present paper.

Similarly to legislative studies scholars, we observe the impact of electoral rules on MPs’

behavior; but we also look at how they impact on MPs’ preferences.

The first part of the paper sets the theoretical background on which we establish a series of

hypotheses linking electoral system design (IV) to PPG factionalism (DV). Drawing on both

the literature on factionalism and legislative studies on party voting unity, section 1.1. defines

what is intended as PPG factionalism. Looking at the factors affecting PPG factionalism,

section 1.2 gives some preliminaries on the link between factionalism, electoral system and

electoral reform. Then, section 1.3 shows that the assumptions regarding the impact of

electoral rules on factionalism clearly merge with those assuming that voting unity or disunity

4 The PARTIREP research project (http://www.partirep.eu/) is funded by the Belgian Federal Science Policy

Office (BELSPO). The responsible partners for the Comparative MP survey in each of the participating countries

are: W.C. Müller (Uniersität Mannheim), K. Deschouwer (Vrije Universiteit Brussel), O. Costa (Université de

Bordeaux), B. Wessels (Wissenschafszentrum Berlin) and M. Edinger (Universität Jena), G. Ilonszki (Corvinus

University Budapest), M. Gallagher (Trinity College Dublin), R. Hazan (Hebrew University Jerusalem), L.

Verzichelli (Universitá di Siena), R. Andeweg (Universiteit Leiden), D.A. Christensen and T. Midbø (Rokkan

Centre Bergen), J. Wasilewski (Warsaw School of Social Sciences and Humanities), A. Freire and J.M. Leite

Viedas ( ISCTE-IUL Lisbon University Institute and CIES-ISCTE), P. Oñate (Universidad Valencia) and M.

Mendez (Centro de Investigaciones Sociologicas), D. Caramani (Universität Sankt Gallen), and J. Bradbury

(Swansea University).

4

is partly determined by electoral system. On these bases, section 1.4. hypothesizes how PPG

factionalism might be affected by electoral system design. The second part of the paper then

empirically tests these hypotheses. First, section 2.1. presents the data used in the analysis,

collected as part of the PARTIREP international project. Second, section 2.2. describes how

we measure ‘PPG factionalism’ across cases. Third, section 2.3. tests the hypotheses built in

part 1 linking electoral system design (IV) to PPG factionalism (DV). Fourth, section 2.4.

attracts attention towards the remaining theoretical and methodological problems of the

analysis and suggests how we could handle them. Finally, part 3 suggests how recent electoral

reforms in the democratized world would further influence parliamentary party factionalism.

1. PPG factionalism as a consequence of electoral system design: theoretical

background.

1.1. Defining PPG factionalism

Factionalism: a complex and multifaceted phenomenon

Defining factions and factionalism is much challenging, as attested by the richness of the

literature, dating back to the writings of the American ‘Founding Fathers’5 and British

thinkers6. The most commonly accepted definition of a faction comes from Raphael Zariski

(1960): “any intra-party combination, clique, or grouping whose members share a sense of

common identity and common purpose and are organized to act collectively – as a distinct

bloc within the party – to achieve their goals” (Zariski 1960: 33). In a more dynamic

perspective, factionalism describes a situation of internal dissent within a party, which can

develop into processes of party fractionalization or “subgroup partitioning” (Boucek

2009:455) organized around the fulfillment of a particular goal. Actually, factionalism can

designate very different realities in a wide range of contexts, what makes comparative study

quite problematic.

Several typologies have attempted to describe the various dimensions factionalism

encompasses (Belloni and Beller 1978; Bettcher 2005; Hine 1982; Panebianco 1988; Sartori

1976). These successive classifications of intra-party conflict have revolved around three

main aspects: first, what are the motivations of group partitioning; second, how these

dissenting groups organize; third, where do conflicts occur within the party. As far as the

motivational dimension is concerned, party factions can be of very different nature: based on

shared ideological preferences (Nyomarkay 1965; Rose 1964); or on interest-based, patron-

client relationships (Key1949). Second, on the organizational dimension, factions can appear

as long-rooted, institutionalized groups (as in the Japanese LDP and Italian DC) characterized

by cohesion and discipline (Rose 1964: 37); or rather as weak and ephemeral sub-party

groupings (Key 1949). Third, factionalism can occur at various levels within the party (in the

party in central office, in the party in public office, or in the party on the ground; or even

throughout the whole party).

This paper deals with factionalism referring to a situation of internal dissent within the PPG,

and operationalizes it in terms of MPs’ preferences, attitudes and behavior. As the next

5 Two chapters within the Federalist papers are dedicated to the defense of the Union against factions. The

Federalist n°9 by Hamilton: “The Utility of the Union as a Safeguard Against Domestic Faction and

Insurrection”; and The Federalist n°10 by Madison: “The Utility of the Union as a Safeguard Against Domestic

Faction and Insurrection (continued)” (http://www.constitution.org/fed/federa00.htm). 6 Among which Bolingbroke, Hume, and Burke. For a comprehensive review, read Sartori (1976).

5

section explains, the paper observes this form of factionalism as a continuous variable -

namely, as a degree of factionalism - along the organizational dimension. Dissension and

factionalism are viewed as dynamic processes, occurring through successive ‘sequences’.

PPG factionalism: a sequential dynamic process

Looking at PPG dynamics implies drawing on the vast array of parliament and legislative

studies, and opening up the unresolved conceptual debates on legislative cohesion or unity,

discipline or loyalty (Andeweg and Thomassen 2011; Hazan 2003; Özbudun 1970). Clearly,

all these concepts relate to, one way or another, factionalism and dissent. Factionalism has

once been linked to the concept of coherence (Janda 1980), defined as “the degree of

congruence in the attitudes and behavior of party members” (Janda 1980: 118)7 that results

from cohesion and discipline. In other studies, factionalism has been opposed -

indistinguishably - to both cohesion and discipline (Bell and Shaw 1994), and associated with

party division, fragmentation, internal conflict and dissent. Defining factionalism within the

parliamentary arena therefore implies to clarify its link with party unity, cohesion and

discipline, and with the way scholars have operationalized these concepts.

This paper echoes recent works from scholars who distinguish between cohesion, discipline,

party loyalty and voting unity within the PPG, and who consider a “sequential relationship”

(van Vonno 2011) between these concepts. The sequential perspective considers MP voting

behavior (with or against his/her party) as a decision-making process, resulting from various

mechanisms. Furthering the works of Ozbudun (1970), Bowler et al. (1999) and Hazan

(2003), sequential scholars identify these mechanisms as follows (Andeweg and Thomassen

2011; Kam 2001; van Vonno 2011): party agreement, resulting from common ideological

preferences; party loyalty, produced by party socialization and solidarity; and party discipline,

referring to positive or negative sanctions used by party leaders to prompt voting unity.

Ultimately, voting unity would be determined by a combination of all these factors, operating

in different ‘sequences’.

At first, in determining his/her vote on a given issue in parliament, an MP would estimate if

he/she agrees or not with the party official line. If there is no conflict between his/her

individual preferences and those of his/her party - i.e. if there is some “homogeneity of

preferences” (Andeweg and Thomassen 2011) - unity is reached. However, if there are

conflicting preferences, two options remain for the individual MP: remaining loyal to the

party by voting with his/her party in spite of his/her disagreement about the policy in

question, or voting against the party line. At this stage, socialization plays a central role. If

norms of socialization do not operate, party leaders would enforce party unity through

sanctions – although sanctions seem to be “more an ultimum remedium” (Andeweg and

Thomassen 2011: 669).

Like scholars who see MP voting decision as a sequential decision-making process, we

consider factionalism as a dynamic process occurring through ‘sequences’. Originally,

7 See Janda’s (1980) ‘Political Parties. A Cross National Survey.’ The book reports the work of the International

Comparative Political Parties (ICPP) project. Within this study, 158 political parties operating in 53 countries

around the world during 1950-1962 are put under scrutiny: parties are described and analyzed through concepts

pertaining to their external relations on the one hand and their internal organization on the other hand. Among

the four concepts relating to a party’s internal organization, one directly concerns “coherence of behavior”

(Janda 1980: 7). Janda splits coherence into several dimensions: legislative cohesion, ideological factionalism,

issue factionalism, leadership factionalism, strategic or tactical factionalism and party purge.

6

factionalism refers to a situation of internal dissent within the PPG. Dissent within the PPG8

can serve as a ‘voice’ (Hirschman 1970)9 function, with MPs (organized or not in factions)

fulfilling the role of representation (Köllner and Basedeau 2005) of ideological currents

within the party or of interests of external groups (unions, localities, women…). These

dissensions might arise before a vote on a particular policy in parliament. As far as the

organizational aspect of factionalism is concerned, these internal dissensions can remain quite

silent (latent disagreement); or they can develop into real party fragmentation if they come to

be expressed publicly (e.g. dissenting MPs criticizing their party leaders in parliament or in

the media, or voting against their party in the parliamentary arena). Further, at a critical stage,

dissenting MPs might organize and form sub-PPG groups, which will work for the fulfillment

of particular goals against the will of the central party office or party leader.

Thus, several degrees of factionalism within a PPG can be identified along the organizational

dimension10

, by examining MPs’ preferences, attitudes, and behaviors. PPG factionalism can

be conceived as a continuum, ranging from very low factionalism with no dissension at one

extreme (MPs very rarely disagree with the party line, and automatically vote with their party)

to openly dissenting MPs, expressing their disagreement through voting against the party line

at the other extreme. In between, we would find latent dissension: a situation where MPs

often share different views on a policy from those of the PPG leadership, but where they still

stick to the party line and vote according to it. This ‘in between situation’ might occur due to

a form of party loyalty, which we think is inversely correlated to voice (Birch 1975).

As mentioned in the introduction, most legislative studies seeking explanations for MP voting

behavior are dominated by a common rational assumption: any incumbent politician’s

primary goal is to secure (re-)election, and to this end he/she displays strategic behaviors,

depending on the structure of incentives or institutional framework in which he/she plays.

This study shares a similar assumption: MPs’ voice behavior can be seen as a strategic

behavior resulting from institutional incentives, such as the electoral system. In the following

sections, we show that the hypotheses linking electoral system design to factionalism (section

1.2.) are quite similar to those linking electoral system design to MP voting behavior (section

1.3.). Finally, in section 1.4. both literatures are used to formulate the hypotheses we test in

the analysis (part 2).

1.2. Factionalism, electoral system design and electoral reforms: preliminaries

Recently, Boucek has argued that “different institutional regimes produce different intra-party

incentives and can shape the behaviour of party dissidents in different ways” (Boucek 2005:

27). Throughout the literature, hypotheses about the influence of institutional arrangements on

factionalism have concerned: state structure (federal or decentralized system) (Grynaviski

2004; Sartori 1976); legislative-executive relations (presidential or parliamentary system)

(Köllner and Basedeau 2005); legal and financial regulations regarding parties (Cox and

Rosenbluth 1993; Kohno 1992); regulations pertaining to the parliamentary functioning

8 Such situation of internal dissent can happen in two different ways: we can observe differences of opinions

between individual MPs and their party official line; or between a group of MPs and another group. For the sake

of simplicity, the paper only considers the first type of relationship (MPs versus party leaders). 9 Hirschman’s (1970) famous trilogy (exit, voice and loyalty) and following works have been associated to the

study of factionalism in a previous paper (see Close 2011). 10

Again for the sake of simplicity, the analysis below ignores the motivational dimension of factionalism, which

would imply a qualitative analysis of the nature of disagreements (depending on the policy content or policy

issue).

7

(Köllner and Basedeau 2005); as well as leadership and candidate selection methods (Cox and

Rosenbluth 1996; Katz 1982; Leiserson 1968).

Rational choice and (neo) institutional researchers have also developed explanations linked to

electoral system design. Sartori (1976) and Boucek (2005, 2009) have hypothesized that

within ‘first-past-the-post’ systems, where the effective number of parties is limited, inter-

party competition is high. As a consequence, discipline within the party is reinforced in order

to make the party win - and thus survive. As discipline is strengthened, intra-party divisions

are stifled. Assessing the impact of intra-party competition, other scholars have pointed out

systems that put co-partisans into direct electoral competition. Within proportional systems,

the use of open and semi-open list has been considered as a potential factor (Morgenstern

2001) that could “exacerbate competition over the choice of candidates” (Obler 1973: 184),

who have to build a personal reputation to defeat candidates of the same party. Single Non

Transferable Vote in Japan (Kohno 1992; Cox and Rosenbluth 1995) and the intraparty

preference voting system in Italy are among the most studied cases (Katz 1980; Katz and

Bardi 1980; Hine 1982). As far as district magnitude is concerned, the Japanese case again

appears as a well-known example: medium-size districts are said to have reinforced the

effects of the SNTV system and to have maintained an ‘optimal number’ of five factions

within the LDP (Kohno 1992).

Japan and Italy not only represent countries where we find exceptional cases of

institutionalized factions; they are also known for the electoral reforms they implemented at

the beginning of the 1990s, along with Israël, Britain and New Zealand11

. In Italy, these

reforms have largely been explained by popular dissatisfaction with proportional system that

was thought to produce government instability and party fragmentation (Donovan 1995). The

Japanese and Italian electoral systems were also accused to cause political corruption and one-

party domination (for Japan see Shiratori 1995; for Italy see Donovan 1995). In fact, people

associated corruption and political instability with the factional politics at play in the two

dominant parties (DC and LDP); reforms were therefore aimed at lessening the incentives to

factionalize. In the 1990s, both systems “moved to variants of mixed member systems, similar

in certain regards to that used in Germany” (Norris 1995:7). These reforms had, according to

some ‘factionalism scholars’, real impact on intra-party politics in both the Japanese LDP

(Cox et al. 1999) and the Italian DC (Bettcher 2005). In both cases, “factional structure ought

to have weakened under the new rules” (Cox et al. 1999:33; Di Virgilio and Kato 2001).

These reforms, their reasons and impacts, suggest that factionalism is genuinely linked to

electoral system design. But few empirical cross-sectional comparisons have been led that

show an evident relationship between electoral system design and party factionalism outside

the well-known cases of LDP and DC, mostly because measuring factionalism across cases

remains much challenging12

. Using cross-national MPs attitudinal data, the paper offers a new

way of conceptualizing a parliamentary party’s degree of factionalism, in terms of MPs’

propensity to dissent. Since this definition merges with the literature on party voting unity,

section 1.3. reviews how electoral system design has been studied as an explanation for voting

11

See the special issue of the International Political Science Review / Revue internationale de science politique

(1995), Vol.16, No. 1, ‘The Politics of Electoral Reform. La politique de la réforme électorale’. 12

See for instance Katz’ (1980) comparative work A Theory of Parties and Electoral Systems. Drawing on ICCP

data, his book compares the impact of the electoral system in Britain, Ireland and Italy on three dimensions:

number of parties, ideological orientation, and internal unity. However, when Katz talks about party cohesion or

disunity, he doesn’t build a precise index of factionalism. In fact, he recognizes that the indicators for

factionalism were highly questionable and thereby he decides to ignore them.

8

(dis)unity. Then, section 1.4. presents a series of hypotheses linking electoral system (IV)

design to PPG factionalism (DV).

1.3. Electoral system design and party voting unity: a review.

As for factionalism, several institutional features have been pointed out as explanations for

voting unity: regime type (presidential or parliamentary), federalism (Carey 2007), control

over candidate nomination (Hazan and Rahat 2011) and over various sorts of intra-

parliamentary resources (Sieberer 2006), and electoral system (Carey and Shugart 1995;

Carey 2007; Hix 2004). Though not unanimously accepted, common explanations state that

voting unity is higher in parliamentary system then in presidential ones (Ozbudun 1970); that

federalism fosters party division between national and sub-national branches (Carey 2007:92);

and that decentralized and inclusive candidate selection methods tends to lessen party unity

(Sieberer 2006, Hazan and Rahat 2011). Besides, some scholars have investigated the effect

of government participation on party unity, some viewing this effect as positive (Carey 2007),

other as negative (Sieberer 2006).

Some hypotheses linking electoral rules to legislative behavior merge with those linking

electoral system design to factionalism: systems that increase intra-party competition - i.e.

that put co-partisans into direct electoral competition – would tend to foster candidate-

centered behavior, whereas electoral systems that lessen intra-party competition are more

likely to produce cohesive parties. In this vein, Carey and Shugart (1995) have identified four

variables that can affect “the extent to which individual politicians can benefit by developing

personal reputations distinct from those of their party” (1995:418): ballot13

(the degree to

which voters or party leaders determine the order in which candidates get elected), pool

(whether votes cast for one candidate of a given party also contribute to the election of that

candidate’s co-partisans in the district), votes (whether voters can cast a single partisan vote,

multiple candidate votes or a single vote below the party level) and district magnitude14

.

Generally, preferential voting has been identified as a major cause of personal vote-seeking

behavior (Bräuninger et al. 2012; Karvonen 2004; Katz 1980). As Norris (2002:4) argues:

where citizens exercise a preference vote […] this strengthens the chances that

particular candidates from the list will be elected and therefore changes their

rank. Under these rules, politicians have a moderately strong incentive to offer

particularistic benefits, to stand out from rivals within their own party.

Drawing on both the studies dealing with the consequences of electoral system on legislative

behavior, and those dealing with electoral system impact on factionalism, section 1.4. builds a

series of hypotheses linking electoral system design to MPs’ dissenting preferences and

13

The ballot dimension defined by Carey and Shugart as “the degree of control party leaders exercise over

access to their party’s label” also encompasses a variable linked to the degree of control leaders exercise over

party endorsements (candidate selection procedures). Though Carey and Shugart consider candidate selection

methods as rules deriving from the electoral system, we argue that candidate and leadership selection should

rather be examined as a ‘meso’ level factor (see for instance Shomer 2011). This paper does not examine the

impact of these selection methods, but as mentioned (see note 1), the broader PhD research includes them as

independent variables distinct from the electoral system design. See also a previous paper (Close 2012), which

exclusively deals with the impact of candidate selection methods on factionalism. 14

As far as district magnitude (M) is concerned, Carey and Shugart (1995) suggest that the way it affects

personal reputation depends on the value of the ballot variable. « In all systems where there is intraparty

competition, as M grows, so does the value of personal reputation. Conversely, in systems where there is no

intraparty competition, as M grows, the value of personal reputation shrinks » (Carey and Shugart 1995:418).

9

behaviors. Three variables are investigated: the general electoral formula, and, drawing on

Carey and Shugart (1995), the ballot (list type) and votes variables (number of preferential

voting that can be expressed).

1.4. Electoral system design and factionalism: hypotheses

This paper argues that different electoral rules can impact on factionalism, not only by

influencing party voting unity, but also by enhancing homogeneity or heterogeneity of

preferences within a party, and by leading MPs to adopt loyal or ‘voice’ behaviors. Looking

for explanations of PPG different degrees of factionalism, we draw on both the literature on

factionalism and legislative studies on party voting unity and roughly hypothesize that

electoral systems that increase intra-party competition tend to produce factionalized parties.

We estimate the degree of intra-party competition by observing three variables that are in fact

linked to each other: electoral formula (proportional or majoritarian); list type (rather closed

or open), which corresponds to Carey and Shugart’s ballot variable; and votes variable, in

order to study the effect of preferential voting on MPs propensity to dissent15

.

First, as far as the general electoral formula is concerned, the literature tends to agree that

proportional list systems, by contrast to single-member plurality or majority systems, are

more prone to produce factionalized parties, since they put co-partisans into direct

competition. According to these assumptions, the PARTIREP dataset should reveal that the

highest degrees of factionalism are found in proportional systems.

H1: Factionalism is higher in proportional systems than in single-member

plurality/majority systems.

Second, the ballot variable constitutes a decisive incentive to intra-party competition (Katz

1980: 116). Ballot or ballot control is intended as the degree to which voters or the party

leaders determine the order in which candidates get elected. Pilet and Renwick (2011) identify

three categories: “those in which intra-party candidate ordering is entirely determined by

voters [open list system] those in which it is determined partly by voters and partly by parties

[flexible list] and those in which it is entirely determined by parties [closed list] ” (Pilet and

Renwick 2011:15). We assume that the less a party retains control over the process, the more

intra-party competition, thus the more dissenting behaviors.

The PARTIREP project classifies PR list systems in three categories: closed, open and

‘flexible, but rather closed’ list systems. The first category implies that voters have no power

to alter the order in which candidates are elected; thus, they express a choice for a party rather

than for one or multiple candidate(s). By contrast, in open list systems, how candidates gain

some seats is not pre-determined by the order in which they appear on the ballot, but highly

depends on preferential votes. Open list systems are therefore clearly candidate oriented,

what implies higher competition between co-partisans. In between, flexible list systems offer

voters the possibility to express their preferences towards one or more candidates;

nevertheless, at the end voters have very little power to alter the list order (Pilet 2003:13) -

15

Eventually, we could also look at two other dimensions included in Carey and Shugart’s (1995) classification:

pool and district magnitude. Vote pooling system is excluded from our analysis because in all the PR cases

investigated, pooling occurs across the whole party, excepted in Ireland where pooling occurs below the party

level (implying a country effect). District magnitude is not included in our analysis mainly for the sake of

simplicity, time and space, since it is a complex variable to play with.

10

this explains why the PARTIREP project describes flexible list systems as rather closed. In

flexible list systems, intra-party competition should be limited.

H2: Factionalism is higher in open list systems than in closed and ‘flexible

rather closed’ list systems.

Third, intra-party competition in proportional list systems could also be determined by the

votes variable, which asks whether “voters are allowed to cast only a single vote for a party,

multiple votes, or a single vote for a candidate” (Carey and Shugart 1995: 422). Indubitably,

this variable is linked to the ballot variable: closed list systems allow voters to cast only a

single vote for a party, while flexible and open list systems allow voters to express

preferences for a single or multiple candidates. This work argues that when voters can vote

for multiple candidates, it moderates intra-party competition; and that when they cast a single

vote below the party level, it increases party competition. Therefore,

H3: Within open and flexible list systems, factionalism is higher when voters

can express their preferences for a single candidate below the party level.

Part 2 here under empirically tests these hypotheses. Section 2.1. presents the data we use in

the analysis, collected as part of the PARTIREP international project. Section 2.2. describes

how we operationalize and measure the dependent variable (PPG factionalism). Section 2.3.

tests the hypotheses built in this section, linking electoral system design (IV) to PPG

factionalism (DV). Then, section 2.4. suggests how we could improve the analysis, at both the

theoretical and the methodological levels. Finally, part 3 gives some considerations on how

recent electoral reforms in the democratized world would further influence parliamentary

party factionalism.

2. PPG factionalism: a consequence of electoral system? Empirical testing.

2.1. Data

This study relies on data collected as part of the PARTIREP international project. The project

surveyed national and regional legislators in fifteen European democracies16

. In this analysis,

only those MPs from national parliaments are included. Data were collected between the

spring 2009 and the spring 2011; MPs were invited to respond either through online web-

survey (46.8%), print questionnaires (33.7%), face-to-face interview (18.7%) or telephone

(0.8%). Table 1 gives information on response rates in each national parliament. The dataset

includes a quantitative cross-national attitudinal survey carried out among members of the

legislatures, and other macro variables, a part of them linked to electoral system. For the

purpose of this analysis, survey questions are used to measure the dependent variable

(factionalism), and among macro variables are selected the independent variables (electoral

system).

16

14 geographically European, plus Israel.

11

Table 1. PARTIREP - country information

Country Pop. (N) Resp. (n) Resp (%)

Austria 183 55 30.0

Belgium 150 70 46.6

France 577 50 8.6

Germany 622 134 21.5

Hungary 386 99 25.6

Ireland 166 34 20.5

Israël 120 39 32.5

Italy 630 45 7.1

Netherlands 150 65 43.3

Norway 169 46 27.2

Poland 460 55 12.0

Portugal 230 76 33.0

Spain 350 104 29.7

Switzerland 200 49 24.5

United Kingdom 639 62 9.7

Total 5032 983 19.5

Here below, the paper is further divided into three parts. First, section 2.2. explains how PPG

factionalism is measured in the fifteen parliaments, using the quantitative attitudinal data.

Second, section 2.3. tests the hypotheses linking electoral system design to PPG factionalism,

using the PARTIREP macro variables (for the independent variables). Third, section 2.4.

brings to attention the theoretical and methodological limits of the analysis. Finally, part 3

suggests how recent waves of electoral reforms could impact on intra-party factionalism in

European democracies.

2.2. Measuring PPG factionalism

As explained above (section 1.1.), PPG factionalism can be observed through MPs’

preferences, attitudes and behaviors. Via the PARTIREP survey, two questions are used in

order to compute a variable denoting each national parliament degree of factionalism. The

first question indicates the frequency of disagreement an MP faces with the party line17

, which

is considered as a valid measure of party agreement18

. Table 2 presents the response to this

question by national parliament, dichotomized between MPs who ‘rarely disagree’ and those

who ‘often disagree’. Intuitively, we consider that MPs who rarely disagree with the party line

are less prone to vote against their party line, and consequently less prone to organize into

factional games19

. At the aggregate level, more than 60% of respondents indicate they have

17

How often, in the past year, would you say you have found yourself in the position that your party had one

opinion on a vote in Parliament, and you personally had a different position? Answer categories are: Once per

month, once per three months, once per year, and (almost) never. 18

This question has already been used in a previous study on MP decision-making by van Vonno (2011). In

order to validate the measure of ‘agreement’ as a measure of ‘homogeneity of preferences’, she compares her

measure of party agreement with the perceived ideological distance between an MP and his/her party (using the

absolute distance between a respondent’s self-placement and his/her placement of his/her political party on 11-

point Left-Right ideological scale). She shows that “at the aggregate level, 47% of those who usually agree with

the party line perceive no ideological distance, where 69% of those who frequently disagree with the party line

perceive at least a 1-point difference”; and concludes that “the question concerning frequency of disagree on a

vote in parliament is probably a valid measure of party agreement as a determinant of party unity” (van Vonno

2011:10). 19

Actually, cross-tabulating party agreement with MP voice or loyalty (see Table 3), we can see that, at the

aggregate, 60.8% of MPs who rarely disagree ought to vote along their party line (loyalty), and 50.3% of those

who often disagree ought to vote according to their own opinion (voice).

12

rarely faced a situation of dissension between their own preferences and those of their party

during the last year; whereas around 40% indicate that dissension has occurred frequently.

Austria (78%), Spain (78%), Ireland (79%), and the Netherlands (71%) appear to have the

highest proportion of MPs who rarely disagree; whereas Italy (55%), Israël (54%), Portugal

(51.3%), Hungary (51%) and France (50%) have more than half of their respondents stating

they often disagree.

Table 2. Party agreement.

Country Rarely disagree

Disagree once a

year/almost never

Often disagree

Disagree once per

month/ once every three

months

Austria 78.3 (36) 21.7 (10)

Belgium 62.9 (44) 34.1 (26)

France 50.0 (25) 50.0 (25)

Germany 57.1 (76) 42.9 (57)

Hungary 49.5 (49) 50.5 (50)

Ireland 78.8 (26) 21.2 (7)

Israël 46.2 (18) 53.8 (21)

Italy 45.5 (20) 54.5 (24)

Netherlands 71.0 (44) 29.0 (18)

Norway 62.2 (28) 37.8 (17)

Poland 66.7 (36) 33.3 (18)

Portugal 48.7 (37) 51.3 (39)

Spain 77.9 (81) 22.1 (23)

Switzerland 56.2 (27) 43.8 (21)

United Kingdom 56.5 (34) 43.5 (27)

Average % (n) 60.3 (582) 39.7 (383)

χ²= 68.558, df = 28, sig.= .000

Φ =.224, sig.=.000

The second question used to compute an index of factionalism indicates how an MP ought to

vote in case of such disagreement20

- vote according to his/her own opinion or in line with

his/her party’s position. This variable serves as a proxy for each MP’s actual voting behavior,

but at the same time as an attitude towards voting behavior. Intuitively, we consider that MPs

voting against their own party would be more prone to join or create factions – since their

behavior would attest a higher degree of dissent. Table 3 shows, by country, the responses to

the question of how an MP ought to vote. At the aggregate level, a majority of MPs declare

that MPs should vote along the party line (56%); whereas 44% of MPs state they should vote

according to their own opinion. Norway (88%), and again the Netherlands (87%), Spain

(79%) and Ireland (76%) appear to have the lowest proportions of ‘dissenting’ MPs (thus

being less prone to factionalized). In contrast, Switzerland (88%) shows an extremely high

proportion of dissenting MPs, followed by Italy (66%) and France (63%).

20

And how should, in your opinion, a Member of Parliament vote in the situation that his party has one opinion

on a vote in Parliament, and he personally has another? Answer categories are: MP should vote according to his

own opinion, or MP should vote according to his party’s position. A question also used by Van Vonno (2011).

13

Table 3. MP voice or loyalty.

Country Vote according to party

opinion (loyalty)

Vote according to own

opinion (voice)

Austria 50.0 (22) 50.0 (22)

Belgium 67.1 (47) 32.9 (23)

France 36.7 (18) 63.3 (31)

Germany 43.5 (54) 56.5 (70)

Hungary 62.1 (59) 37.9 (36)

Ireland 75.8 (25) 24.2 (8)

Israël 51.4 (19) 48.6 (18)

Italy 34.1 (15) 65.9 (29)

Netherlands 86.7 (39) 13.3 (6)

Norway 88.1 (37) 11.9 (5)

Poland 49.1 (26) 50.9 (27)

Portugal 56.0 (42) 44.0 (33)

Spain 79.4 (81) 20.6 (21)

Switzerland 12.5 (6) 87.5 (42)

United Kingdom 45.5 (25) 54.5 (30)

Average % (n) 56.2 (515) 43.8 (401)

χ²= 132.848, df = 14, sig.= .000

Φ = .381, sig.=.000

These two variables help us to build an index of ‘factionalism’, which rather indicates a

propensity or a risk to factionalize. Factionalism, as mentioned above, is intended as a process

resulting from dissenting preferences, attitudes and behaviors. In order to obtain an aggregate

index of factionalism for each national parliament, a ‘dissent’ variable is computed. This

variable describes three different scenarios of dissent, and takes three different values: 0 (no

dissent), 1 (latent dissent), 2 (open dissent).

Table 4. ‘Dissent’ scenarios and factionalism index.

Country No dissent (0) Latent dissent (1) Open dissent (2) Factionalism

index

Austria 38.6 (17) 11.4 (5) 50.0 (22) 1.11

Belgium 37.1 (26) 30.0 (21) 32.9 (23) .96

France 24.5 (12) 12.2 (6) 63.3 (31) 1.39

Germany 25.8 (32) 17.7 (22) 56.5 (70) 1.31

Hungary 34.7 (33) 27.4 (26) 37.9 (36) 1.03

Ireland 65.6 (21) 12.5 (4) 21.9 (7) .56

Israël 27.0 (10) 24.3 (9) 48.6 (18) 1.22

Italy 18.6 (8) 16.3 (7) 65.1 (28) 1.47

Netherlands 64.4 (29) 22.2 (10) 13.3 (6) .49

Norway 57.1 (24) 31.0 (13) 11.9 (5) .55

Poland 40.4 (21) 9.6 (5) 50.0 (26) 1.10

Portugal 25.3 (19) 30.7 (23) 44.0 (33) 1.19

Spain 60.8 (62) 18.6 (19) 20.6 (21) .60

Switzerland 8.3 (4) 4.2 (2) 87.5 (42) 1.79

United Kingdom 32.7 (18) 12.7 (7) 54.5 (30) 1.22

Average % (n) 36.8 (336) 19.6 (179) 43.6 (398) 1.06

χ²= 129.869, df =28, sig.= .000

Φ = .426, sig.=.000

The first scenario - no dissent - refers to a situation of an MP who rarely disagrees with

his/her party’s opinion and votes according to his/her party opinion. The second scenario

describes a situation called latent dissent in which an MP often disagrees, but again votes

14

according to his/her party position. The third scenario is the most susceptible to give rise to

organized factionalism, since it indicates a situation of open dissent, where an MP disagrees

(rarely or often), but whenever he/she disagrees with the party line, he/she votes against

his/her party. Table 4 shows country variations of these scenarios, and the aggregate

‘factionalism index’. The closer the index is to 0, the less factionalized the party would tend

to be; the closer to 2, the more factionalized the party would tend to be.

As table 4 shows, Switzerland (1.79), Italy (1.47), France (1.39) and Germany (1.31) display

the highest degrees of factionalism (or risk of factionalism), with an index higher than 1.30.

These countries also have higher proportions of MPs openly dissenting (MPs who would vote

according to their own opinion in case of disagreement with the party line). The lowest

degrees of factionalism are found in the Netherlands (0.49), Norway (.55), Ireland (0.56) and

Spain (0.60), all also showing the four highest proportions of MPs who do not dissent (rare

disagreement between them and their party). Latent dissent scores high in Norway (31%),

Portugal (30.7%) and Belgium (30.0%), indicating that almost a third of MPs consider that

they often disagree with the party line, but still remain loyal.

Factionalism is not a static phenomenon, but rather a dynamic process (Boucek 2009), which

can change through time in response to different types of incentives. Factionalism, as defined

in this paper, is generated from divergent individual preferences, dissenting individual

attitudes and behaviors. Drawing on the neo-institutional paradigm, this paper argues that

MPs’ preferences, attitudes and behaviors can be influenced by institutional factors.

Especially, the analysis focuses on specific political institutions, which are external to the

legislature itself21

: electoral rules. Section 2.3. below analyzes how several aspects of the

electoral system design (electoral formula, list type and preferential vote) can affect PPG

dissent and factionalism, first separately, then it evaluates their combined effects on PPG

factionalism and dissent. Indeed, although each aspect probably has a proper impact, they are

indubitably linked to each other, and it is more likely that MPs’ preferences and behaviors are

influenced by a combination of these factors.

2.3. Electoral system design and factionalism in European parliaments: analysis and

findings

Before exploring the association between our dependent and independent variables, Table 5

gives information on the fifteen national electoral systems and their factionalism index. As

Table 5 shows, there is hardly a clear relationship between one of the independent variables

and the degree of factionalism.

The four highest degrees of factionalism occur in parliaments which are elected either by a

proportional (Italy and Switzerland), plurality (59% of German MPs) or two-round majority

(France) system; within the PR systems associated with high degrees of factionalism, lists are

either open (Switzerland) or closed (Italy); and contrary to H3, the highest degree of

factionalism (Switzerland) is not associated with a preferential voting towards a single

candidate below the party level. A look at the lowest degrees of factionalism does not clarify

the story: within Netherlands, Norway, Ireland, and Spain, four different electoral systems are

found. The relationship between these systems and the low factionalism indexes appears

rather counter intuitive: all systems are PR. Especially, the case of Ireland is really puzzling:

although the STV represents one of the most candidate-centered systems (Mitchell 2000), the

21

Though some scholars consider that electoral systems might be endogenous to the party system (Benoît 2007).

15

Irish legislators appear among the less ‘dissenting’ MPs. This paradox has been examined by

Farrell et al. (2012), who show that Irish TDs “are at one and the same time excessively voter-

oriented and excessively loyal to their party” (2012:20).

Table 5. Factionalism and Electoral system design - country information.

Country Factionalism

index

Electoral formula Ballot Vote

Austria 1.11 PR Flexible but rather closed

(80%) + closed (20%)

For a single candidate

below party level (80%)

+ vote for a party (20%)

Belgium .96 PR Flexible but rather closed For multiple candidates

France 1.39 2RMajority - -

Germany 1.31 PR (41%) + plurality

(59%) Closed - Vote for a party -

Hungary 1.03 PR (45,5%) + 2R

Majority (54,5%) Closed - Vote for a party -

Ireland .56 PR STV - For multiple candidates

Israël 1.22 PR Closed Vote for a party

Italy 1.47 PR Closed Vote for a party

Netherlands .49 PR Flexible but rather closed For a single candidate

below party level

Norway .55 PR Flexible but rather closed For multiple candidates

Poland 1.10 PR Open For a single candidate

below party level

Portugal 1.19 PR Closed Vote for a party

Spain .60 PR Closed Vote for a party

Switzerland 1.79 PR Open For multiple candidates

United

Kingdom 1.22 Plurality - -

These preliminary observations do not help validating or invalidating our hypotheses. In

order to investigate more adequately the impact of electoral system design on MPs’

preferences and behaviors, especially in mixed systems (Austria, Germany, and Hungary), the

impact of the electoral formula, ballot and votes variables is examined at the MP level.

Moreover, in order to better grasp the impact of the IV respectively on agreement, loyalty and

voice, in the following tables the observed dependent variable is the ‘dissent’ variable - the

three different levels or ‘sequences’ of dissent (no dissent, latent dissent and open dissent).

Dissent and electoral formula

Table 6 shows the bivariate relationship between the four categories of electoral formula

through which PARTIREP MPs have been elected: two-round majority (French MPs and 54%

of Hungarian respondents), plurality (British MPs and 59 % of German respondents), PR STV

(Irish MPs) and PR for the remaining MPs (75% of the whole sample). According to Table 6,

our first hypothesis seems invalidated: among the MPs elected through majoritarian or

16

plurality rule, those who openly dissent (vote according to their own opinion whenever they

disagree) are over-represented; whereas among those elected through PR and PR-STV

systems, those who do not dissent are over-represented. As far as latent dissent is concerned -

which denotes a situation of frequent disagreement, but where loyalty prevents from voice -, it

seems more correlated with PR as well as with two round-majority systems.

Table 6. Dissent*Electoral formula

2R-Majority Plurality PR PR-STV Total

No dissent 26.6 (25) 34.9 (38) 37.2 (252) 65.6 (21) 36.8 (336)

Latent dissent 20.2 (19) 15.6 (17) 20.5 (139) 12.5 (4) 19.6 (179)

Open dissent 53.2 (50) 49.5 (54) 42.3 (287) 21.9 (7) 43.6 (398)

Total % (n) 100 (94) 100 (109) 100 (678) 100 (32) 100 (913)

χ²= 18.615, df =6, sig.= .005

Φ = .143, sig.=.005

These results suggest that intra-party competition between co-partisans in list systems is not

the sole determinant of dissenting behavior, nor does it lead to current disagreements within a

PPG22

. Here, it is rather the extent to which the vote is personalized (Pilet 2003) that impact

on MPs’ dissension. In single-member plurality or majority system, Pilet argues that voters

express their choice towards a candidate before they chose a party platform (2003:5). Other

scholars point out that in single-member plurality systems, in comparison to list systems,

representatives are more accountable to voters: “Single-member electoral districts create a

"singular" relationship between voters and MPs: local voters' surveillance task is facilitated

by identifying only one representative amongst a crowd who they must monitor” (Dunleavy

and Margetts 1995:14). Therefore, we can hypothesize that, in single-member plurality or

majority systems more than in PR systems (at least in closed and flexible list systems),

incumbent MPs will tend to put themselves forward within the parliamentary arena (as well as

in the media etc) in order to attract popular attention towards their personal competences and

qualities, or defend their ‘local’ voters’ interests - strategically using the voice option rather

than the loyalty one.

Since PR-STV only applies in the specific Irish case, there is a clear country effect that has to

be taken into account regarding the effect of PR-STV. As far as plurality and two-round

majority systems are concerned, there might also exist a country effect since each system

occur in only one country: France (plus a proportion of Hungarian MPs) and United Kingdom

(plus a proportion of German MPs). Actually, 75% of MPs in the whole sample are elected

through PR systems (melting the respective effect of closed, open and flexible list systems).

This might explain that, among MPs elected through PR systems, the proportions of MPs who

do not dissent, dissent in a latent way or openly dissent are close to the proportions existing in

the whole sample. Consequently, a closer look at the effect of list type appears necessary.

22

The associations between electoral formula and party agreement on the one hand, and electoral formula and

voting behavior on the other, have been examined respectively through cross-tabs. Their results show a similar

tendency: the proportion of MPs who often disagree is higher in plurality (42.2%) and two-round majority

system (51.6) than in the whole sample (39.7); the proportion of MPs who vote against their party is higher in

plurality (49.5%) and two-round majority system (53.2) than in the whole sample (43.8).

17

Dissent and ballot

As mentioned, the PARTIREP project classifies PR list systems in three categories: closed,

open and ‘flexible, but rather closed’ list. Closed and flexible list systems are identified as

rather party-centered systems, whereas open list systems lead to candidate-oriented votes.

According to our second hypothesis (H2), we expect to find more openly dissenting MPs

among MPs elected through open list system than among those elected through closed and

flexible list system.

Table 7 seems to confirm our second hypothesis: there is a significant over-representation of

MPs who openly dissent among those elected through open list system, and a clear over-

representation of those who do not dissent among MPs elected through ‘flexible rather closed’

list system. However, we would have expected that the proportion of MPs who do not dissent

would be higher among those elected through closed list system; since this system is supposed

to lead to lower intra-party competition (voters have no power at all to alter the list order).

Even more surprising: the proportion of MPs who openly dissent is slightly higher in closed-

list system than in the whole sample, and the proportion of those who do not dissent is slightly

lower in closed-list system than in the whole sample.

Table 7. Dissent*List type (ballot)

Closed list Flexible list, but

rather closed

Open list Total

No dissent 35.2 (135) 47.4 (92) 25.0 (25) 37.2 (252)

Latent dissent 21.6 (83) 25.3 (49) 7.0 (7) 20.5 (139)

Open dissent 43.2 (166) 27.3 (53) 68.0 (68) 42.3 (287)

Total % (n) 100 (384) 100 (194) 100 (100) 100 (678)

χ²= 47.121, df =4, sig.= .000

Φ = .264, sig.=.000

According to these results, it seems that in ‘flexible rather closed’ list systems, MPs would

display less dissenting behaviors and less frequent disagreement than in open list and closed

list systems23

. Latent dissent (which implies that even in case of disagreement with the party

line, MPs still vote along with their party) is logically more associated with closed and

flexible list system and not with open list system: it seems that only when voters have a real

power to alter the list, do MPs have an interest to openly express their disagreement - voice

instead of loyalty.

Dissent and vote

Thirdly, within flexible and open list systems, do MPs often disagree with their party and

adopt dissenting behaviors depending on whether voters can cast multiple preference votes

below the party level or only a single one? The third hypothesis suggests that when voters can

only cast a single vote for a candidate within a list, this increases intra-party competition and

increases the ‘risk’ of party factionalism. Table 8 does not indicate that such connection exists

23

Again, the associations between list type and party agreement (rarely/often) on the one hand, and list type and

voting behavior on the other, have been examined respectively through cross-tabs. Surprisingly, the proportion

of MPs who often disagree is higher in closed list system (41.8%) than in the whole sample (38.6), contrary to

what happens in open list (38.2) and flexible list system (32.6). As far as voting behavior alone is concerned, the

proportion of MPs who vote against their party is much higher in open list system (68.3) and slightly higher in

closed list system (43.4) than in the whole sample (42.5).

18

within European democracies: our third hypothesis cannot be supported by the data (sig.=

0.154)24

.

Table 8. Dissent*Number of Preferences (vote)

Vote for a party Vote for multiple

candidates

Vote for a single

candidate below

party level

Total

No dissent 35.2 (135) 39.1 (75) 47.0 (63) 38.5 (273)

Latent dissent 21.6 (83) 20.8 (40) 14.9 (20) 20.1 (143)

Open dissent 43.2 (166) 40.1 (77) 38.1 (51) 41.4 (294)

Total % (n) 100 (384) 100 (192) 100 (134) 100 (710)

χ²= 6.677, df =4, sig.= .154

Φ = .097, sig.=.154

Party, candidate and preference ballot: impact on factionalism

Finally, two among our three variables appear statistically significant in explaining MPs

preferences and behaviors: electoral formula and list type. Although the analysis suffers from

country effects and from an overrepresentation of MPs elected through proportional system,

the results suggest that: first, single-member plurality or majority systems - where voters

chose a candidate before a party platform and where representatives are more accountable to

their constituents - lead to more frequent disagreement and encourage dissenting behaviors;

second, open list systems, in which the order in which candidates get elected is entirely in the

hands of voters, lead to more frequent disagreement and encourage MPs to voice their

dissension in order to differentiate from their co-partisans. Both systems would thus be more

prone to PPG factionalism.

Mixing Mitchell’s (2000)25

classification of electoral systems and Norris’ (2002) terminology

for ballot structure, Table 9 classifies electoral systems in three categories: party-ballot

systems (closed and ‘flexible rather closed’ list, where preferential voting is usually

ineffective), candidate-ballot systems (single-member plurality or majority) and preference-

ballot systems (open list and STV).

Table 9. Dissent*Electoral system categories.

Party-ballot Candidate-ballot Preference-ballot Total

No dissent 39.3 (227) 31.0 (63) 34.8 (46) 36.8 (336)

Latent dissent 22.8 (132) 17.7 (36) 8.3 (11) 19.6 (179)

Open dissent 37.9 (219) 51.2 (104) 56.8 (75) 43.6 (398)

Total % (n) 100 (578) 100 (203) 100 (132) 100 (913)

χ²= 27.255, df =4, sig.= .000

Φ = .173, sig.=.000

24

However, the association between vote and party agreement (rarely/often) is significant (sig. = 0.02); whereas

the association between vote and voting behavior is not (sig.=0.556). Surprisingly, the proportion of MPs who

often disagree is higher in systems where voters can only express a preference towards the whole party (closed

list, 41.8%) than in the whole sample (37.8). It is within systems where voters can express a vote towards a

single candidate below party level that the proportion of MPs who rarely disagree is the highest (70.8, compared

to 62.2 in the whole sample). 25

van Vonno (2011), also using PARTIREP data, proceeds similarly when she investigates the impact of

electoral systems on party agreement, loyalty and discipline (van Vonno 2011: 21-22).

19

In the first category, the chances for a candidate to win a seat depend heavily on the party’s

popularity; and since voters have no power over the order in which candidates get elected,

dissenting from the party line is far from strategic. In view of the high proportion of latent

dissent in this category, it might be that in such list system, MPs often disagree but stick to the

party line because the extent to and the order in which they appear on the ballot (and further

get elected) heavily depends on party central office and leaders. Besides, their loyalty might

be seen as a strategic behavior depending on candidate selection rules; but we need to further

investigate this relationship.

In the second category (candidate-ballot system), candidates are elected mostly because of

their personal abilities; incumbents thus need to appear as the more accountable to their voters

as possible - even if this implies to express their personal position (or voice their

constituency’s interests) against their party’s official position. However, once they are assured

to be on the ballot, candidates do not have to differentiate themselves from co-partisans; and

since their party’s popularity remains an asset, they also adopt more loyal behavior than

candidates elected through open list system. In the third category, “voters can really make a

difference to the choice of candidates within parties” (Mitchell 2000:342); candidates’

accountability to voters should then be high, and incumbents have a strong incentive to

differentiate themselves from their party and co-partisans.

2.4. Limits

According to these results, electoral systems seem to have an impact on MPs’ preferences,

and on their propensity to express their disagreement within their PPG. But before drawing

such conclusions, the analysis should be improved in several ways.

On the one hand, comments have to be expressed towards the way we intend the dependent

variable factionalism. First, factionalism is much defined and measured in terms of atomistic

attitudes and behaviors, a conception that does not take into account the collective dimension

of such phenomenon26

. But as Cox et al. (1999) suggest, atomism and factionalism are linked,

both resulting from the “inability of individual candidates to rely solely on the party label”

(Cox et al. 1999: 34). Further investigation has to be conducted on how these phenomena

relate to each other. Second, the index of factionalism should be completed with other

variables, such as data pertaining to MPs’ ideological left/right positioning, or data

concerning their inclination to represent particular groups within the society or regional

collectivities or localities. Third, the dissent variable and the index we use for factionalism are

rather data collected at the individual MP level or aggregated at the parliamentary level.

Though this allows us to investigate the impact of national electoral systems on individuals

and parliaments, in future research we should build a ‘party level’ index of factionalism if we

want to look at the effect of party organizational rules.

On the other hand, this analysis does not pretend that electoral system is the only determinant

of party factionalism. We simply suggest that it should be considered as one among other

factors. Sticking to an institutional approach27

, other factors could include those already cited:

26

a faction being a group of party members who “share a sense of common identity and common purpose and

are organized to act collectively– as a distinct bloc within the party – to achieve their goals” (Zariski 1960: 33). 27

Within the literature, other approaches have been used in order to examine the causes of party factionalism at

different levels: at the macro-level, functionalist (Belloni and Beller 1978), socio-historical (Obler 1973;

Panebianco 1988) and cultural approaches (Hine 1982); at the meso level, party organization studies; and at the

micro-level, socio-psychological and rational choice approaches (Boucek 2005; Cox and Rosenbluth 1995).

20

regime type, federalism, parliamentary rules and resources, government participation, etc.

Particularly, factors impacting on intra-party competition at the nomination stage, such as the

decentralization and inclusiveness of the candidate selection processes, should be further

included in the analysis. Besides, in order to better assess the impact of electoral system and

electoral system changes, longitudinal analyzes should be conducted, in addition to cross-

sectional examinations.

Despite these limitations, and with some improvements, this analysis brings new perspectives

in the study of factionalism (e.g. new cross-national data) and of electoral system

consequences on the intra-party dimension. Section 4 below extends the discussion on the

impact of electoral system on factionalism by considering recent electoral reforms, and their

potential consequences for PPG atomism or factionalism.

3. Electoral reforms and party factionalism: discussion

From the end of WWII until the 1990s, democratized countries hardly experienced any

significant change in their electoral systems, feeding the traditional conception of electoral

systems as ‘fixed’ institutions (Taagepera and Shugart 1989). Since the mid-1990s, following

the exceptional electoral reforms in Italy, Japan, and New-Zealand and widespread debates on

electoral reforms in other countries (Renwick and Pilet 2011:3), scholars have attempted to

identify global trends in the adoption of electoral systems, in new and old democracies

(Dunleavy and Margetts 1995). Some have focused on changes from majority system towards

the adoption of more proportionality or vice-versa. At first, scholars have believed in a

common trend towards the adoption of mixed electoral system (Dunleavy and Margetts 1995),

in particular for reasons of electoral efficiency (Shugart 2001). Recently, other scholars

(Karvonen 2010; Pilet and Renwick 2011; Renwick 2011) have examined a trend towards a

personalization of electoral rules: typologies of electoral systems have been established

according to their degree of personalization – evaluated on the basis of many dimensions and

variables such as Carey and Shugart’s (1995) ballot, pool, votes and district magnitude (Pilet

and Renwick 2011).

Our classification of electoral system according to their impact on MPs’ dissenting

preferences and behaviors partly merges with these latter classifications. In view of our

results, it seems that closed and flexible list systems would produce less dissension within a

PPG, probably because no dissent or latent dissent constitute better strategies for MPs seeking

re-election. Open list systems however, in which candidates’ destiny depends on voters’

preferences, lead to open dissent within PPG, with MPs strategically advancing their personal

abilities, personalities and competences. In between, single-member plurality or majority

systems imply that MPs are elected partly thanks to their personal abilities, but also depending

on their party’s popularity. Consequently, MPs in single-member plurality or majority

systems adopt personal vote-seeking behavior, but in a lesser extent than MPs in open list

systems.

According to these results, it seems that it is not only about more or less proportionality of the

system producing more or less intra-party dissensions. Any shift from a closed to a more open

list system, or a shift from a PR closed list towards a single-member majority or plurality

system would increase PPG dissent, and increase the risk of factional politics. A shift from

single-member majority or plurality towards a PR list system would lead either to increased

dissension (in case of open-list system) either to lessened dissension (in case of closed or

‘flexible rather closed’ list system). Roughly, it seems that any electoral reform increasing

21

voters’ choice for candidate(s) rather than for a party would produce more dissension within a

PPG. Nevertheless, the surprising results we obtained as far as flexible systems are concerned

might blur these assumptions.

Among the fifteen countries included in the PARTIREP survey, half have experienced an

electoral reform during the past 10 years28

. No clear tendency emerges in the adoption of

these national reforms. Some have concerned gender quota on party lists (Belgium, Spain),

other deals with the formula translating votes into seats (Germany, Poland), district magnitude

(Norway) or districting (Belgium). A few deals with the openness of the list (Belgium) and

electoral formula (Italy). Belgium has passed a legislation making candidates more central,

but still shows a rather low index of factionalism (0.96, see Tables 4 and 5). Italy however, for

the reasons mentioned in section 1.2, has changed its system towards less centeredness of

candidates, but now display a rather high factionalism index (1.47, see Tables 4 and 5). Does

it mean that reforms have been ineffective? Does it show that they are still needed, or that

actors need time to adapt their strategies? Or does it simply mean that electoral rules are not

the sole determinant of MPs’ behaviors? Without further longitudinal data, we are left with no

answer. Moreover, since factionalism does not only result from electoral rules, other factors

have to be taken into account.

28

Austria, Belgium, Germany, Israël, Italy, Norway, Poland, Spain.

22

References

Andeweg R.B., Thomassen J. (2011) ‘Pathway to party unity: Sanctions, loyalty, homogeneity and division of

labour in the Ducth parliament’, Party Politics, vol.17, No.5, pp. 655-672.

André A., Depauw S., Deschouwer K. (2011) ‘Institutional Constraints and Conceptions of Political

Representation’, paper presented at the APSA Annual Meeting, Panel ‘Descriptive representation in practice’,

Seattle (WA), 1st-4th September 2011.

Bell D.S.,Shaw E. (1994) Conflict and Cohesion in Western European Social Democratic Parties, London:

Pinter Publishers

Belloni F.P., Beller D.C. (eds.) (1978) Faction Politics: Political Parties and Factionalism Comparative

Perspective, Santa Barbara and Oxford: ABC-Clio.

Benoit K. (2007) ‘Electoral Laws as Political Consequences: Explaining the Origins and Change of Electoral

Institutions’, Annual Review of Political Science, Vol.10, pp.363-390.

Bettcher K.E. (2005) ‘Factions of Interest in Japan and Italy: The Organizational and Motivational Dimensions

of Factionalism’, Party Politics, Vol.11, No.3, pp.339-358.

Blais A., Indridason I.H. (2007) ‘Making Candidate Count: The Logic of Electoral Alliances in Two-Round

Legislative Elections’, The Journal of Politics, Vol. 69, No. 1, pp. 193-205.

Boucek, F. (2002) “The Growth and Management of Factionalism in Long-lived Dominant Parties: Comparing

Britain, Italy, Canada and Japan”. Unpublished PhD thesis in Government, London School of Economics and

Political Science, winner of LSE’s William Robson Memorial Prize.

Boucek F. (2005) ‘Why does party dominance end in factionalism and what are the implications for democracy?’

Paper presented at the ECPR General Conference, Budapest, Session 31-8 - Dominant Parties and Factionalism.

Boucek, F. (2009) ‘Rethinking Factionalism: Typologies, Intra-Party Dynamics and Three Faces of

Factionalism’, Party Politics, 15(4), pp.455-485.

Bowler S., Farrell D.M., Katz R.S. (1999) Party Discipline and Parliamentary Government, Columbus: Ohio

State University Press.

Bräuninger T., Brunner M., Däubler T. (2012) ‘Personal vote-seeking in flexible list systems: How electoral

incentives shape Belgian MPs’ bill initiation behaviour’, European Journal of Political Research, Articlefirst

published online: 12 DEC 2011, DOI: 10.1111/j.1475-6765.2011.02047.x

Carey J.M., Shugart M.S. (1995) ‘Incentives to Cultivate a Personal Vote: a Rank Ordering of Electoral

Formulas’, Electoral Studies, Vol. 14, No.4, pp. 417-439.

Carey J.M. (2007) ‘Competing Principals, Political Institutions, and Party Unity in Legislative Voting’,

American Journal of Political Science, Vol. 51, No.1, pp. 92-107.

Carrubba C;, Gabel M; Hug S. (2008) ‘Legislative Voting Behavior, Seen and Unseen: A Theory of Roll-Call

Vote Selection’, Legislative Studies Quarterly, Vol. 33, pp. 533-572.

Clark P.B., Wilson, J.Q. (1961) ‘Incentive Systems: A Theory of Organizations’, Administrative Science

Quarterly, Vol. 6, No. 2, pp. 129-166.

Close C. (2011) ‘Le factionnalisme partisan: imbrication ou concurrence des loyautés?’, paper

presented at the 4ème Congrès international des associations francophones de science politique, Section

Thématique 11 ‘Être partisan. Loyauté, prise de parole et défection dans les partis politiques et les

mouvements sociaux’, Brussels, 20th

- 22th

April.

23

Close C. (2012) ‘Parliamentary party coherence or factionalism: a consequence of candidate selection n

methods? A theoretical framework’, paper presented at the ECPR Joint Sessions of Workshops, Workshop

‘Party primaries in Europe. Consequences and challenges’, 10th

-15th

April.

Cox G.W. and Rosenbluth F. (1993) ‘The Electoral Fortunes of Legislative Factions in Japan’, American

Political Science Review, Vol. 87, No.3, pp. 577-589.

Cox G.W. and Rosenbluth F. (1995) ‘Anatomy of a split: the Liberal Democrats of Japan’, Electoral Studies,

Vol. 14, No.4, pp. 355-376.

Cox G.W. and Rosenbluth F. (1996) ‘Factional Competition for the Party Endorsement: The Case of Japan's

Liberal Democratic Party’, British Journal of Political Science, Vol. 26, No. 2, pp. 259-269.

Cox G.W., Rosenbluth F., Thies M.F. (1999) ‘Electoral Reform and the Fate of Factions: The Case of Japan’s

Liberal Democratic Party’, British Journal of Political Science, Vol.29, No.1, pp.33-56.

Di Virgilio A., Katô J. (2001) ‘Factionnalisme, 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, 51ème

année, No. 4, pp. 587-619.

Donovan M. (1995) ‘The Politics of Electoral Reform in Italy’, International Political Science Review/Revue

international de science politique, Vol.16, No. 1, ‘The Politics of Electoral Reform. La politique de la réforme

électorale’, pp. 47-64.

Dunleavy P., Margetts H. (1995) ‘Understanding the Dynamics of Electoral Reform’, International Political

Science Review/Revue international de science politique, Vol.16, No.1, The Politics of Electoral Reform. La

politique de la réforme électorale’, pp. 9-29.

Duverger M. (1951) Les partis politiques, Paris: Armand Colin.

Farrell D.M., Mair P., Muineacháin S.O., Wall M. (2011) ‘Courting, but not always serving: Perverted

Burkeanism and the puzzle of Irish TD under STV’, paper presented at University of British Columbia, May

2011.

Golembiewski R.T. (1958) ‘A Taxonomic Approach to State Political Party Strength’, The Western Political

Quarterly, Vol. 11, No.3, pp. 494-513.

Grynaviski J.D. (2004) ‘The Impact of Electoral Rules on Factional Competition in the Democratic South, 1919-

48’, Party Politics, Vol.10, No. 5, pp. 499-519.

Hazan R.Y. (1999) ‘Yes, institutions matter: The impact of institutional reform on parliamentary members and

leaders in Israel’, The Journal of Legislative Studies, Vol.5, No. 3-4, pp. 303-326.

Hazan R. Y. (2003) ‘Introduction. Does Cohesion Equal Discipline? Towards a Conceptual Delineation’, The

Journal of Legislative Studies, Vol. 9, No.4, pp.1-11.

Hazan R.Y., Rahat G.. (2011) ‘Parties, Politicians and Parliaments: The Impact of Democratizing Candidate

Selection on Party Unity’, Paper prepared for presentation at the Workshop on ‘Parties as Organizations and

Parties as Systems’, Vancouver, Canada, 19-21 May 2011.

Hine D. (1982) ‘Factionalism in West European parties: A framework for analysis’, West European Politics,

Vol.5, No.1, pp. 36-53.

Hirschman A. O. (1970) Exit, Voice and Loyalty. Responses to Decline in Firms, Organizations, and States.

Cambridge: Harvard University Press.

Hix S. (2004) ‘Electoral Institutions and Legislative Behavior: Explaining Voting Defection in the European

Parliament’, World Politics, Vol. 56, No.2, pp. 194-223.

24

Hix S., Noury A. and Roland G. (2005) ‘Power to the Parties: Cohesion and Competition in the European

Parliament, 1979–2001’, British Journal of Political Science, Vol.35, No.2, 209–34.

Hug S. (2010) ‘Selection Effects in Roll Call Votes’, British Journal of Political Science, Vol. 40, No.1, pp. 225-

235.

Janda K. (1980) Political Parties: A Cross-National Survey. New York: The Free Press.

Kam C. (2001) ‘Do Ideological Preferences Explain Parliamentary Behaviour? Evidence from Great Britain and

Canada’, The Journal of Legislative Studies, Vol.7, No. 4, pp. 89-126.

Karp J.A. (2006) ‘Political Knowledge about Electoral Rules: Comparing Mixed-Member Electoral Systems in

Germany and New Zealand’, Electoral Studies, Vol.25, No. 4, 2006, pp. 714-730.

Karvonen L. (2004) ‘Preferential Voting: Incidence and Effects’, International Political Science Review, Vol 25,

No. 2, pp. 203–226.

Karvonen L. (2010) The personalization of politics. A study of parliamentary democracies, ECPR monographs,

ECPR Press, London.

Katz R.S. (1980/2007) A Theory of Parties and Electoral Systems, Baltimore: The Johns Hopkins University

Press.

Katz R.S., Bardi L. (1980)’Preference Voting and Turnover in Italian Parliamentary Elections’, American

Journal of Political Science, Vol. 24, No. 1, pp. 97-114.

Key V. O. (1949) Southern Politics in State and Nation, New York: Knopf.

Klingemann H.D. (eds) (2009) The Comparative Study of Electoral Systems, Oxford: Oxford University Press.

Kohno M. (1992) ‘Rational Foundations for the Organization of the Liberal Democratic Party in Japan’, World

Politics, Vol. 44, No.3, pp. 369-397.

Köllner P., Basedau M. (2005) ‘Factionalism in Political Parties: An Analytical Framework for Comparative

Studies’, Global and Area Studies WORKING PAPERS, No.12.

Krehbiel K. (2000) ‘Party Discipline and Measures of Partisanship’, American Journal of Political Science, Vol.

44, No.2, pp. 212-227.

Leiserson M. (1968) ‘Factions and Coalitions in One-Party Japan: An Interpretation Based on the Theory of

Games’, The American Political Science Review, Vol. 62, No. 3, pp. 770-787.

Leyenaar M., Hazan R.Y. (2011) ‘Reconceptualising Electoral Reform’, West European Politics, Vol.34, No.3,

pp. 437-455.

Lijphart A. (1994) Electoral Systems and Party Systems, Oxford: Oxford University Press.

McAllister I. (2007) ‘The personalization of politics’ in Dalton R.J., Klingemann H.D. (eds), Oxford Handbook

of Political Behavior, Oxford: Oxford University Press, pp. 571-88.

Mitchell P. (2000) ‘Voters and their Representatives: Electoral Institutions and Delegations in Parliamentary

Democracies’, European Journal of Political Research, Vol. 37, No. 3, pp. 335-351.

Morgenstern S. (2001) ‘Organized Factions and Disorganized Parties. Electoral Incentives in Uruguay’, Party

Politics, Vol; 7, No. 2, pp. 235-256.

Norris, P. (1995) ‘Introduction: The Politics of Electoral Reform’, International Political Science Review / Revue

internationale de science politique, Vol.16, No. 1, ‘The Politics of Electoral Reform. La politique de la réforme

électorale’, pp. 3-8.

25

Norris P. (2002) ‘Ballot Structure and Legislative Behaviour’, Paper for the conference ‘Exporting Congress?

The Influence of the U.S. Congress on World Legislatures’ Jack D. Gordon Institute for Public Policy and

Citizenship Studies, Florida International University, University Park Campus, Miami, Florida, 6th-7th

December.

Nyomarkay J.L. (1965) “Factionalism in the Nationalist Socialist German Workers' Party, 1925-1926: The Myth

and Reality of the 'Northern Faction'”, Political Science Quarterly, 80, pp. 22-47.

Obler J. (1973) ‘The Role of National Party Leaders in the Selection of Parliamentary Candidates: The Belgian

Case’, Comparative Politics, Vol.5, No.2, pp. 157-184.

Ozbudun E. (1970) Party Cohesion in Western Democracies: A Causal Analysis, Beverly Hill, CA: Sage.

Panebianco A. (1988) Political Parties: Organization and Power, Cambridge: Cambridge University Press.

Pilet J.B. (2003) ‘Les jeux sont ouverts. Les effets du vote préférentiel dans les scrutins proportionnels de listes.

Analyse comparée : Espagne, Belgique, Finlande et Suisse’, Les Cahiers du CEVIPOL, No.1, Université Libre

de Bruxelles.

Pilet J.B. (2008) ‘Les nouveaux souffles dans l’analyse des systèmes électoraux’, Revue française de science

politique, Vol. 58, No.1, pp. 137-143.

Pilet J.B., Renwick A. (2011) ‘The Personalization of Electoral Systems: Theory and European Evidence’, Paper

prepared for presentation at the ECPR General Conference, University of Reykjavik, 25th

-27th

August.

Rae D.W. (1967). The Political Consequences of Electoral Laws. New Haven: Yale University Press.

Renwick A. (2011) ‘Electoral Reform in Europe since 1945’, West European Politics, Vol.34, No.3, pp. 456-

477.

Rose R. (1964) ‘Parties, Factions, and Tendencies in Britain’, Political Studies, vol.12, n°1, pp. 33-46.

Sartori G. (1976) Parties and Party Systems: A Framework for Analysis, New York: Cambridge University

Press.

Shiratori R. (1995) ‘The Politics of Electoral Reform in Japan’, International Political Science Review / Revue

internationale de science politique, Vol.16, No. 1, ‘The Politics of Electoral Reform. La politique de la réforme

électorale’, pp.79-94

Shomer Y. (2011) ‘Electoral Systems and Intra-Party Candidate Selection Processes: Influences on Legislators

Behavior’, Paper Prepared for the ECPR Joint Session of Workshops, Workshop ‘Cohesion, Dissent and Partisan

Politics in European Legislatures’, St. Gallen, 12th-17th April.

Shugart M.S. (2001) ‘Electoral “efficiency” and the move to mixed-member systems’, Electoral Studies, Vol.20,

pp. 173-193.

Taagepera R., Shugart M.S. (1989) Seats and Votes: The Effects and Determinants of Electoral Systems. New

Haven, CT and London: Yale University Press

van Vonno C.M.C. (2011) ‘MP decision-making: A sequential model’, Paper prepared for the ECPR Standing

Group on Parliaments Conference, Leiden University Institute of Political Science, 24th

-25th

April.

Zariski R. (1960) ‘Party Factions and Comparative Politics: Some Preliminary Observations’, Midwest Journal

of Political Science, pp.372-90.