Intra-Party Democracy and Leadership Election Methods A Comparative Analysis of Political Parties in European Democracies

Transcription

1 Intra-Party Democracy and Leadership Election Methods A Comparative Analysis of Political Parties in European Democracies (Work in progress) Annalisa Cappellini PhD Student Annalisa.cappellini@kcl.ac.uk King s College London Prepared for the ECPR General Conference 7-10 September 2016 Prague

2 Abstract To what extent parties must be, should be and are internally democratic in terms of their internal distribution of power and decision-making practices? This is a crucial question yet still under researched by political scientists. This paper aims to address this research deficit and assess the extent to which a more inclusive and open leadership selection method increases the level of Intra Party Democracy (IPD). Considering the importance of parties to the realisation of democracy at the system level it is imperative to understand their internal workings. In normative debates the argument is sometimes raised that making parties more internally democratic may reverse the negative trend and tackle their crisis. The concept of IPD is gaining increasing attention given its apparent potential 'to promote a virtuous circle' able to link ordinary citizens to government and to contribute to the stability and legitimacy of the democracies in which these parties compete for power. At its core IPD is about the internal distribution of power within a political party and according to some scholars the most used instrument to pursue IPD is the enhancement of the inclusiveness of leadership selection methods. Even though some attention has been paid to the study of more inclusive methods for selecting candidates, the literature has been far less engaged in analysing the changes in terms of IPD generated by the adoption of more inclusive leadership election method. Furthermore, the field of leadership election is still under researched and the existing literature is both quite fragmented and lacks in extensive cross-national and cross-party comparison. Drawing from the tradition of participatory democracy this paper sheds a light on the very concept of IPD and its practical realisation in Europe by providing a snapshot of the current state of affairs. This paper classifies the existing patterns of party leadership election methods and IPD in 30 European democracies and examines the practices of 167 major parliamentary political parties, thus showing the links between different categories and the level of IPD existing in specific parties and countries. By presenting a cross-party, crosscountry large N analysis it offers a wider picture of the leadership election methods used across European democracies and tests whether the trends and the findings that have been so far identified by small n comparative studies can be applied and are valid on a larger scale. This paper provides a valuable empirical contribution which will help the development of theoretical and normative debates around IPD and its implications for contemporary political parties. 1

3 Introduction The aim of this paper is to shed a light on the methods used by political parties to elect their leadership and on the by providing a comparative snapshot of the current state of art, also trying to specify why certain methods are used. The primary goal is to assess whether political parties have started to engage in organisation reforms moving towards the enhancement of intra-party democracy (IPD), by analysing how they elect their leaders. To this end, this paper provides a comprehensive classification of the leadership election methods adopted by political parties across European democracies by carrying out a comparative analysis of these methods using the lenses of IPD. The office of party leader can often be considered as gatekeeper 1 to lead the country, hence the methods through which political parties elect their leaders can be seen as important political institutions, key institutional crossroad which deserve to be analysed (Rahat & Hazan, 2010: 10). Leadership election can also be seen as a good indicator of the democratic degree of the wider organisational dynamic of a party, as the way in which political parties select their [leaders] may be used as an acid test of how democratically they conduct their internal affairs (1988:1). Given the importance of parties to the realisation of democracy at the system level it is also important to understand their internal workings. Political parties are nowadays facing a legitimacy crisis; if it is true that citizens across Europe seem to be generally disengaged with their governing national institutions, not trusting them much, 2 political parties appear to be the least trusted political institutions in 23 of the 28 EU member state, as the latest Eurobarometer report has shown. In addition, the report indicates that in more than half of the EU countries political parties register levels of trust at least 5% lower than those recorded for other national institutions (Eurobarometer, 2015: 9). In order to tackle their legitimacy crisis and counter the decline in party membership, political parties seem to have changed their organisational structures by adopting procedures that allow the direct participation and involvement of their members in their policy development process (Gauja, 2013). This strategic shift is said to have happened also in relation to the party leadership election methods, across political parties in Europe. However doubts remain on whether this is actually the case and hence the question arises regarding how political parties elect their leaders nowadays. 2

4 Do parties use more inclusive, less centralised, hence democratic method to elect their leaders? Do political parties allow the direct participation of their members in the leadership election procedures thus fostering IPD? Is it possible to find patterns in terms of leadership election methods used by political parties according to their ideological positioning and geographical area of development? These are the core research questions that this paper aims to answer. A comprehensive comparative study of the leadership election methods used by political parties at this moment in time across all European democracies is still lacking, this paper aims to fill this gap. By understanding how they currently elect their leaders and by providing a new comprehensive classification of these methods this paper will provide a new reference framework for further analysis and comparisons. The paper proceeds as follow: in the next section the relevant literature and theories on the relationship between leadership election methods and IPD will be briefly reviewed; through this overview the main hypotheses to be tested in the empirical analysis will be derived. The following section will see the data, the methods and the overall research design outlined. In the last section the empirical analysis will be presented and the hypotheses investigated so to answer to the main research questions. Before proceeding with the literature review and with the outline of the main hypotheses, a clarification regarding the main concepts at the base of the analysis is however necessary. First of all, in order to avoid a conceptual stretching a distinction needs to be made between those processes aimed to select candidates and those aimed to elect party leaders, at the core of this paper. Although these two distinct processes may, to a degree, overlap, they need to be treated - and defined - differently, in virtue of their very different nature and scope. Hence, also a terminological differentiation needs to be applied. While a selection procedure implies a two stage process, where someone is selected to run as a candidate in a specific election, a party leadership election, on the contrary, does not requires neither implies a second stage, a person is elected to hold the specific office of the party leaders. It follows that also the methods should be defined differently (Pasquino and Valbruzzi, 2016), so to avoid to 3

5 inappropriate generalisation those aimed at electing party leaders will be called Leadership Election. 4

6 Literature Review The idea of Intra-Party Democracy emerged in the era of mass parties (Duverger, 1961), at its base lies the assumption that at the system level it requires the central collective actors of representative democracy, that is political parties, to be internally democratic. This concept has gained increasing attention in recent years given its apparent potential to promote a virtuous circle able to link ordinary citizens to government and to contribute to the stability and legitimacy of the democracies in which these parties compete for power (Scarrow, 2005). Through this lens it can be read, for example, the call made by the Parliamentary Assembly of the Council of Europe on member states to ensure that the legislative framework promotes the implementation by political parties of internal party democracy principles. 3 On the same line the European Commission for Democracy through Law - the Venice Commission - issued its Code of Good Practice in the Field of Political Parties and it identified to reinforce political parties internal democracy and increase their credibility in the eyes of citizens, thus contributing to the legitimacy of the democratic process and institutions as a whole and fostering participation in political life as its explicit aim. If democracy on one hand has been defined as constituted by participation (Cohen, 1971), on the other hand it is seen as a political method, a certain type of institutional arrangement for arriving at [ ] decisions (Shumpeter, 1962 [1942]; 242). According to Bobbio, however, democracy is without any doubts government by the people (Bobbio, 2005[1988]: 25) and Barry points out that, the normal way of understanding democracy is to suppose that it refers to internal distribution of power within a political unit (Barry, 1974: 493). Thus the fundamental meaning of democracy appears to involve the equal participation of citizens in the governing of a state, however links and balances between participation and institutions are essential as participation without institutions is chaotic [ ] institutions without participation are an empty exercise at best (Johnston, 2005: 5). Political parties are among the most crucial institutions in these processes as they represent one of the principal channels for political participation, playing the distinctive role of transmission belts between voters and policy outcome in the electoral arena ; as Sartori put it, they should be an instrument, or an agency, for representing the people by expressing their demands (1976: 24). Although political scientists keep quoting Schattschneider by saying that, political parties created 5

7 democracy and [ ] modern democracy is unthinkable save in terms of parties (1942:1), the reality shows that citizens are becoming increasingly disaffected with political parties. 4 Among scholars concerns are growing on whether parties are still capable of performing the functions traditionally allocated to them; Peter Mair used a quite striking opening line for his last book by saying: The age of party democracy has passed. Although the parties themselves remain, they have become so disconnected from the wider society, and pursue a form of competition that is so lacking in meaning, that they no longer seem capable of sustaining democracy in its present form (Mair, 2013:1). Few scholars have underlined that the dimension facing the real crisis is the party on the ground (Cross and Katz, 2013: 65). It seems that being aware of the loss of a stable electoral base and of the dropping of membership rates, political parties began to address the attention to their internal distribution of power and started enhancing their members role in terms of policy making, candidate and leadership selection. This trend has been interpreted by many as a shift towards an improvement of Intra-Party Democracy (Scarrow, Webb & Farrell, 2000). On one hand IPD is seen as relating to the participation and voice of parties ranks-and-file, on the other hand to the responsiveness of parties to the national electoral arena; this distinction can be traced back to Duverger (1961) and Michaels (1911). An agreement, however, has not been reached yet on what it means to be internally democratic, and as the very notion of democracy, Intra-Party Democracy is still a contested concept both in terms of the meaning and operationalisation (Cross & Katz, 2013). At its core, however, IPD is about the internal distribution of power within a political party (Cross, 2013: 100) and it seems to require at least some element of participation by the party on the ground in the (s)election of the leading members of the party central office (Katz, 2014, 188). Candidate selection and leadership election is indeed one of the main defining functions performed by political parties; indeed according to Sartori, a political party is any political group that presents at elections, and is capable of placing through elections, candidates for public office (1976: 63, 64). Choosing candidates and leaders is a strategic process with 6

8 important implication on the nature of the party itself as it might reflect its democratic ethos (Cross, 2013: 115). In addition, some scholars have also argued that the phenomena of presidentialization or personalisation of politics (Poguntke & Webb, 2005), has increased the importance of the role played by party leaders in modern representative democracies. While it is usually lamented that candidate selection receives considerably less scholarly attention than the study of electoral systems, it is fair to say that, party leadership (s)election is the poorest cousin of all (Cross & Blais, 2010: 4). The literature in this field is quite fragmented, usually focusing on a single case study, 5 single-party, 6 or single-country study, 7 it is skewed in focus, (Kenig 2009) given that the majority of these studies deal with British parties. The special issue of the European Journal of Political Research edited by Marsh (1993), represents a notable exception in this respect, presenting a collection of interesting findings related to six different countries. However those studies, despite providing an insight of the functioning of the leadership election methods in the respective cases, do not present an actual cross-national comparison. Limited comparative works have been presented by Punnett (1992) and Courtney (1995); while LeDuc (2001) undertook a systematic comparison between three English speaking parties. The issue of Party Politics, 8 publishing the contribution of LeDuc, can be easily defined as the cornerstone for the development of this research field; its importance is indeed declared and an initial analytical framework for cross national comparison is proposed. 9 Hazan & Rahat (2010), by presenting a thoughtful and detailed analytical framework for evaluating the functioning and consequences of leadership (s)election methods based on four dimensions, have given a considerable push to the enhancement of the discipline. The major challenge for the research in this field is to bring it close to the one of the electoral system; to do so more cross-party and cross-national empirical studies are needed. Astudillo (2012) has focussed specifically on the participation of the rank-and-file members in the leadership election process in advanced industrial democracies and produced an intense study of 23+1 countries and 134 parties. As well as the new research project the Comparative Study of Party Leaders (COSPAL), is studying party leaders in fourteen countries from a comparative perspective covering a period of study from 1965 until This group has so far published two major studies; one in 2014 describing, on a country by country basis, the leadership election procedures in thirteen countries (Pilet and Cross, 2014) and a more recent one presenting a comprehensive 7

9 comparative study of the leadership election in fourteen countries and 107 parties across almost five decades (Cross and Pilet, 2015). This paper aims to fit within this research framework by presenting a comparative analysis of how political parties in contemporary European democracies currently elect their leaders. 8

10 Conceptual Analysis and Operationalisation IPD is here defined as involving the equal distribution of power within a political party (Cross, 2013: 100), where the decision-making processes involve the participation of the party on the ground. It is here operationalised on the basis of the degree of inclusiveness of the leadership election processes, the dependent variable, that can be considered as a general measure of the level of IPD (Cross and Katz, 2013). The operational definition of party on the ground - or in other words the specification of those eligible to participate in the election process - has been expanding over time, both in numbers and intensity, where a less intense connection and less durable engagement is required. (Katz, 2014; Cross and Blais, 2012; Pennings and Hazan, 2001). Thus, through these lenses the party leadership contest organised by the Italian Democratic Party in 2009 can be read as representing, an extreme case of a more general trend (Katz, 2014:188). This trend sees political parties shifting the locus of their decision making practices to a political space that at its extreme goes beyond the party on the ground to encompass the whole citizenship. But is this really the case? Have party elites renounced to their power choosing prerogatives and handed it over - at least formally - to their members and supporters? The reference point for assessing intra-party democracy is party leadership election institutionalisation as measured by the adoption of formal rules for the election of party leaders (Sindre, 2016). Hence, the focus needs to be on the presence - or lack - of formal and codified rules in official party documents, like party Statutes and Constitutions. It is important to underline that formal institutions might co-exist with informal procedures and the former may actually say little about the reality of the leadership election practice (Sindre, 2016, Helmke and Levitsky, 2004: 726). However, it is crucial to analyse the content of the formal institutions - here defined as the formal rules regulating the leadership election that are contained in party statutes and constitutions - before looking at the informal ones. After having done so it will be the possible to assess whether discrepancies exist between the two and to what extent parties stick to their rules. 10 The codification of formal rules for the institutions and functioning of political parties is becoming a major topic in contemporary democracies. While few countries have already introduced in their national legislation laws regarding the requirement for political parties to adopt both statutes and codified rules for 9

11 their leadership election and general functioning, others are still addressing this issue. 11 Who is the Leader? Given that the objective of this paper is to provide a general picture, a comprehensive classification of the leadership election methods used by political parties across European democracies, the focus is solely on those processes aimed to elect the formal and actual leader of any given political party. To identify the leader in a political party is not always an easy task; for example in the Christian Democratic Union of Germany (CDU) we can find a Chairman, that is elected by the Congress, and a Secretary General - which is elected on a proposal of the Chairman - responsible for the daily managing of the party. 12 However, the formal (and real) leader remains the Chairman. Different parties use different names to define their leaders, thus this role might be played by the Party Chairman, the Party President or by the General Secretary. Regardless to the different label used by different parties the attention is on how - and if - parties elect their leaders. A further clarification needs to be done; we should refrain from simplistically assuming a universal personalisation and presidentialisation of political parties (Poguntke and Webb, 2005; Astudillo, 2012). In some parties for example the locus of power is shared by collective forms of leadership, as it happens in several Green Parties, like the Danish Red-Green Alliance that has a collective leadership of 21 people and one spokesperson. Other political parties present a dual form of leadership, like in the case of the Belgian Ecolo that has two co-chairs. Regardless to the type of leadership chosen by each party here it has been analysed studied and classified how the leadership is elected; thus defining as the party leadership the political figure(s) that according to the formal and codified rules utilised and accepted in any given political party has the main decision-making power, along with the responsibility to set the political agenda and externally represent the party. Who is the Party Leader Electorate? Thanks to the previous theoretical work carried out by LeDuc (2001), Kenig (2009) and Hazat and Rahat, we know that one of the most important dimension that needs to be taken into account when addressing the study of leadership election is the composition of the body that elects the leader. The very classification of different type of methods can be indeed built according to this dimension; however, a conceptual and terminological clarification is once 10

12 again needed. Previous literature has always referred to this body as a selectorate, Rahat & Hazan (2001) represent its composition as a continuum, where at one end there is just one person, identifying the most exclusive type of selectorate, and at the opposite extreme there is the whole electorate, representing the most inclusive one. While agreeing on the conceptual starting point of this elaboration, in the light of the previous differentiation suggested, it is important to apply the same criteria, hence to define the body that elect the party leadership as Electorate. It s however imperative to do not confuse the meaning of term Electorate, as used in this context, with the more general one indicating the entire electorate of a Country which has the right and duty to vote for national elections. In addition, the idea to position at one pole of the continuum the entire national electorate is here challenged; the rationale being that even the most open and inclusive leadership election method, that allows also supporters and sympathisers to take part in the leadership election, 13 requires the potential voter to formally show his support and commitment to the principles and values of the party by signing a declaration, and/or paying a fee, and/or registering to the party. To present the different composition of the electorate on a continuum is a helpful exercise, but some modification in terms of its composition are necessary; it is also useful to present a categorical classification that can be then positioned on the continuum. Moreover, given the ultimate scope to measure and assess the different level of IPD across European democracies, by using as an indicator the type of leadership election method adopted, it naturally follows the necessity to create a scale that will be helpful in providing an Index of IPD across countries. The proposed scale goes from 0 to 5, with the value of 5 assigned to the most inclusive - and democratic - method and the value of 0 to the least and it is summarised in the inclusiveness continuum pictured below: 11

13 At one pole of the continuum there is the most inclusive of all methods, represented by the Direct Election Method (DEM). As it has been mentioned earlier, it is important to operate a clear distinction between the procedure aimed to elect the party leader - at the core of this paper - and the one aimed to select candidates. Although the literature has come to generally accept, and use, the term primaries for both, in the light of their very different nature and scope this generalisation is here rejected and a differentiation - also terminological - of the two is called. We should also refrain from adopting the term one-member-onevote (OMOV), used by some authors as synonym of party primaries (Kenig, 2008: 243), as it fails to capture the core of the issue, that is not the equality of the votes but the direct access to it (Astudillo, 2012). Hence, it is more appropriate to define as Direct Election Method the procedure to elect the party leadership that involves the direct participation of all party members, and sometimes supporters and sympathisers of the party - where obviously each vote has also the same weigh. For the sake of simplicity in this category every party that has adopted some forms of Direct Election is included. A further differentiation, within this category, might have been drawn, both in terms of openness of the electorate (all party members versus party members + sympathisers and supporters) and in terms of electoral arena. For example, a distinction between those leadership election carried out in party assemblies or congresses, where the delegates are formed by the whole membership - mainly due to the relatively small size of the party or of the country; and those organised with polling station and/or postal ballots across the country where the result is then ratified by the specific party organs like extraordinary congresses. An example for the first type can be provided by the parties in Luxembourg, where every members has the right to participate and vote in specific party congress aimed to elect the party leader. On the contrary the second type can be represented by the leadership election rules established by the Spanish PSOE, which allows all its members to vote in the leadership contest organised via electoral polling stations across the country. The members vote is then ratified by the party congress. The rationale behind the choice is twofold; first of all at the core of the overall classification there is the aim to provide macro categories to support the elaboration of a wider picture of the state of arts of IPD across European democracies. This will be done by taking as a 12

14 benchmark the degree of inclusiveness of members in the leadership methods according to the formal and codified rules of the political party. The second reason lies in the expectation that overall DEM is not widely spread in political parties in European democracies - contrarily to what the current literature seems ti suggest - hence, if the final results of this investigation will confirm the expectation it won t be because of the very strict criteria applied to the category. At the opposite extreme of the continuum - as the least inclusive method - there is the one where no election actually ever takes place and no formal rules or practices are established within the party. In this category there are parties that are led by their funders and that don t present any kind of rules, neither for potential future leadership elections. Hence, this category is defined as Personalistic with No Rules (PNR), with a value of 0. One step further up on the scale there is a similar category but presenting a crucial difference; the presence of formal rules within the party. Although these rules have not been used so far, the possibility for a leadership election to take place exists. Thus, it can be defined as Personalistic with Rules (PWR) with the value of 1. With the value of 2 there is the category that can be named as Party Elites (PE); here there are political parties that carry out their leadership election behind closed doors. Only selected party elites are allowed to take part in this practice and usually they are MPs members of the Parliamentary Party Group. Although this was a common practice in the last century the research shows that it has been fading, nowadays less than an handful of parties uses this method. The literature so far has suggested a distinction between the category of Party Elite and the one defined as Parliamentary Party Group, (Kenig, 2008); however, considering the relative small number of parties - at least across European democracies - that use these types of methods and in virtue of the predominance of the involvement of the MPs, it seems appropriate to merge both types in one single category - identifying the leadership election that are in the hands of the party elites. The next category is an hybrid one, represented by what could be defined as Mixed Method (MM) with a value of 3. In this category the leadership election is usually split in multi stages that see the involvement of both party elites and party members. Although there are just two parties across European democracies using this method, it is nevertheless important to specify it, in virtue of its dual character. Despite the involvement in principle of all members this category does not score more than 3, due to the fact that the party elite still 13

15 control the first and fundamental stage of the leadership election, that is the decision regarding the final candidates. In addition, as the most recent British Conservative Election has shown, it might easily happen that the election process ends after the very first stage, with new leader not having to go through the public scrutiny and vote of all party members. The last category is represented by the Selected Party Agency (SPA), with a value of 4. This is the largest of all of the categories, where there is the highest number of parties fitting in; party delegates and/or representatives are in charge of the leadership election and the procedure is carried out in specific arenas like congresses, assemblies, conferences or conventions. Party statutes set the formal rules regarding the electoral procedures in terms of the occurrence, the number of party delegates allowed to vote and how the delegates are selected or appointed. Usually the leader is elected for a fixed term, normally ranging from 2 to 4 years and it is not uncommon for the statute to set a limit in terms of number of mandates for every leader. The number of delegates/representatives allowed to vote varies from party to party and it is usually set according to the size of the party and the overall size of the membership, thus ranging between a few hundreds to a thousand of voters. It is important to underline that these categories might be regarded as pure types, hence according to each party few differences may occur. These categories present a good effort to reach a fair balance and compromise between the broadness that a large N study might require and the obvious need of specification for an in-depth study. Thus, this categorisation will be helpful in identifying trends across parties in European democracies and providing a comprehensive overview of the current state of art. Unfortunately there are few parties part of the study that do not fit in any of these categories, the main reason being the difficulties in finding reliable and certain information on how they effectively organise and carry out their leadership election. These parties thus are comprehended within the Other/Not Clear category, that is positioned outside the continuum due to the difficulties in finding a justification for any possible value. 14

16 Research Design In order to offer a robust contribution to the debate on IPD and Leadership Election Method the focus will be on providing a comparative classification of the methods adopted by political parties represented in national parliaments across European democracies. The primary goal, thanks to a cross-party, cross country large N analysis, is to present a comprehensive comparative overview of the state of arts of leadership election methods and to provide a detailed classification that can be used for future comparative studies. Whilst both large N and small n approaches have pros and cons, virtues and pitfalls, a large N analysis can provide the reference framework that can be applied to in-depth case studies (Ebbinghaus, 2005). A large N approach can also be useful in testing on a wider scale hypotheses formulated through specific case studies; moreover it might help in assessing whether a trend that sees parties transforming their internal organisation towards the enhancement of IPD actually exists. In addition, to the best of my knowledge a comparative study focused on the analysis of formal and codified rules of party leadership election adopted by the political parties represented in national parliaments across European democracies is still lacking, this paper aims to fill this gap. Given that this paper wants to focus on the analysis of the formal institutions of party leadership election, the data have been mainly gathered from official party regulations, party statutes and constitutions - when these were accessible. In addition, information have been collected from previous case studies and empirical research and from online newspapers and reports from different countries. Country and Parties Selection To select the countries it has been adopted the Freedom House Index and the European countries labelled as free democracies identified. It is important to mention that this study did not want to focus solely on the countries member of the European Union member states, but on the liberal democracies in Europe. Thus a total of 30 countries were selected and namely Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia Lithuania, Luxembourg, Malta, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland, The Netherlands, The United Kingdom. 14 For each country, the political parties that have succeeded in guaranteeing a relevant representation in their national parliaments during the 15

17 last national election have been taken into account. A total of 167 political parties have been analysed and are part of this study; having as the main objective to analyse how the party currently elect their leaders, the time frame of my research covers the last leadership election organised by each party. For every party it has been researched the presence of official party statute/constitution/regulation and analysed the content of the formal and codified rules related to the leadership election procedures. To best of my knowledge this is the first study covering all the parties that in Europe democracies have gained a relevant representation in their national parliaments. 15 The division into geographical sub-regions has been done by following the composition used by the Statistic Division of the United Nations Secretariat. In table 1 below there is the overview of the countries and parties studied with the summary of the results both in terms of leadership election methods and presence of party statutes and/or regulations. 16

18 Intra-Party Democracy and Leadership Election Methods Table 1 Overview of the countries and parties researched in this paper.16 The basic distribution of the dependent variable - inclusiveness of party leadership election in any given party - has 5 valid categories on the continuum and one category - the not clear one - that is not comprehended in the continuum. Means for electorate variable have been calculated in order to elaborate the IPD index and create a distribution of all the European democracies part of this study. This variable has then been crossed with the categorical independent variables - party family affiliation and ideology, geographical region, age of the party. Hypotheses 17

19 The analysis of how party leadership election varies across different parties in European democracies is still at an embryonic stage, especially from a wider comparative perspective. One of the question that still seems to be left unanswered is whether, in the wake of the crisis of legitimacy and support that parties have been experiencing, they have decided to tackle this crisis by adopting similar organisational strategies. Or in other words, whether regardless to their genetic model and background they have implemented similar organisation reforms in relation to their leadership election procedures. It is not uncommon to find in the current literature a somehow implicit assumption regarding the presence of a generalised shift towards direct election methods across western and industrialised countries, regardless to the ideological connotation and positioning of the party (Astudillo, 2012). This paper aims to clarify this point by analysing whether the ideological components of political parties do matter or not for their organisational decision in terms of leadership election method. In order to do so this paper will go beyond the dichotomous classification of left and right wing parties to analyse instead political parties according to a classification based on their affiliation to transnational party families. The rationale behind this decision lies first of all in the absence of a clear theoretical framework and explanation regarding how and when to classify parties according to the left/right dichotomy. (Astudillo, 2012). The data provided by Bakket et al. (2015), along with the continuum offered by Benoit and Laver (2006) could have been applied; however when looking closely at this classification it emerges that for example the British Liberal Democrats are classified as left-wing parties and the Dutch 66 as right-wing parties, while in terms of their European Affiliation both parties are part of the Alliance of Liberal and Democrats in Europe (ALDE). Thus, by adopting this approach and using as a reference of their ideological positioning their rational choice of their affiliation with transnational party families this research will be based on a more reliable and less biased information. In addition, the affiliation to a transnational party family does not just represent a strategic choice, but it is perceived by the Party and its members as a crucial ideological decision that somehow highlights and defines the ethos of the party itself. As the example of the Italian Democratic Party (PD) shows; one of the most important ideological debate - and struggle - that occurred within the PD in the wake of its creation was regarding its European affiliation. The party, was indeed born by the merge of the Democrats of the Left and the Daisy. The first one originally affiliated to the Party of European Socialists (PES) and the 18

20 second member of the Alliance of Liberal and Democrats for Europe (ALDE). In order to solve this deadlock and to find a European and transnational house for the newly born political party, the PES party first agreed to change the name of the group represented in the European Parliament into Progressive Alliance of Socialists and Democrats (S&D) so to welcome the PD into its ranks. Lately the PD, during one of its national assembly meeting, voted to join the European Party PES, which now includes the words socialists & democrats in its symbol. Hence, it naturally follows the importance to analyse the organisational dynamic of the leadership election through this lenses and to see whether a correlation between party affiliation and leadership election method exists. Given the focus of this research on parties represented within European democracies, it naturally follows to look at the affiliation of parties to European families. However, considering that the samples includes countries that are not member of the EU and parties that are not represented in the EU Parliament, the point of reference won t be the EU Party Groups but specifically the Transnational Party families. This research will take into account both the cases where parties are full members of these families and those where they are associate or observers - due to the likelihood of their shared vision and ideology. One issue encountered while adopting this approach is that there are parties not affiliated with any transnational party. Nevertheless, to study the party ideological positioning according to the rational choice of its affiliation provides a more objective analysis. 17 One alternative could have been to apply the traditional distinction of the eight party families that have been drawn following the development of parties in Western Europe. However, also this approach presents pitfalls; first of all these labels might be difficult to apply to every party - especially when taking into account new political parties - and secondly it might be biased with subjectivity from the researcher perspective. To the best of my knowledge, a widely accepted and complete classification - according to the traditional distinction of the eight party families - comprehending all the parties that are part of this study has not been yet formulated. 18 By looking at the affiliation of the 167 political parties under investigation the following main transnational party family have been identified; the Party of European Socialists - Socialists and Democrats (PES), the Alliance of Democrats in Europe (ALDE), the European People s Party (EPP), the European Green Party, the Party of the European Left, the Alliance 19

21 of European Conservatives and Reformist (AECR); few minor families have been grouped into one single category labelled as others, both because of their similar character and due to the relative small number of parties affiliated to them. By looking closely at these families and at their official ideological positioning and stances it appears possible to collocate them within the traditional ideological framework. Thus, the PES Family can be identified as the one representing the socialist, social-democratic and labourist ideological stances, and in terms of the left-right spectrum it could be positioned on the centre-left; the ALDE family can be associated with the Liberal-Democrat values and positioned at the centre of the traditional spectrum; the EPP can be positioned at the centre-right of the spectrum representing the Christian-Democratic values; the European Green Party is the one representing all the Green oriented stances and values, while the Party of the European Left is home for parties coming mainly from a communist, left and non-socialist tradition, and it can be positioned on the leftfar left side of the continuum. At the opposite end it can be found the AECR, formed by Conservatives and Reformists parties. Drawing from the so-called convergence hypothesis, which has been put forward by few scholars (Katz and Mair 1995, 2009; Scarrow, Farrell, and Webb 2000, Kittilson and Scarrow, 2003) and building on the findings of the latest comparative analysis carried out by Lisi, Freire and Barbera (2015), this paper wants to examine whether is it possible - or not - to validate the convergence hypothesis across political parties affiliated to a European Party Families. Or in other words, whether it is possible to assess that parties in European Democracies, have adopted similar leadership selection methods regardless to their ideological affiliation. In the light of previous studies and findings, the expectation here is that parties affiliated to the same family present significant similarities in terms of leadership election methods. The first hypothesis (H1) thus argues that parties affiliated to the same family have adopted similar leadership election methods, while the subsequent hypothesis (H1a) claims that parties affiliated to the PES family are more likely to have adopted inclusive hence democratic methods to elect their leaders. The rationale behind this hypothesis lies in the assumption that parties affiliated to the PES family mainly come from the mass-party model and from the social-democratic tradition, historically more inclined to include grassroots members in their decision-making processes. However, in order to test the validity of the converge hypothesis, it is important to analyse it also from a geographical 20

22 perspective, hence to assess whether regardless to the geographical proximity parties have followed the same trajectory and have adopted the same leadership election method. The second hypothesis (H2) argues that parties in the same geographical area have adopted similar leadership election methods, thus it will be likely to find a concentration of the same method in the same region. In their comparative study of the five principal English-speaking Westminster countries Cross and Blais, 19 formulated the hypothesis about the likelihood for new parties to adopt more easily rules that allow a greater role for rank-and-file members, compared to older parties (2012). The scholars concluded their investigation by positively confirming the hypothesis, finding the rationale in the wish of the new parties [ ] to appear more democratic and less hierarchical than the old-line opponents (2012: 145). Is it possible to confirm this finding going beyond the study of the English-speaking Westminster countries and thus assessing that new parties are more likely and keen to adopt more inclusive, hence democratic, leadership election methods? The third hypothesis (H3), built upon Cross and Blais s findings, argues that in any given country is more likely to find parties with inclusive leadership election methods in newer parties compared to the old-line opponents. To do so it will be compared the leadership election method adopted by parties that have been established after the year Only the cases of newly born parties, not those resulting from the merge between preexisting parties or those that have simply changed their name, will be taken into account. In addition, a control based on the party affiliation to the above mentioned families will be used. One last dimension fundamental to focus on is the gender equality in political leadership. It appears important to evaluate the effect on descriptive representation that a specific election method might have. The decision to concentrate on the representation of women can be justified as the best empirical proxy for representation overall (Rahat, Hazan, Katz, 2008: 669). Moreover, one of the most crucial issues for contemporary democracies is the lack of it, as in advanced industrial society the female leadership style tends to be more effective and the feminisation of leadership style are closely linked with the spread of democratic institutions (Inglehart, Norris, and Welzel, 2002: 343). Hence, to analyse the presence of female party leaders and whether different type of leadership election methods might facilitate, or hinder, the likelihood of women assuming the highest office in a party is a useful 21

23 exercise. In their study on the consequences produced by a more inclusive candidate selection methods Rahat, Hazan and Katz (2008) showed that high levels of inclusiveness result in low levels of representation (2008: 673) thus arguing that, enhanced inclusiveness in political parties can have a negative impact [ ] on the overall health of democracy (2008: 674). Is it possible to draw the same conclusion when looking at the party leadership election? Do more inclusive methods produce less representative party leaders? Is it more difficult for women to be elected as party leaders when the election is open to the participation of all members and/ or supporters of the party? The expectation here is that Rahat, Hazan and Kats s findings might just partly apply to the case of leadership election, to the extent that more inclusive leadership election method might not have a significant positive impact on the female representation. However, this paper argues that more centralised and closed methods do not allow any female representation. The fourth hypothesis (H4) thus states that the lower the degree of inclusiveness of a given leadership election method and the less likely it is to find female representation in the party leadership. 22

24 Intra-Party Democracy and Leadership Election Methods Empirical Analysis The first hypothesis (H1), argued about the likelihood for parties affiliated to the same transnational party family to have adopted similar leadership election methods, H1 is tested in Table 2 below. The most used method across all families is the one involving a selected number of members, the SPA method. Out of the 167 parties that have been investigated 91 use SPA to elect their leaders; within the EPP family 26 out of the 40 parties use this type of method while of the 32 political parties that are part of the PES family 16 use SPA. The predominance of this method can be seen also among the ALDE, the Greens, strikingly 11 out of the 15 parties members of the European Left or United Left use the SPA method. It seems possible to affirm the validity of the convergence hypothesis, contrarily to the original expectation; regardless of their genetic model, parties appeared to have opted for a method that, while presenting a relative high degree of inclusivity, is not completely democratic and inclusive yet. Table 2 Party leadership election methods according to the party affiliation to the European party family. The leadership election is indeed mainly entrusted in the hands of a selected number of appointed and/or (s)elected group of members, participating to specific party fora, like Party Congresses or Conventions; this fora meets on a regular basis to elect the new party leader - 23

25 or confirm the incumbent one. This type of leadership election appears to be often just a formality as the party leader is rarely challenged (Pilot and Cross, 2014). One other main feature of this type of leadership contest is that it can be subject to a high degree of influence and conditioning by the party elites. While with Direct Election Method is rather uncommon to see episodes of coronation, as this may prompt negative media exposure (Kenig, 2009), leadership contests carried out via SPA are likely to present this feature. Quite often congress delegates are simply called to ratify a decision - or agreement - previously achieved among party elites. Especially in those case where there is only one candidate and/or when this candidate is essentially chosen by the former party leader, a coronation is the only choice left. One illustrative example is represented by the Spanish People s Party, affiliated to the EPP party family, where the leader Mariano Rajoy was elected by the 14 th party congress in 2004 after the previous leader José María Aznar proposed him as his successor (Astudillo and Garcia - Guereta, 2006). Despite the fact the formal rules indicate in three years the length of the leadership term, Mariano Rajoy hasn t yet been challenged and he is still the leader of the Party. Some parties in Sweden have clearly stated in their statute the cheerleader role of the congress delegates; for example in the Moderate Party (M) affiliated to the EPP, the leadership election has to be by acclamation, while the delegates of the Centre Party (C) have simply to ratify a decision that has already been taken by a selection committee. Another interesting example is represented by the German Party SPD, affiliated to the PES and led by Sigmar Gabriel since After an electoral debacle the then leader Franz Müntefering decided to step down and Gabriel run unchallenged; at the party convention in Dresden he received 94.2 percent of the delegates votes, 472 out of 501 valid votes. Gabriel was then confirmed in 2011 by the party delegates with 91.5 percent of the votes and again in 2013 with 83.6 percent of the votes, still unchallenged. 20 As Detterbeck and Rohlfing have noted, it is almost an unwritten rule for German parties to settle conflicts before the party convention and then present just one candidate (Pilet and Cross, 2014). The same dynamic can be indeed found within the Gruene - the German Green party - affiliated to the European Green Party. Although it is true that the German legally entrenched rule is election by delegates at a national convention, parties would have the option to hold a direct election involving every member, and then call the delegates to ratify the results. 21 However, it seems that party elites 24

26 in every German party, regardless to their background and ideological connotation, prefers to maintain a strong grip in the Party Leadership Election dynamic, and involve - some - members just to ratify a decision previously taken. As the data show across parties in European democracies the SPA method is the most popular, regardless of their ideological positioning as its adoption is traversal to the affiliation to different party families. However, by taking a closer look at the chart we can notice that the concentration of the Direct Election Method can be found within parties belonging to the PES family. The hypothesis (H1a) claimed that parties affiliated to the PES family are more likely to have adopted inclusive hence democratic methods to elect their leaders; results presented in Table 3 below confirm it. The transnational party family that has the greatest percentage of the Direct Election Method is the PES, confirming the hypothesis that parties coming from the mass party and social democratic tradition are more keen to entrust their members with decision-making powers. Conservatives and far right parties, on the contrary, are the less keen to give powers to their members in terms of leadership election. Thus, despite having used a different framework to define the ideological position of parties, this research confirms Astudillo s findings (2012); parties coming from the social-democratic tradition are more prone to use DEM, while both parties coming from the conservative and communist tradition are the more reluctant to move towards this kind of model. 22 In sum, it is possible to conclude that ideology indeed matters, not all party families are equally prone to empower their members in the leadership election procedure. 25

27 Table 3 Direct leadership election method across European party families. As the German example has hinted the convergence hypothesis might be disproved in terms of geographical proximity; if parties in the same country are likely to have adopted the same type of leadership election method, it follows that this might be true also in terms of geographical area. H2, that argued about the likelihood to find a concentration of the same method in the same European region, is tested in table 4 below. By looking at the distribution of leadership election methods used by parties across European regions we can see that in three regions out of four the most used method is the one involving Selected Party Agencies, that is the election happens in specific party fora composed by (s)elected members. The majority of parties both in Northern and Eastern Europe have opted for this type of method. Eastern Europe however, is also the region presenting the highest number of parties falling within the two Personalistic categories. Specifically these are parties represented in the national parliaments of Bulgaria, Czech Republic, Poland and Slovakia. These four countries were part of the communist block and are considered new democracies, as they are all part of the so-called Third Wave of Democratisation that began in Southern Europe in the late 70 s and then spread to the Eastern area following the fall of the Berlin Wall in 1989 (Huntington, 1991). Studies on post-communist countries are quite straightforward in stating the general 26

28 weakness of party organisation, despite the formal presence of members, branches, congresses, etc; many scholars have assessed their low degree of institutionalisation and their dependence on a narrow coterie of leaders (Mair and van Biezen, 2001; Lewis, 2001; van Biezen, 2003, 2005). A striking example is provided by Slovakia, where out of seven parties analysed, four fit within the Personalistic without Rules categories. These parties not only are still guided by their founders, but they neither have formal rules for the leadership election nor general party statutes and/or regulations. Interestingly enough the transnational party families which these parties are affiliated to are the AECR, EPP and the Alliance for Peace and Freedom, a far right, ultra-nationalist and Eurosceptic European party. Two other Slovakian parties fit within the Personalistic With Rules categories; from a formal point of view they have rules governing the party leadership election but they never used it. It is worth mentioning in this respect the case of the Direction-Social Democracy Party (SMER-SD), which is affiliated to the PES family and that is currently in power with 49 out of 150 seats in the National Council. Originally named Direction (SMER), the party emerged as a breakaway from the Party of the Democratic Left (SLD) in November 1999 under the leadership of Roberto Fico, then one of the most popular politicians of SLD. In 2005 the Party merged with few other smaller parties coming from the social-democratic tradition thus adopting the current name of SMER-SD. The party statute states that the Congress, that is the highest body of the party, is in charge of the election and the removal of the party leader; however, no mention of length of the leadership term(s) is made and Roberto Fico is still the unchallenged leader of the party after 17 years. These data thus show that political parties in post-communist states appears to be still weak and leader-dependent organisation, without many open contests and/or collegial and open decision-making practices (Enyedi & Linek, 2008). In addition, as the case of the SMER-SD suggests, this feature is traversal to any ideological positioning. Finally it is worth noting that Eastern Europe is the only region with just one party that, according to its formal rules, has a Direct Election method, the Alternative for Bulgarian Revival (ABV). This newly formed Bulgarian party, created in 2014 by former President Georgi Parvanov, saw at its inaugural congress the delegates confirming Parvanov as the party leader. However, according to the party statute the leader will have to be elected by the direct vote of all party members every two years; it will have to be seen whether the 27

29 leadership of ABV will follow its formal rule. The analysis of the party leadership election rules adopted by parties in Eastern Europe seems to suggest that new democracies struggle most - or are less keen - to move towards an enhancement of IPD by opening the election process to their members. Can be said the same for the other new democracies in other European regions? Southern Europe is the other area that shows a high number of parties scoring poorly in terms of IPD when analysed through the lens of their leadership election procedures. Six parties fit within the PNR category, and they are evenly distributed among the countries, they don t present a clear pattern in terms of the age of the democracy. For example in Italy - a formally long established democracy - is possible to find a party that meets the PNR criteria, namely the Five Stars Movement (5SM), founded in 2009 by the comedian Beppe Grillo. This party, which in principle advocates direct democracy, is structured around the figure of the leader, 23 whom has absolute power in terms of the policies and the direction of the party, to the point that he has can expel members and elected representatives of the Movement in public offices. The party - that prefers to be called Movement - doesn t have a statute and clear rules for the leadership election, it claims however to be based on a democracy of the internet in virtue to the fact that the candidate for public offices are selected by an online voting of the members. Rosanna De Rosa has wittingly defined it as cybercratic centralism. This could be considered as the 2.0 evolution of the democratic centralism, common feature of the communist parties, characterised by the presence of strong internal discipline and a centralisation of all steering functions (2013). However, in the case of the 5SM, decision-making procedures, far from being decentralised, are totally embedded in and overlapping with the figure of its leader (Miconi, 2014). Thus, it does not surprise that the 5SM MPs tried to fight the approval of a new Party Law that would require political parties to formally establish rules governing their functioning, financing and the internal democracy, including the leadership election procedures. 24 These observations also confirm that the empowerment of members to elect party leader has not occurred only in Northern and Western Countries; Southern Europe is indeed the region that, after the Western one, has more parties with Direct Election Method. 25 Five countries out of the nine investigated in the South have parties that have opened up their leadership 28

30 election procedures to all their members - and in quite few cases also to their supporters. Especially Spain and Portugal, two countries listed among the new democracies, are cases worth mentioning. Portugal counts three parties fitting within the DEM category; two major parties, the Socialist Party (PS) - affiliated to the PES and the Social Democratic Party (PSD) - affiliated to the EPP and a small one, the Left Block (BE). This party is affiliated to the Party of the European Left, and has a collective form of leadership formed by six members of the so-called permanent commission and one spokesperson, elected with the participation of all party members. PS was the first one to introduce the direct election of party leader in 1998, while the PSD started the internal debate over the organisation reform in the 90s and it finally adopted the method in 2006 (Lisi, 2009). In Spain the direct of involvement of members was, until recently, only used for the selection of candidates and in some cases for the election of regional and local leaders. However, the picture has now changed as two parties have empowered all their members to elect their leaders; the Spanish Socialist Worker s Party (PSOE) - affiliated to the PES and the newly form Podemos - affiliated to the Party of the European Left. While Podemos made of the direct involvement of all members and supporters one of its founding principles, PSOE has recently introduced this practice. The Socialist Party used it for the first time in 2014 to elect the current leader Pedro Sánchez, which won with 49% of the members votes against the two other contenders Eduardo Madina and José Antonio Pérez-Tapias (Barberà & Teruel, 2015). This observation thus goes partly against what previous studies have shown, that is that the organisational configuration of parties in new democracies strongly differed from parties in long established ones, especially in terms of autonomy of party leaders and lack of opportunities for members to influence the election of the leader (Biezen, 2003). While this argument still stands for political parties in new Eastern European democracies, it fails relatively to parties in new Southern European democracies, thus suggesting that geographical proximity might indeed play a role in the organisation and institutional development of political parties. One last note worth giving is the confirmation that Western Europe seems to have moved in bulk towards a more inclusive leadership election method (Krouwell, 2012). Germany is indeed the only country that does not have parties with Direct Election Method; as it was 29

31 mentioned beforehand this might be due to the specific legislation in terms of party organisation of the country, although it is worth noting that the German Party Law does not prevent parties from adopting this type of method. Table 4 Party leadership election methods across European regions H3 argued that in any given country is more likely to find parties with inclusive leadership election methods in newer parties compared to the old-line opponents. By looking at the year of foundation of the parties in the sample it results that, excluding the Italian Democratic Party, 46 out of the 167 parties have been established from the year 2000 onwards. Of these parties only five have introduced DEM. Thus it appears clear that regardless to the specific country we might look at, new parties are not that prone to empower members. More interestingly is to note that 20 of the new parties fit within the Personalistic categories and 16 30

32 of those within the less inclusive of all the categories, the PNR. Moreover, these new political parties are not concentrated in one specific area but they are spread across European democracies. Hence, not only H3 is disproved but it also seems that new political parties have been created around the figure of a strong leader - the very founder of the party - which has total power on the development and agenda of the party. The party, in most cases, is also named after the leader. Thus a question arises, is emerging across Europe a new form of political party based on a sort of business-firm model (Mazzoleni, 2016) which sees the absolute predominance of the leader and the emptying of the role of members? Although more research is need in this area, these preliminary data suggest so; a new cleavage seems to start emerging in terms of parties organisation across Europe. On one side new political parties like Podemos, which are built around the very concept of IPD and which promote a new idea of inclusivity that goes beyond the traditional grassroots segment to encompass the whole society, are emerging. On the other side, there are business-firm parties like the Austrian Team Stronach or the Dutch Freedom Party, which are identified with the figure of leader itself and serve not other purposes but providing him a platform. Furthermore all these parties are built around the figure of a male leader, thus providing some key data to analyse the last hypothesis of this research. H4 argued that the lower the degree of inclusiveness of a given leadership election method and the less likely it is to find female representation in the party leadership; H4 is analysed in Table 5. Female party leaders are clearly still a rare phenomena in politics, with men maintaining a strong grip on the leadership of the party. Women are more likely to assume the highest office of the party when the electorate is composed by delegates, while the opening of the electorate towards all members doesn t seem to have a significant effect on their representation. It is worth noting though the concentration of female party leaders in Scandinavia countries; both Denmark and Norway have major parties leaded by women, and in the Norwegian case they have been ruling for over a decade. For example, Erna Solberg has been leader of the Norwegian Conservative Party (H) since 2004, while Siv Jensen, has been leader of the Norwegian Progress Party (FrP) since 2006, with her leadership unchallenged since she took over from her mentor and predecessor Carl I. Hagen. Although the rare cases of female party leadership appear to be evenly distributed among parties, both in terms of size of the party and ideological positioning, one aspect strikes the attention; the 31

33 majority of charismatic and strong female party leaders are at the head of centre right and right wing parties. Apart from the above mentioned examples of the Norwegian parties, two important cases are represented by the German Christian Democrats Party (CDU), which is affiliated to the EPP and that has been guided by Angela Merkel since the year 2000, and the French Front National, affiliated to MENL and leaded by Marine Le Pen, daughter of the founder. Thus the suspicion arises that to analyse the female party leadership path from an ideological perspective might pave the way to new fascinating research findings. Table 5 Gender of the party leader according to the leadership election method. 32

34 Intra-Party Democracy Index As it was originally stated, the aim of this paper is to present a comparative snapshot of the level of IPD across European Democracies measured according to the inclusiveness of the leadership election method formally adopted by political parties in any given country. The result of the analysis, with the IPD ranking is showed in Table 6 below. The ranking shows the overall IPD level scored by each European democracy according to the leadership election method adopted by each political party. It is based on the analysis of the formal institution of the Leadership Election Method adopted by every party represented in the National Parliament and/or European Parliament. The Index is calculated according to the values assigned to each category on the continuum that has been previously presented, with N=total number of parties in a country; ri is the rank of the particular party; the sum runs over all parties in the country N, as per the following formula: As the figures below shows, just one single country reaches the highest scores; the best performing of all is Belgium, scoring 5 out of 5. This is not surprising given that the first wave of reforms towards an enhancement of IPD started in 70 s and by the late 90 s all main Belgian political parties had empowered their members to choose the leader. At the opposite end there is Slovakia and again this result doesn t come as a surprise; out of the seven Slovakian parties investigated only one party, the Slovak National Party (SNS), has adopted, and used, formal rules to elect its leader. The remaining six parties have never elected a leader, being still guided by their founders. By looking at the overall figure it appears clear that IPD can t be considered a general trend in European democracies; the party on the ground haven t been widely empowered by the party in central office ; it s the latter that still has a strong grip on the crucial dimension of the leadership election. A final note regarding the suggested classification is due. Although its basic configuration, 33

35 Intra-Party Democracy and Leadership Election Methods containing only the dimensions of inclusiveness of the leadership election methods and the presence of formal rules regulating these elections, this index and the approach behind it, might provide the starting and reference point for more sophisticated comparative analysis of IPD. In fact, with more dimensions combined an IPD Index could give a further insight on the general status and the development of contemporary democracies. Table 6 IPD Index of European democracies.26 34

36 Preliminary Conclusions This paper focused on the analysis of the Leadership Election Methods adopted by political parties across European democracies. The main aim was to assess how political parties formally elect their leaders and provide a classification of these methods through the lenses of Intra-Party Democracy; thus to clarify whether a shift towards IPD has occurred. To provide a general overview a big N approach was adopted and a cross-party, cross-country analysis of parties in European democracies was performed. The leadership election methods have been classified according to five different categories and positioned on a continuum representing the level of inclusivity of each method. By providing a new classification and conceptualisation of the leadership election methods used by political parties this paper provided a new framework for analysis that can be used for future comparative studies. By filling a gap in the existing research, this comprehensive analysis of the methods used in this moment in time by political parties in European democracies, brings an important contribution to the enhancement of the discipline. The IPD Index presented wants to be a starting and reference point for future comparative analysis, where a diachronic approach can also be applied to understand how and if the love of IPD has changed across countries. The preliminary finding of this research show that the Direct Election Method is not as spread as previously suggested; despite very low criteria were set to fit within the category. Among the sample of 167 parties distributed across 30 countries, only 24% of political parties have introduced some forms of DEM to elect their leaders, as table 7 below shows. Although the leadership election procedure does not happen behind closed doors any longer, and the party elite are not formally solely in charge, the most popular method, SAP, in few cases appears to be a mere coronation. The party delegates entrusted to elect the leader often are simply called to ratify a decision that has already been taken by the party leadership. 35

37 Table 7 Total breakdown of the party leadership election methods used by the 167 parties researched in this paper. The analysis proved that, when looking at the likelihood the adoption of more inclusive election methods ideology does matter. Although the convergence hypothesis seems confirmed relatively to the wide adoption of SPA method, the data have also shown that parties affiliated to the PES are more prone to empower their members to elect their leaders. The convergence hypothesis, on the contrary was disproved in terms of geographical proximity; the geographic area of development of the parties and the path taken by neighbouring countries plays a role in the type of method adopted. Thus, we have seen that Eastern Europe seems to be immune to the shift towards IPD and on the contrary it sees a strong concentration of personalistic parties. If Western Europe has been confirmed to be the area with the highest concentration of DEM, as every country but Germany has at least on party using a Direct Election Method, also Southern Europe appears to be moving in that direction, especially when looking at the neighbouring Spain and Portugal. The data also showed that new political parties are generally less prone that their old line opponents to foster IPD, while a new business firm model of party organisation seems to be emerging. Finally, we have observed that female party leadership is still a rare phenomena and the 36