The three arenas of personal leaders

Party leaders have become more powerful and autonomous actors in recent years, developing a direct and personal relationship with citizens. As anticipated in the United States (Lowi, Reference Lowi1985), the rise of the ‘personal leader’ seems to have occurred in many European democracies, both in old parties and in more recently formed parties with the ‘diffuse tendency that they be promoted and controlled by individual leaders’ (Calise, Reference Calise2012). Nevertheless, party leadership remains a theme that is largely overlooked in political science. Nearly three decades on, the view expressed by Sartori still remains valid: ‘the vital role of leadership is frequently acknowledged, nonetheless it obtains only a negligible status within the theory of democracy’ (Sartori, 1987: 171; Körösényi, Reference Körösényi2005). Moreover, although the substitution of the traditional party oligarchies with more personalized actors and roles has largely impacted upon all Western democracies over the last few years, both at the political and institutional levels, we can still register the lack of a comprehensive empirical investigation into the phenomenon of ‘personal leaders’. In the authoritative Handbook of Party Politics (Katz and Crotty, Reference Katz and Crotty2006), despite the relevance of the process of personalization in changing political parties, no specific chapters are devoted to party leaders. More recently, the Oxford Handbook of Political Leadership (Rhodes and t’Hart, Reference Rhodes and t’Hart2014) provides a significant contribution on party and electoral leaders, but notes that ‘in studies of both party organization and the electoral arena there seems to have been some reticence in recognizing the role of party leaders’ (Lobo, Reference Lobo2014: 362).

With the aim of providing a partial response to this lacuna, this article will focus on the personalization of party leadership by comparing Italy with other Western countries through a data set including ~500 party leaders in 13 democracies (see Methodological Appendix 1). More particularly, this process will be analysed in three relevant arenas of political parties.

The first regards the way the leaders are nominated, and leads to a focus on the introduction of direct election by party rank and file, as this tends to lend to the party president a personal mandate of sorts. This is broadly pertinent to a more general trend towards personalization of electoral campaigns, with their growing emphasis on leadership appeals (McAllister, Reference McAllister2007). Although cross-national comparative research has shown a significant tendency towards the adoption of more open methods in nominating party candidates for public office, especially through the use of primary elections (Hazan and Rahat, Reference Hazan and Rahat2010), political parties in several democracies have given their members a role in leadership selection too (Denham, Reference Denham2009; Kenig, Reference Kenig2009; Lisi, Reference Lisi2010; Cross and Blais, Reference Cross and Blais2012a; Pilet and Cross, Reference Pilet and Cross2014; Seddone and Venturino, Reference Seddone and Venturino2013). Despite of the fact that these developments may vary across countries and across parties, they may be regarded as ‘a clear trend in Western political parties’ (Wauters, Reference von Beyme2015: 218). This is particularly true in the Westminster world, for instance, by analysing 25 parties in Australia, Canada, Ireland, New Zealand, and the United Kingdom from 1965 onwards, Cross and Blais (Reference Cross and Blais2012b) found plenty of evidence for the expansion of the leadership selectorate. In this framework, Italy is very relevant in that the Democratic Party became the first European party to elect its leader through direct election. After the 2005 open election that acclaimed him as the leader of centre-left coalition, Romano Prodi underlined the great novelty of inclusive methods for the choice of party leader in an interview for the New York Times: ‘Never in Europe have we had primary elections – never, never, never. This will be the thermometer, the measure of a new involvement in national politics’ (Fisher, Reference Fisher2005). In fact, the event paved the way for important changes for Italian Democratic Party. Subsequently, three of its leaders were nominated though primary elections as follows: Walter Veltroni in 2007, Pierluigi Bersani in 2009, Matteo Renzi in 2013. Although the story of rightist parties is very different, the transformation of typical personal parties such as Forza Italia and Lega Nord brought about a new debate about the adoption of primary elections for choosing party leaders (Sandri et al., Reference Seddone and Venturino2014: 94). A more recent example of innovation is offered by Matteo Salvini, who beat Umberto Bossi, the historical Northern League leader, by running in a closed primary.

The second aspect of analysis refers to the centralization of party organization. Despite the fact that the creation of a direct relationship between leader and citizens facilitates the formation of a strong leadership (Katz, Reference Katz2001; Penning and Hazan, Reference Pennings and Hazan2001), party organizational changes can, independently of this, alter structures and processes in order to endow the leader with full prerogatives to determine the candidacies for all levels of national and local government, nominate or remove party managers and staff, and define policy lines in several domains. Centralization of party organization is one of the fundamental features of the personal party that has found in Forza Italia one of the best examples on the international stage (Calise, Reference Calise2010). Set up on the basis of the enormous financial power and the organizational resources of the Cavaliere’s business empire, this party has had a marked patrimonial nature since its birth, so that it could be interpreted as ‘the first European experiment in a large mass party made up out of a private enterprise’ (Poli, Reference Poli2001: 42). Moreover, the boundaries between Berlusconi’s business concerns and the new party remained unclear in terms of personnel, procedures, and culture throughout the party’s consolidation (Hopkin and Paolucci, Reference Hopkin and Paolucci1999). The phenomenon of personalistic parties has also spread in other European countries, through the diffusion of party organizations so centralized that they appear as ‘constructed or converted by an incumbent or aspiring national leader exclusively to advance his or her national political ambitions’ (Gunther and Diamond, Reference Gunther and Diamond2003: 187). Moreover, in Italy in the last 20 years, Forza Italia was largely imitated by both centre-right and centre-left political formations, from Italia dei Valori and Lega Nord to the Democratic Party re-founded by Matteo Renzi on a more personalistic basis. More traditional parties are not immune from such developments, as they may also be subject to the tendency to centralize decision- and policy-making powers in the hands of party executive committees and leaders, and ‘the extensive resort to co-optees and ex-officio members in the collective bodies by the leadership’ (Ignazi et al., Reference Ignazi, Bardi and Massari2010: 212). In this regard, specific structures created for managing and controlling party communication through old and new media are especially interesting, in that they may favour the creation of a direct relationship between the leader and the people.

Finally, the third area of leader-centred politics concerns the role of the party leader in the governmental and legislative arena. Although the leader performs several institutional duties, and is often a member of Parliament (MP), minister, or head of government too, this is the arena where they have experienced more problems during the last few years. In fact, while personalization of politics has strengthened the leader, it does not relate just to the higher levels of party organizations, but to all its levels and articulations, reducing the degree of their cohesiveness and deeply influencing party behaviour in representative institutions (Musella, Reference Musella2012). Thus, it remains to be investigated how and to what extent the party leader is able to realize his/her political programme in the new context. One may ask whether, while in many modern democracies the leader is more independent of the party, the party also feels more independent of the leader, and therefore is more inclined to rebel (Webb et al., Reference Wauters2012). So, while the leader becomes more autonomous and often more powerful, a sword of Damocles hangs over his head when he governs: the risk of ending up as a leader without a party to call his own.

The ‘direct election’ of party leaders

Several research studies have shown that the reinforcement of party leaders is often the result of the introduction of direct election by party rank and file that ended up attributing a sort of personal mandate to the party president. The pioneering cross-national work on leader selection (Gallagher and Marsh, Reference Gallagher and Marsh1987; Punnett, Reference Punnett1992; Marsh, Reference Marsh1993) has been followed by more recent studies that focussed attention on the expansion of the electorate for selecting leaders and candidates (Kenig, Reference Kenig2009; Wauters, Reference Wauters2009; Pasquino and Venturino, Reference Pasquino and Venturino2009, Reference Pasquino and Venturino2010; Hazan and Rahat, Reference Hazan and Rahat2010), on the norms, rules and behaviour surrounding political party leadership recruitment [in particular Politics at the Centr e by Cross and Blais (Reference Cross and Blais2012b)], and on the question of ejecting party leaders (Quinn, Reference Quinn2012). This literature has pointed out that, on the one hand, candidate and leader selection is a crucial political arrangement, which varies from country to country, that can be easily altered and can cause a transformation of behavioural patterns in a given political system (Hazan and Rahat, Reference Hazan and Rahat2010: 11), and on the other, that a general trend is developing towards the adoption of more inclusive methods. Even though until recently party leadership selection was an oligarchic intra-elite selection procedure, so that ‘despite factionalism in many political parties, national party leaders seldom faced an open contest’ (von Beyme, Reference Bickerton and Invernizzi Accetti1985: 226–232; Krouwel, Reference Krouwel2012: 257), currently an increasing number of parties allow members to participate directly in the selection of the party chairman.

As it can be noted in Table 1, while in 1965, 72% of selection procedures in 13 parliamentary regimes were accorded to restricted groups, today only 50% of them are given to party delegates, thus paving the way for the wider involvement of party members or supporters (Pilet and Cross, Reference Pilet and Cross2014). In fact, 26.8% of leader selections over the last 20 years have involved party members, and two cases of open primaries may be also registered (see Table 1). A paradigmatic case is represented by the UK political parties, traditionally controlled by their parliamentary caucus, from whose ranks the leader was usually recruited, and that are now moving towards a system in which votes are shared between parliamentarians and party members (Heppell, Reference Heppell2010; Bale and Webb, Reference Bale and Webb2014; Low, Reference Low2014). With further recent reforms to the procedure for electing the Labour Party leader, the balance has now plainly shifted in favour of the mass membership in the United Kingdom. Selection through a direct vote by party members has been introduced in several countries such as Belgium, Canada, and Israel: out of 31 political parties that changed their rules in the last four decades, 30 involved party members. The empowerment of party members has also expanded in Southern Europe, for instance, Portugal deserves particular attention to open procedures as all major parties have adopted more inclusive methods in the last few years, from the Socialist Party to the Social Democrats and the Social and Democratic Centre-Popular Party (Lisi, Reference Lisi2010). We also have two examples of open primaries for the choice of leader in this country, with the possibility of participation on the part of broadly defined ‘supporters’. Thus the evolution towards greater intra-party democracy in the process of leadership selection is largely proved, albeit to different degrees in contemporary democracies.

Table 1 The selectorate of party leaders (1965–2012)

Pilet and Cross (Reference Pilet and Cross2014): 227.

Italy is the country that has experimented most innovation in the field of leader selection in two very different directions. The first one is represented by the introduction of selection open to basic supporters, which revolutionized the traditional procedure of nominating the leader. Indeed, the party oligarchy had a sort of monopoly on leader selection in Italian political parties, with the rare exception of a few cases such as the Socialist Party of Bettino Craxi, who since the ‘Seventies developed an unchallenged supremacy over the party’s National Executive Committee’ (Massari, Reference Massari1989: 563). A hugely innovative method for leadership selection was adopted by the Italian Partito Democratico when it used open ‘primaries’ to select the general secretary for the first time in Europe (Bolgherini and Musella, Reference Bolgherini and Musella2010).Footnote ¹ This election also led to other consequences for the future of the party. On 14 October 2007, Walter Veltroni became party secretary, raking in 76% of the poll; in other words, three Democratic Party sympathizers in four were in favour of him – a plebiscite that, however, ‘only underlined the absence of heavyweight opponent candidates, as Veltroni’s resignation in 2009 and Pierluigi Bersani’s subsequent victory in the leadership primaries will show’.Footnote ² Indeed after 2 years, in the October of 2009, in a more competitive consultation, Pierluigi Bersani reached the top of the party with a lower consensus: he won half of the preferences, with a gap between him and other candidates of 16%. In 2012, Bersani was confirmed leader, this time of the centre-left coalition, but competitiveness in primary elections grew significantly. After obtaining a plurality but not an absolute majority in the first round, Bersani took only 62% of the votes in a run-off with the young mayor of Florence and his main rival, Matteo Renzi. He won by only 10% in the first round, and by 20% in the second round.Footnote ³ Finally, in the 2013 PD primaries, Matteo Renzi emerged as the leader with seven electors in 10 voting for him, so confirming the plebiscitary nature of the primary elections in Italy.

As can be noted in Table 2, Lega Nord used a closed primary election to select its leader in 2013 too. In a sort of ‘contagion from the left’, to borrow Duverger’s (Reference Duverger1951) famous phrase, for the first time we see an European right-wing party having recourse to an inclusive method like this. After Roberto Maroni announced that he was stepping down from the party leadership, a competition occurred between the party founder Umberto Bossi and the emerging leader Matteo Salvini, who was able to obtain a landslide victory with 82% of the vote in the ‘primary’ and >8000 ballots.

Table 2 The selection of party leaders in Italian political parties

Data on political parties from ParlGov database (Döring and Manow, 2012); own elaboration for variables on leader selection.

Yet, the presence of open primaries is not the most significant and widespread trend that may be observed. The second innovation in Italy regards the spread of personal parties, where the leader is endowed with a plebiscitary support or tends to create a political party on his own. Eight out of 12 Italian political parties included in Table 2 have been founded in the last few years the direct expression of a personal leader. The most famous example is Forza Italia, created by Berlusconi as a vehicle of personal political ambition. On the other side of the right-left continuum, Nichi Vendola founded Left Ecology and Freedom in 2009 from the former party faction he led in the Communist Refoundation Party. He was unanimously acclaimed president by the constituent assembly of 1500 party members. More recently, we find the case of the personal party created by Mario Monti after leaving the prime ministerial office. The list he launched for the 2013 national election, Civic Choice, was part of the centrist coalition, which broadcast his name: With Monti For Italy. Only in a tiny number of cases does selection of the leader depend on the decision of an oligarchic committee or on the deliberation of a restricted council of party affiliates.

The way in which the party leader emerges also influences his profile and how long he stays in office. A comparative analysis of the role exercised by the party heads before they became leader (Table 3) confirms high levels of seniority: more than 70% of them had been a government minister or MP in the period 1965–2013. Yet, during the Italian Second Republic, this traditional institutional path to party leadership changed as a result of Silvio Berlusconi’s direct passage from the world of management and publishing to politics in an incredibly short space of time. A novelty is represented by Beppe Grillo too, who worked as an actor and comedian from the 1970s, especially in theatres, sport centres, and city squares, after being ousted by Italian public television. In 2005, he created the blog, beppegrillo.it, where his political initiatives would find expression, and in 2009 he officially founded and became the leader of the Movimento5Stelle, the party that in the general election of 2013 emerged as the big winner, taking 25% of the vote. A different course was also taken by Matteo Renzi, who was the mayor of Florence when he became the leader of Partito Democratico and then the youngest person to become Prime Minister of Italy since unification in 1861. Nichi Vendola was president of the region Apulia when he became the leader of his party, whereas Bruno Tabacci was a member of the Milan city executive. Matteo Salvini reached the apex of Lega Nord through an non-traditional path too, as he was a member of the Milan City Council and a Member of the European Parliament (MEP), with no position at the Italian national level.

Table 3 Role of party leaders before their office

Own elaboration (Appendix 1).

Moreover, Italian party chiefs remain in power longer than their foreign colleagues, probably owing to the stricter control they exercise on party organization. The length of time in office is 71.3 months during the Second Republic, whereas the average time for the party leaders who ended their mandate in the period 1990–2013 in other Western democracies is 52.4 months. This result occurred despite the fact that the adoption of open primaries for the Democratic Party contributed to the rapid changes at the head of this party, with four leaders in only 4 years.

Other personal data of Italian leaders are more related to the political culture of the country. In comparison with other parliamentary democracies, Italy shows a very limited presence of female leaders, even in a common international context of strong under-representation of women: Giorgia Meloni is the only one in the last two decades, for a party which obtained only 2% of the vote in the most recent national elections, compared with approximately one party leader in six in other West European democracies. In addition to this, Italian party leaders are older than those in other countries, with an average age of 51.5, compared the overall mean of 48 in other countries during the same period.

The iron law of leadership

The idea that parties tend to concentrate power at the top of their organization dates back to Michels’ iron law of oligarchy (Michels, 1961[1911]), and was followed by Weberian reflection on charismatic leadership. Since the decline of mass parties, new typologies of political parties have emphasized the relevance of party leadership for electoral purposes, as demonstrated by ‘the contagion from the right’ defined by Epstein (Reference Epstein1967), in Kirchheimer’s ‘catch-all party’ (Reference Kirchheimer1966), or in Panebianco’s ‘electoral-professional party’ (Reference Panebianco1982).

In the last few years, this field has been the subject of major innovation, as the substitution of the traditional party oligarchies with more personalized actors and roles has impacted upon all Western democracies, both at the political and institutional levels. For instance, during the 1990s, following the model of Ross Perot’s Reform Party in United States and Tony Blair’s New Labour in United Kingdom, Forza Italia presented a new type of party political organization where personalization, professionalization, and centralization represented the keys to success that led Silvio Berlusconi to three general election victories, and that were soon variously taken up by both the centre-right and centre-left coalitions. Yet, the creation of the ‘Personal Party’ (Calise, Reference Calise2010), rather than being intrinsic to the Italian case, is part of a more general trend: as Gunther and Diamond (Reference Gunther and Diamond2003: 187) put it, personalistic parties are considered the most recent type of electoral party, which aim to provide ‘a vehicle for the leader to win an election and exercise power’. Throughout Europe, personalization leads to – and is the product of – a process of reconfiguration of party organization, which produces ‘the shift of intra-party power to the benefit of the leader’ (Poguntke and Webb, Reference Poguntke and Webb2005: 9). Candidate nomination, as one of the most significant and ancient prerogatives of political parties, represents one of the best fields of observation of the reinforcement of party leaders’ power. In fact, in the words of Schattschneider, ‘who can make the nomination is the owner of the party’ (Schattschneider, Reference Schattschneider1942: 101, cited in Norris, Reference Norris2006: 92). Focussing attention on the most relevant cross-national trends in the last few years, more inclusiveness in formal rules produces a sort of democratization of processes of selection, so that party members or registered supporters have more opportunity to express their view. If we consider, for example, the power of selecting candidates for national legislature, a clear shift may be identified towards widening participation within the selectorate. Thus, during recent decades ‘these changes are evident in the British Labour party, the ÖVP and SPÖ in Austria, the CDU and SPD in Germany, and by Fine Gael in Ireland’ (Bille, Reference Bolgherini and Musella2001; see also Norris, Reference Norris2006). Nevertheless, openness coexists with the transformation of party organization to the benefit of its leader. This happens owing to the effect of three interrelated processes. First, while participation in the selection processes increases, a ‘less intense (atomistic, unorganized, unstable) audience of party members is more likely to take cues from the highly visible party leadership’ (Hazan and Rahat, Reference Hazan and Rahat2006: 117). Second, empirical support has been found to show that, despite increasing inclusiveness in candidate selections, party leaders still retain many instruments to indicate preferred candidates or to exercise a veto over unwanted nominees (Scarrow et al., Reference Scarrow, Webb and Farrell2000). Finally, the introduction of more inclusive methods has the latent function of weakening the middle party activists, who constitute the only force able to challenge the autonomy of party leaders.

As we can observe in Table 4, primaries open to party members (internal or closed primaries) have become the most widespread procedure for the selection of candidates, with a cross-national trend towards the enlargement of the selectorate and a still relevant role for party leaders and national elites in conditioning or vetoing the list of candidates. UK political parties, whose parliamentary candidates have been selected traditionally by autonomous local constituency associations, have recently introduced changes in order to broaden participation in the candidate selection process through primary election, by a vote of all members at a hustings meeting or via postal ballot. These transformations were often viewed as an attempt by the leadership to sideline party activists in favour of the membership at large or specific party factions (Russell, Reference Russell2005). In Belgium, while most parties moved in the direction of a system with larger involvement of party members in the late 1990s, they introduced additional rules to ensure a certain control of the national leaders on the drafting of electoral lists (Fiers and Pilet, Reference Fiers and Pilet2006). Although Spanish parties opened selective procedures to party members, and the Catalan Socialists involved non-members registered as ‘sympathizers’ too (Hazan and Rahat, 2006), centralized mechanisms of control are exercised on nominations (Cordero and Coller, Reference Cordero and Coller2014). The same is true of Portugal where, for instance, the Socialist Party national executive has the formal right to nominate up to 30% of all ‘winning’ candidates and the leader intervenes in the choices made by the party local branches (Montabes and Ortega, Reference Montabes and Ortega1999). Norway represents an exception as the procedure for nominating candidates is regulated by the 1921 Act of Nomination, which forbids the leaders from directly intervening in the selection, with the consequence that selection procedures appear highly decentralized and with no role for national leaders (Krouwel, Reference Krouwel2012; Aylott et al., Reference Aylott, Ikstens and Lilliefeldt2014). On the contrary, Romania constitutes one of the more exclusive cases as in the post-communist period members’ involvement has remained very marginal, so showing an ‘interrupted oligarchical inertia’ (Chiru and Gherghina, Reference Chiru and Gherghina2012: 511; Gherghina, Reference Gherghina2013).

Table 4 Modalities of candidate selection in parliamentary democracies

Own elaboration (Appendix 1).

Regarding methods of candidate selection, Italy combines the two more extreme poles on the scale of inclusiveness: candidates may be selected by primaries open to all eligible voters or, in most cases, picked up by single leaders (Gallagher and Marsh, Reference Gallagher and Marsh1987; Rahat and Hazan, Reference Rahat and Hazan2001). This is the opposite of what occurred in the First Republic, when candidate selection methods occupied a more central position, as both local and central organs exercised an influence over the selection process, with a more relevant role for the national level in right-wing political parties such as Movimento Sociale Italiano or in charismatic ones such as Socialist Party (Lundell, Reference Lundell2004).

More recently, new forms of citizen engagement have been experimented with. One of the most significant innovations made by the Italian Democratic Party was the use of primaries to nominate candidates to Parliament, which were called ‘parlamentarie’. These took place on 29–30 December 2013 to choose about 85% (782 out of 918) of PD candidates for the two parliamentary houses, with the remaining candidates directly nominated by the PD leader (Musella, Reference Musella2014a). Primary elections were open to all supporters who had been involved in PD primaries or were willing to sign party programme documents: more than two million participants took part in the event. In the same period, open primaries for parliamentary candidates were also launched by Beppe Grillo for his Movimento5Stelle, through a consultation that was organized entirely online. Yet, ‘criticism was raised about the transparency of the method used, and the low number of participants, about 40,000 according to the MoVimento sources’ (Bordignon and Ceccarini, Reference Brusini2013: 12).

On the other hand, parties that approximate the personal party model typically adopted an approach of leader control over all party decisions; this has been typical of Forza Italia, and other parties that have emulated it in the last two decades. Indeed, Berlusconi has so completely embodied and directed his part that he chose the members of the party executive from among his associates and, although from a formal point of view regional co-ordinators have a say in nominations of candidates, he was endowed with full prerogatives to determine the candidacies for all levels of national and local government (Ignazi et al., Reference Ignazi, Bardi and Massari2010: 207). Italia dei Valori, created by Antonio Di Pietro, is probably the most similar to Berlusconi’s personal party, almost entirely dependent on the charismatic appeal of its founder-leader: until 2010 its statute was the only one, apart from Forza Italia’s, to attribute to the founder-leader the indisputable power to decide candidacies for national and European elections (Di Virgilio and Giannetti, Reference Di Virgilio and Giannetti2011). Lega Nord has also manifested the centralized nature of a personal party, presenting a small executive of 25 out of 30 people with no elected members and a wide prerogative of co-optation on the part of the leader. Bossi was able to field candidates at all elections and the party apparatus was ‘filled by means of appointments of the leader […] leading to the development of an internal elite hitherto recruited from among Bossi’s early political collaborators’ (Venturino, Reference Venturino2010: 180). More recently, its current leader Matteo Salvini confirmed that he is alone in the choice of the candidates (Brusini, Reference Webb, Poguntke and Kolodny2014). Thus, the analysis of the organization of the Italian political parties proves an evident process of centralization of the deliberative structures, realized through the direct attribution of nomination to the party leader or to very restricted committees, as well as the enlargement of the proportion of ex-officio members elected by the presidency. Even when the statutes give regional or local party branches the right to choose some of the members of the deliberative councils, the leader is able to decide or sponsor the co-optees. This is quite rare, though, for instance, in Forza Italia only 50 out of 400 members of the national council and six out of 50 members of the presidency committee are elected though a bottom–up process (Bardi, Ignazi, Massari, Reference Bardi, Ignazi and Massari2007: 289).

Thus, as far as the nominations for parliamentary elections is concerned, there are two strategies for allowing the leader to control the process of candidate selection, so protecting himself from ‘the party on the ground’ (Katz, Reference Katz2001). Party leaders may formally centralize the power around the party leader or empower the ordinary party members, or an even broader range of party supporters, at the expense of the power or influence of middle-level activists. Both these alternatives have been widely implemented in Italy, though with some reservations as to whether new modes of candidate selection are leading to effective democratization of internal party life.

The party leader in government

Party leaders also obtain central positions in key democratic institutions such as parliament or government. Indeed, while centralizing, or at least controlling, the most relevant functions of political parties such as the recruitment of the political class, they also maintain different roles in the wider political system. During the period 1965–2013, 40% of them belonged to national governmental teams, either as head of the executive (21.8%) or minister (22.3%). Moreover, a similar percentage of party leaders sat in national legislatures (41.5). Some 13.3% of party leaders occupy representative positions at the local (regional presidents, mayors, councillors) or supranational levels (MEPs or EU commissioners). (Table 5).

Table 5 Other institutional positions of party leaders

Own elaboration (Appendix 1).

In particular, in Table 6 we find the number of prime ministers who are also party leaders per country for the period 1991–2013. In most countries, the overlap between the two roles is an established rule, especially in political systems with a strong premier such as United Kingdom, Germany, and Spain. Executive chiefs lead their party in about two-third of the cases (>70% if we exclude Romania as an outlier in terms of the high number of prime ministers – 14).

Table 6 Prime ministers who are party leaders in parliamentary democracies (1991–2013)

Own elaboration (Appendix 1).

In Italy, a very important break in republican history is constituted by the fact that during the First Republic the party leader never became prime minister. This was due to the predominance of political parties in the Italian political system, which meant that secretaries of the major political parties did not enter the executive because their position was supposed to be higher than the premier’s. Apart from De Gasperi at the beginning of the republican history, only two further attempts to combine the offices of party leader and head of government were made in the Democrazia Cristiana, by Fanfani in 1958 and by De Mita in 1988, which, however, provoked strong resistance by some Democrazia Cristiana currents (Bettcher, Reference Bettcher2005; Vercesi, Reference Vercesi2013). A great novelty of the 1990s, after the introduction of the new majoritarian electoral rules, was that the leaders of centre-right and centre-left coalitions came to be regarded as candidates for prime minister. In 1994, Berlusconi became premier as the leader of the centre-right coalition: this was considered one of the more relevant innovations of the Italian transition, despite of the fact that centre-left premiers did not combine the roles of prime minister and party secretary during the same period (Campus and Pasquino, Reference Campus and Pasquino2006). Things took a further step forward after the 2013 national election, however, when three parties each received >20% of the vote, so creating a tri-polar Italian party system. Now, for the first time, the prime minister, rather than representing the head of the winning coalition, corresponded to the leader of the major party. Thus, Matteo Renzi, besides being Italy’s youngest ever Prime Minister, is also the first Italian party leader to become head of government by choice of a single party.

Prime ministerial office constitutes a great opportunity for party leaders to develop an emotional connection with voters by presenting themselves as the person in charge of executive action: a process that goes along with the citizens’ desire ‘to hold an individual accountable for government performance (or, occasionally, for the performance of the opposition), rather than an abstract institution or a political ideal’ (McAllister, Reference McAllister2007: 578). This is also the effect of new media strategies, which emphasize personal traits of leaders and bypass the intermediary role of parties in political communication, as we can note in the ‘mediatized leadership’ of Berlusconi and Sarkozy (Campus, Reference Campus2010). At the same time, the overlapping between the figures of premier and party leader may serve to distance the latter from his own party. A recent example of this phenomenon is represented by Matteo Renzi himself, who, on the one hand, has carefully managed his public image by developing direct communication with citizens through old and new media; on the other, he has tried to involve supporters and sympathizers as a counterbalance to party apparatchiks and internal opposition: ‘as he explained in speeches, his goal was to unite people around a common project more than it was to divide them along partisan or party lines’ (Bickerton and Invernizzi Accetti, Reference Bille2014: 25). From this point of view, when in government the party leader, rather than communicating with his party base, tends to speak to the majority of the citizens.

A critical element in this process is the centralized management of mass communication and the use of polling, both of which are representative instruments of the way the leaders interact with the public at large in the era of personalization. Following the US route, where the spread of the plebiscitary presidency meant that by 1960 ‘the pollster became a regular part of the inner core of each White House and presidential candidate organization’ (Lowi, Reference Lowi1985: 62), European prime ministers have institutionalized offices at the national level in order to monitor public opinion and to become more visible and popular in the eyes of the citizens in a sort of permanent campaign (Heith, Reference Heith2004; Reda, Reference Reda2013). However, popularity may also result in a boomerang effect for political leaders. By using television and polls to commune directly with the masses, bypassing mediating institutions such as parliament and the political parties, they begin to be regarded as the main governmental driving force and become the focal point of mass expectations. Yet, it remains to be investigated to what extent the party leader is able to realize political programmes, in a context in which personalization, while leading to a reinforcement of the figure of party leaders, also challenges party cohesion. With respect to the Italian case, it has to be observed that strong and visible party leaders have coexisted with high levels of party indiscipline, thus showing that the autonomy of the leader in the party’s internal decision-making procedures is not the same thing as the ability to direct MPs (Musella, Reference Musella2014b). Although after Tangentopoli the rise of the personal party identified with Forza Italia has implied both the creation of a new model of party organization and the adoption of a more plebiscitary idea of political representation (Calise, 2010 [Reference Calise2000]), two decades later the personalization of politics has shown another side, as it has involved not simply the higher levels of a party’s organization, but all levels. Indeed a form of personalization – which we can call micro- or parliamentary personalization (Musella, Reference Musella2012) – has affected Italian personal parties, reducing their cohesiveness and deeply influencing party behaviour in representative institutions (the party in public office). Data on parliamentary defections confirms a declining loyalty to party and coalition. A growing number of MPs decide to change parliamentary group during the legislature, and the quantity of dissenting votes in parliamentary activities goes far beyond the levels of other Western democracies (Giannetti and Laver, Reference Giannetti and Laver2009), so that there were thousands of rebellions of legislators against the positions of the groups to which they belonged. The clearest signal of the difficulty of party leaders to control their party/coalition while prime minister concerns the exponential use of the vote of confidence in order to avoid internal dissension in the process of approval of a law. It is not a coincidence that Matteo Renzi has reached a record in republican history for votes of confidence on parliamentary laws, with 18 votes in only 6 months.

Conclusion

Personalization has affected political parties in many contemporary democracies, leading to the prevalence of monocratic power over oligarchies. Italy plays a leading role in this scenario, presenting ‘completely new parties supporting personalities (Forza Italia), a deeply reshuffled party supporting alternative leaders (PD), a regional party with charismatic leader (Lega), and several local parties based on clientelism and patronage (UDC, Udeur, MPA)’ (Blondel and Thiébault, Reference Blondel and Thiébault2010: 175). Moreover, the list can be updated, adding the innovative case of Movimento5Stelle, a sort of personal party characterized by cybercratic centralism, with Grillo retaining exclusive rights to the party brand name and over every decision in terms of party organization and policy.

Italy seems a step more advanced than most in the three different arenas of the ‘personal party’. The enlargement of the selectorate for the choice of the party leader represents a cross-national trend, so that party members have been involved in decision-making processes traditionally anchored to party oligarchies. The push towards more open modes of selection is not only in response to the need to strengthen internal party democracy. In fact, party leaders are the ones to benefit most from these recent transformations, as they obtain consent directly from the party. This represents a massive change if we consider the party elites’ old prerogative of electing the highest party representative: it is no coincidence that in the tradition of mass parties the leader was often called ‘secretary’. Democratization of the nomination of party leaders occurred in Italy in two different and extreme ways. On the one hand, the Partito Democratico was the first party in Europe to open the leader’s nomination to electors at large, so allowing for the participation of millions of supporters. On the other, in personal parties, from Berlusconi’s Forza Italia to its more recent imitators, the leader founded his own party and was then confirmed leader by acclamation. In both cases, the intermediate levels of the party tend to be overlooked, in favour of a direct leader–follower relationship.

Moreover, personalization of political parties also means their organizational centralization. It is worth considering one of the most relevant – and constitutive – functions of political parties: the recruitment of the political class. Party leaders reinforce their control over political nominations as a result of two different processes. In new parties founded by their party leader, the president acts as a plenipotentiary figure by electing ex-officio members in deliberative organs, sponsoring elected members, or directly imposing his associates. As a consequence, candidacies for national as well as local consultation rest on the leader’s diktat, as occurred in parties such as Forza Italia, Lega Nord, or Italia dei Valori. In reformed mass parties consultation may be open to party members in order to favour grassroots participation. Once again, we find significant examples in Italy, where open primaries have been organized by Partito Democratico and Movimento5Stelle in order to choose parliamentary candidates for the 2013 general election, so offering some of the most inclusive procedures in Europe. The choice of more participative methods, however, may also be used in order to sideline the party middle strata in the process of political recruitment.

Italy is following the pattern of Western democracies in terms of correspondence of party leaders and prime ministers too. Thanks to the bipolarization of the party system brought about by the majoritarian electoral system reforms after Tangentopoli, coalitions started to nominate a candidate for premiership before elections. A path followed by Matteo Renzi who is currently premier and the secretary of the Democratici, the winning party in 2013 national electoral competition. This circumstance gives him the chance to present himself as the only person responsible for executive action, a daunting task in a period of political and economic crisis. In addition to this, he has inherited Berlusconi’s legacy as a great communicator, constructing his image in a very accurate way and developing a direct and personal relationship with citizens through old and new media. Yet, highly visible party leaders may have an Achille’s heel too. Although they reinforce their governmental role, on the basis of their personal mandate, they also experience a low degree of party discipline and cohesiveness in parliamentary activities. The leader becomes stronger only at the cost of fractionalizing and weakening his own party, a sad circumstance if one considers that ‘as presidential success advances arithmetically, public expectations advance geometrically’ (Lowi, Reference Lowi1985: 20). Here we find, in a nutshell, the most dangerous dilemma for personal leaders.