Microsoft Word - PDF_Issue_13_1_6.docx Italian Political Science, VOLUME 13 ISSUE 1, MAY 2018 © 2018 Italian Political Science. ISSN 2420-8434. Volume 13, Issue 1, pp. 82–94. * This article is the result of work conducted fully in common. However, Eugenio Salvati specifically worked on Sections 2 and 5, while Michelangelo Vercesi specifically worked on Sections 3 and 4. The Introduction is a joint effort. We are grateful to two anonymous reviewers of Italian Political Science for their valuable com- ments and suggestions. Contact Author: Michelangelo Vercesi, Leuphana University of Lüneburg. E-mail address: michelangelo.vercesi@leuphana.de Party Organizations and Legislative Turnover: Signals of an Unstable Parliamentary Class? Eugenio Salvati UNIVERSITY OF PAVIA Michelangelo Vercesi LEUPHANA UNIVERSITY OF LÜNEBURG Abstract Parliamentary turnover is an important question in political science due to its connection to salient topics like elite circulation and parliamentary roles, legislature institutionalization, executive-legislative relations, and the quality of the policy-making process. In this article we contribute to the debate by focusing on the Italian case from a longitudinal perspective. The 2018 general election was conducive to a relatively high level of turnover. This level is roughly equivalent to two previous electoral rounds. However, the empirical evidence shows that the conditions usually thought to account for another two exceptional elections seem not to apply to 2018. After introducing the literature on legislative turnover and arguing that this topic is worth tackling, we stress the use- fulness of the Italian case for comparative research and we provide a picture of our case study, presenting the theoretical puzzle and research questions. Subsequently we tackle this issue, claiming that the usual explana- tions fit only partially with our case. In particular, we argue that neglected intra-party factors can be a viable explanatory alternative. The article ends with a brief discussion of the findings. 1. Introduction arliamentary turnover is an important question in political science due to its con- nection to salient topics like elite circulation and parliamentary roles, legislature institutionalization, executive-legislative relations, and the quality of the policy- making process. Despite its importance, it is a relatively underdeveloped research field. Starting from the two influential works by Polsby (1968) and Rosenthal (1974), the great bulk of studies on turnover have focused on single cases (Moncrief et al. 1998; Samuels 2000; Manow 2007; Kerby and Blidook 2011); only rarely have they attempted to engage with cross-country analyses (Matland and Studlar 2004; Kuklys 2013; Gouglas et al. 2017). The main exception is the stream of studies on subnational parliaments. Follow- ing the US tradition, these works have provided within-country comparisons (Squire 1998; Moncrief et al. 2004; Heinsohn and Freitag 2012; Heinsohn 2014). In this article we contribute to the debate by focusing on the Italian case from a lon- gitudinal perspective. The 2018 general election was conducive to a relatively high level of turnover. This level is roughly equivalent to two previous electoral rounds, with rates ‘definitely deviant for a context of democratic continuity’ (Verzichelli 2009: 61). P WHO’S THE WINNER? AN ANALYSIS OF THE 2018 ITALIAN GENERAL ELECTION 83 However, the empirical evidence shows that the conditions usually thought to account for another two exceptional elections seem not to apply to 2018. This suggests that we should look for further explanatory factors. After introducing the literature on legislative turnover and arguing that this topic is worth tackling, we stress the usefulness of the Italian case for comparative research and we provide a picture of our case study, presenting the theoretical puzzle and re- search questions. Subsequently we tackle this issue, claiming that the usual explanations fit only partially with our case. In particular, we argue that neglected in- tra-party factors can be a viable explanatory alternative. The article ends with a brief discussion of the findings. 2. Legislative Turnover: Definition and Determinants Parliamentary turnover is linked to two facets of the life of legislatures: institutionaliza- tion and professionalization. Polsby (1968) has defined the turnover rate as the central element that can help measure the degree of institutionalization achieved by a parlia- ment. According to him, long legislative careers and low turnover rates play a central role in the strengthening of a legislature. This is why a stable group of incumbents within leg- islatures is fundamental to structure and routinizing organizational behaviours. Moreover, they promote specialization and differentiation, which are considered crucial features of efficient legislature. Professionalization refers to a condition in which MPs can conceive of their parlia- mentary activity as their main occupation. Professionalization can be measured by three indicators: the MPs’ remuneration and benefits, length of legislative sessions, and avail- ability of resources (staff, working spaces, funds, etc.) (Squire 1988, 1992, 1998). Often, subnational legislatures are less professionalized, being thus inherently much more ex- posed to high turnover rates (Squire 1992; Moncrief et al. 2004; Heinsohn and Freitag 2012; Heinsohn 2014). Following Matland and Studlar (2004: 92), legislative turnover can be defined as the proportion of membership that changes from one election to the next. A new leg- islature is composed of re-elected MPs, incoming MPs elected for the first time and former MPs who come back after an interruption (Francois and Grossman 2015; Sal- vati 2016a). Legislative turnover defines the level of renewal or continuity of parliamentary elites and membership stability within an institution (Squire 1998). Additionally, it in- directly tells us something about the criteria that parties follow to select their personnel and/or party gate-keeping strength. Studies on parliamentary turnover can be gathered into two groups: the micro-level approach looks at individual motivations for exiting, while the macro-level approach takes institutional factors into account. Francois and Grossman (2015) have suggested three levels for studying turnover: the legislature (macro-level), the parliamentary party (meso-level), and, finally, the individual (micro- level). According to the two scholars, micro and macro levels are not per se adequate for studying turnover. Indeed, only the party level could account for party strategies, party elites’ circulation, and party selection processes. All in all, a concomitant consideration of both party and legislative factors seems to be the most fruitful approach. SALVATI and VERCESI, Party Organizations and Legislative Turnover 84 The nature of the exit of MPs from parliament can be either voluntary or involun- tary. Voluntary exits stem from individual choices, determined by cost/benefit analyses of running for office (Squire 1988; Hall and Van Houweling 1995; Matland and Studlar 2004; Kerby and Bildook 2011). On the other hand, an involuntary exit may be due to party gate-keeping or failure to get re-elected (Somit et al. 1994; Moncrief 1998; Matland and Studlar 2004). The most important source of involuntary turnover is electoral volatility, that is, the degree of change in voting behaviour between elections (Matland and Studlar 2004; Moncrief et al. 2004; Manow 2007; Heinshon 2014): higher volatility increases the likelihood of higher turnover and substantial changes of parliamentary seats (Rosenthal 1974; Matland and Studlar 2004). Overall, we can expect to find higher turnover rates when there is institutional instability, parties and party systems char- acterized by lower institutionalization, and high electoral volatility (Matthews 1984; Gouglas and Maddens 2017). A second determinant of involuntary turnover is the electoral system: proportional systems tend to increase turnover, while majoritarian and double-list systems are condu- cive to higher re-election rates (Mutland and Stadlar 2004; Manow 2007; Heinsohn and Freitag 2012). In majoritarian systems, candidates rely on their status and resources; in contrast, in closed party lists in proportional environments, candidates depend more on parties’ decisions (Heinsohn and Freitag 2012; Francois and Grossman 2015). Parties take a step back in their gate-keeping activity in majoritarian systems because they ‘fear the loss of what is perhaps a personal vote’ (Matland and Studlar 2004: 101). Party selec- tion prior to the campaigning phase has proved to be a major factor in explaining the entry of newcomers into parliament (Gouglas and Maddens 2017). However, it remains a ne- glected aspect of turnover research. Frequently, turnover is affected by the inability of incumbents to obtain re-nomination (Moncrief 1998; Swain et al. 2000; Marino and Martocchia Diodati 2017). Re-nomination can fail due to resignations, party deselection (more frequent in PR systems), and death (Gouglas and Maddens 2017). In addition to the electoral system, Heinshon (2014) has listed four further variables that can affect turnover rates: 1. length of the legislative term: the more frequent the elections, the lower the turnover rate; 2. level of professionalization: MPs are less likely to abandon the legislature when benefits are greater and the institutional context is professionalized (Squire 1988; Hibbing 1999; Kerby and Blidook 2011; Gouglas et al. 2017); 3. degree of institutionalization: the stronger and more influential the legis- lature, the less likely it is that MPs will leave voluntarily (Sisson 1973; Hall and Van Houweling 1995; Hibbing 1999; Matland and Studlar 2004; Salvati 2016a); 4. legislature size: lower numbers of seats have a negative effect on turnover. Overall, the literature tells us that legislative turnover calls for a continuous re- search effort, due to the complexity produced by the interaction between variables at both the institutional and individual levels. Furthermore, it is extremely interesting to consider turnover in relation to the institutionalization of parties and to party system stability. From this viewpoint, high turnover can be interpreted as an indicator of struc- tural de-institutionalization (Huntington 1968). WHO’S THE WINNER? AN ANALYSIS OF THE 2018 ITALIAN GENERAL ELECTION 85 3. The Italian Case between Stability and Deep Transformations Italy has proved to be an interesting case study for the analysis of legislative turnover in a twofold sense. On the one hand, it is possible to observe significant variations over time in the structural conditions that are likely to affect turnover ratios. This allows us to make longitudinal comparisons within the same context of democratic continuity, thus keeping fixed the possible effect of intervening regime and cross-country idiosyncratic factors. On the other hand, the relevance of the Italian case ensues from the empirical mixture of stability and quick renewals which has characterized the parliamentary class over the years. As noticed by Verzichelli (2009: 60), this research topic has great poten- tial, especially when it comes to investigating political change in Italy from the 1990s. We operationalize legislative turnover simply as the percentage of new MPs out of the total membership, compared to the previous legislature (Matland and Studlar 2004; Gouglas et al. 2017). We do not distinguish between newcomers and those who re-enter after a ‘break’. Finally, we decided to focus only on the first parliamentary chamber. This choice is the most beneficial strategy from a comparative perspective, for reasons of lower comparability between second chambers due to their deep inter-chamber differ- ences (e.g., Vercesi 2017). In this regard, it is worth noting that, however, the low share of passing between the two parliamentary branches in Italian history (usually ranging between 4% and 8% of the total number of MPs) does not appear to be a potential signifi- cant source of bias (Verzichelli 2009: 61). Circumscribing the focus to the republican period only, we notice that – from the first general election in 1948 to 2013 – Italy has shown relatively high levels of turnover compared to other Western European countries. However, there has been a general in- crease especially from 1994 onwards (Gouglas et al. 2017). Before 2018, the turnover rate of the Chamber of Deputies surpassed the indicative threshold of 50% only in three cir- cumstances: in 1948, 1994, and 2013 (Verzichelli 2009: 53). Excluding the exceptional case of the 1948 election, held to substitute the then Constitutional Assembly with the new democratic Parliament, we easily find – mutatis mutandis – important similarities between the two remaining periods. The general election of 1994 was the first held with a new electoral system since 1953 and marked the final breakdown of the former party system (Bartolini and D’Alimonte 1996); this ground-breaking phase was subsequently followed by the new stabilization of the parliamentary class (Verzichelli 2006). In 2013, the party system underwent a new earthquake, due to the pressing entry of the Five Star Movement (M5S) in Parliament and the consequent shift from bipolarism to tripolar party competition (D’Alimonte 2013). Both electoral rounds were characterized by sig- nificant voter realignment and high electoral volatility. In Table 1, we present descriptive statistics on electoral turnout, volatility and legislative turnover since 1994, including 2018 for comparative purposes. Information about the mean age of the elected and the number of ‘new parties’ in Parliament is also included. To exclude minor changes, we consider as new parties only those that were not in the previous Parliament with a parliamentary group and reached at least 5% of votes. However, we do not count those lists resulting from the merger of former parties. Electoral volatility is calculated following in the footsteps of Pedersen (1979) and Chiaramonte and Emanuele (2015). With regard to the 2018 general election, an interesting puzzle stems from our pic- ture. The percentage of legislative turnover is as high as the one of the two critical SALVATI and VERCESI, Party Organizations and Legislative Turnover 86 passages which established the birth of the so-called ‘Second Republic’ in 1994 and its end in 2013 (Almagisti et al. 2014). The 2018 turnover equals the ratio of 66% of 2013 and it is only one percentage point below that of 1994. The mean age of MPs can be a further indicator of turnover. Indeed, the more people that get re-elected, the greater is the like- lihood of older Deputies. In this regard, the data are striking: in 2018 the age reaches its historical minimum, that is, 44 years.1 However, we observe a concomitant significant decrease in overall electoral volatility, compared to the previous election. In 1996, the downfall of volatility appeared together with the stabilization of a new parliamentary class compared to 1994: between 1994 and 1996, the legislative turnover moved from 67 to 44 percent. Moreover, the two peaks of legislative turnover in 1994 and 2013 are coin- cident with the entry of three and two, respectively, new major parties in the legislative arena. Nothing similar has happened in 2018: all the largest parties had already run the previous electoral campaign. This would lead us to argue that volatility has not involved new party actors. Table 1. Electoral turnout, volatility, and turnover in the Italian Chamber of Deputies, 1994-2018 Election Turnout (%) Total volatility (%) Legislative turnover (%) Mean age New parties 1994 86.3 39.3 66.8 47.1 3 1996 82.9 12.3 43.5 48.1 0 2001 81.4 20.4 41.9 50.4 0 2006 83.6 8.2 42.3 52.2 0 2008 80.5 11.3 40.8 50.8 0 2013 75.2 36.7 65.5 45.7 2 2018 72.9 26.7 65.9 44.3 0 Note: data for 2018 volatility are drawn from the ‘New Data Release (May 2, 2018)’ of Emanuele (2015). The parties People of Freedom (Popolo della Libertà – PDL) for 2013 and Go Italy! (Forza Italia – FI) for 2018 have been counted as the same party (e.g., Vercesi 2015). The same applies when a party has changed its label. Sources: data from the archives on the Italian parliamentary class of CIRCaP (University of Siena) presented in Verzichelli (2006, 2009) and Cotta and Verzichelli (2016: 78), Emanuele (2015), Openpolis (2018: 9), and own update based on data from the website of the Italian Ministry of the Interior (www.interno.gov.it). We have touched upon the fact that electoral volatility and party system changes are major factors when it comes to explaining legislative turnover (Gouglas et al. 2017). Overall, our data on the Italian Chamber of Deputies from 1994 and 2018 confirm this tendency. The positive Pearson correlational value between the two measures is indeed 0.91 (p< 0.01); if we exclude 2018, it even rises to 0.95 (p< 0.01). Overall, the 2018 general election appears to be a deviant case. If it cannot be explained by electoral volatility as straightforwardly as previous rounds, what accounts more for the prominent legislative turnover of 2018? Are there any factors that hold for 2018 and that distinguish this election from previous ones in terms of elected candidates? 1 It is worth remembering that the minimum age to be elected for the Italian first chamber is 25. WHO’S THE WINNER? AN ANALYSIS OF THE 2018 ITALIAN GENERAL ELECTION 87 4. Legislative Turnover in the 2018 Italian General Election: What Role for Parties? Recent changes in legislative turnover rates in Italy have been theoretically associated with a couple of conditions: the emergence of new parties, voters’ realignments, chang- ing electoral rules, the decline of mass-based parties, and the change of the multi-level state structure which provides career opportunities to ambitious politicians at other lev- els of government (Lanza and Piazza 2002; Verzichelli 2006; Cotta and Verzichelli 2016: 78, 177; Grimaldi and Vercesi 2018). We have seen that the first two conditions are not fully convincing when accounting for the high level of 2018. The fourth and fifth, in turn, can perhaps explain differences between the pre-1994 period and subsequent increases, but they cannot explain oscillations from one election to the next (e.g., Verzichelli 2010). The electoral system, too, does not account for all the variations between subsequent leg- islative terms. Moreover, it is worth pointing out that similar turnover rates in Italy in 1994, 2013, and 2018 were reached under three different electoral laws. In contrast, the significant decrease in 1996 occurred with the same mixed system of 1994 in force. In 2018, candidates were elected through a new system that allocates a minority part via single member districts and the majority of votes proportionally to closed party lists, on which the same candidate can be placed for more than one district. However, no split- vote was allowed. Except for the majority premium, the rule applied in 2013 resembled this logic to some extent (Chiaramonte 2015). If we discard the usual explanatory arguments, we need to look for alternatives. We have seen that political parties remain an intriguing option in this respect. Regression analyses have found no evidence of an impact of ideological positions (Gouglas et al. 2017: 18). According to Gouglas and Maddens (2017), selection is often more important than election as a drive behind new entries into Parliament. With regard to Italy, the im- pact of political parties seems visible in the higher percentages of non-rerunning outgoing MPs after 1994, associated with higher turnover rates (Verzichelli 2009: 54). Gouglas et al. (2017: 20-21) have mentioned the possible effect of intra-party rules. In particular, they have suggested analysing turnover at the party level and looking at fac- tors such as quotas and candidate selection methods. What they do not directly mention is the impact that strong leaderships can have. Scholarship on Italian political parties has underlined the flourishing of personal- ist/personal parties after the breakdown of the ‘First Republic’ (Pasquino 2014; Musella 2015). These parties have become increasingly dependent on their own leaders in terms of functioning and party strategies. To one extreme, leaders of these parties use party or- ganizations as vehicles for their own success, undermining the notion of party as a composite and autonomous organization (Blondel and Thiébault 2010; Vercesi 2015; Mu- sella 2018). It has been claimed that even the Democratic Party – traditionally based on cadres and a thicker organization – has recently moved in this direction (Salvati 2016b). Here, we posit that large renewals of parliamentary elites can be the result of a spe- cific way of organizing parties. In particular, we argue that leader-centred parties boost turnover, because in these parties ‘loyalty to the leader is the basis on which asymmetrical power relationships are established […]. A key indicator of this would be the authority to make unilateral decisions on nominations’ (Kostadinova and Levitt 2014). Thus, our SALVATI and VERCESI, Party Organizations and Legislative Turnover 88 expectation is that legislative turnover in 2018 in Italy has ensued from a substantial con- vergence of political parties towards a personalist model, larger than in the past. Before entering into an overall comparison, we first check if in 2018 turnover af- fected some parties more than others. Moreover, we gauge whether the renewal of MPs is proportional to the gains of votes for parties with better electoral performances. This al- lows us to observe if there was a cross-party trend and to pinpoint possible parties where selection was more important than election to reshape parliamentary groups (Table 2). Table 2. Variation in votes, turnover, and candidate renewal by party in 2018 (Chamber of Deputies) Party Votes 2013 (%) Votes 2018 (%) Variation in votes (%) 2013-2018 Turnover (%) New candidates (%) M5S 25.6 32.7 +7.1 72.3 75.0 (85.8) PD 25.4 18.7 -6.7 34.5 54.3 (57.5) League 4.1 17.4 +13.3 87.8 92.7 FI 21.6 14.0 -7.6 64.7 76.1 FdI 2.0 4.4 +2.4 85.2 94.7 Centre- right 27.6 37.0 +9.4 79.2 75.4 (only SMD) LeU - 3.4 +3.4 21.4 84.8 (91.9) Note: calculations based on proportional votes to party lists, national territory (Aosta Valley excluded). FdI: Fratellid’Italia, Brothers of Italy; FI: Forza Italia, Go Italy!;LeU: Liberi e Uguali, Free and Equal; M5S:Movimento 5 stelle, Five Star Movement; PD: Partito democratico, Democratic Party. Only parties with parliamentary representation in 2018 are included. For simplicity’s sake, the counting of new candidates is based on a stricter criterion, compared to that used to calculate turnover. We indicate the percent- age of candidates without previous parliamentary experience in plurinominal lists. Values between brackets refer to single member districts. In the latter case, the PD ran in coalition with other minor parties. Sources: Openpolis (2018), Istituto Cattaneo (2018a), Italian Ministry of the Interior’s website (www.interno.gov.it), and own up- dates. Data show that parties have promoted high levels of turnover, irrespective of their electoral performances. If we look at the ‘new candidates’ placed on party lists, we see that a significant majority had not had previous parliamentary experience. However, the net difference between new candidates and turnover confirms what previous research has shown: often, new candidacies occupied lower positions on the party lists, thus de- creasing the likelihood of election (Istituto Cattaneo 2018a). In this respect, the cases of the PD and LeU are clear-cut examples. However, it is interesting to note that the PD was affected by a renewal of 35% of its MPs, although it lost 7% of its votes. In this regard, a caveat must be addressed. As we have said, the electoral law used in 2013 provided for a majority premium, which granted about 55% of seats in the first chamber to the coalition or list with more votes (Chiaramonte 2015). In particular, this overrepresented the Dem- ocratic Party (PD) and the electoral coalition it had led which won the election with less than 1% of votes, compared to the centre-right coalition; the M5S, in turn, ranked third, about 4 percentage points below the winning coalition. The move towards lower dispro- portionality could be seen as a possible drive behind the high turnover of some parties other than the PD. For example, the M5S has more than doubled its number of seats in 2018 subsequent to an increase of only 7% of votes, thus being confronted with the need to allocate a proportionally higher percentage of new seats. On the other hand, the loss of WHO’S THE WINNER? AN ANALYSIS OF THE 2018 ITALIAN GENERAL ELECTION 89 the majority bonus together with the decrease in votes should have reduced the PD’s room for manoeuvre to change representatives beyond its parliamentary elite. However, we observe indications of high turnover in this party as well. All in all, thus, our numbers suggest that the trend towards the renewal of the parliamentary class involved all parties, albeit with different gradients. To answer our questions, we compare a few central characteristics of parties in 2013 and 2018, so that we can see any changes within the same organizations. For control pur- poses, we double-checked our findings with similar information about parties in 1994 and 1996. According to classic explanations, after a round of adjustment, new stabiliza- tion should have come. This was true in 1996, but not in 2018. By conducting the comparison by pairs of elections, we should be able to see if there are clues that party fac- tors actually could play a role. If the degree of variation between 1994-1996 and 2013- 2018 is similar, then we should reject our conjecture. A first rough indicator of party change is the change of leadership from one election to the next. Parties can become more personalized by modifying the actual functioning, but substantial moves are more likely if there is a new leader. In addition, we present a few proxies of party personalism and leader’s strength over nomination. They refer to leadership powers and intra-party democracy (Poguntke et al. 2016). Building on Sandri et al. (2013), we look at the selectorate’s inclusiveness of leadership contests: we argue that inclusive methods can lead to more personalized leaderships (plebiscitary effect); however, if there is no election and the leader holds the position ‘by right’, this probably further increases his or her strength within the party. Another important aspect is the possibility and the likelihood of being able to remove the leader from his or her office: stronger personal leaderships are hard to remove. Moreover, we need to focus on the leader’s actual control over nominations. We also refer to Hazan and Rahat (2010: 35) to distinguish between inclusive and exclusive candidate selection methods: once more, moving from party circles to broader selectorates would foster personalization, but turn- over should be even more likely when the single leader decides on his or her own. Finally, we claim that more inclusive candidacy (Hazan and Rahat 2010: 20) favours turnover. We provide the overview of party changes in Table 3. For clarity’s sake, we limit our focus only to the four largest parties in parliament in 1996 and 2018, so that we can ex- clude parties with residual impact on turnover. Changes in party organizations are marked in italics in the table. Findings support our expectations. While from 1994 to 1996 no substantial changes occurred (except for the change of party leader in the PDS), between 2013 and 2018 three out of the four main Italian parties moved in the expected direction. Methods of leader selection moved to- wards higher inclusivity. On the other hand, leaders increased their control over the party. Moreover, unlike in 2013, in 2018 all leaders had a final say on nominations, often raising complaints among excluded party cadres, former MPs nominated in the lower positions of party lists, and underrepresented party minorities. FI did not experience substantial changes in its organizational arrangements, but this indicates nothing but the confirmation of Berlusconi’s role of absolute party domi- nus. FI’s lists were formed with the aim of opening lists to civil society and specifically to a new wave of MPs recruited from Berlusconi’s enterprises (as previously happened in 1994): this renewal was pursued together with confirmation of the most loyal party SALVATI and VERCESI, Party Organizations and Legislative Turnover 90 members.2 Even in a personal party such as FI, this method has been conducive to ten- sions with excluded candidates and penalized incumbents.3 The impact on turnover has been amplified by the poor electoral performance of FI, which suffered from an intra- coalitional competition with Salvini’s League, which was able to attract many former FI voters (Istituto Cattaneo 2018b). Table 3. Party Features, 1994-1996 and 2013-2018 1994 1996 PDS FI AN LN PDS FI AN LN Leadership change No - No No Yes No No No Leader’s selectorate Party convent. Party council Party convent. Party convent. Party convent. Party council Party convent. Party convent. Leader’s deselection No No No No No No No No Control over nominations Very low Very high Low High Very low Very high Low High Candidate selection Party council Single leader Party council Party council Party council Single leader Party council Party council Candidacy Citizens Citizens Citizens Citizens Citizens Citizens Citizens Citizens 2013 2018 M5S PD LN PDL M5S PD League FI Leadership change - Yes Yes No Partial Yes Yes No Leader’s selec- torate Single leader Voters Party convent. Party convent. Members / single leader Voters Members Party convent. Leader’s deselection No Party congress No No No Party congress No No Control over nominations Very high Moderate High Very high Very high Very high Very high Very high Candidate se- lection Members Members/ single leader Party elite Single leader Members / single leader Single leader Single leader Single leader Candidacy Members + requirements Citizens Citizens Citizens Citizens Citizens Citizens Citizens Note: see Table 1. PDS: Partito democratico della Sinistra, Democratic Party of the Left; AN: Alleanza nazionale, National Alliance; LN: Lega Nord, Northern League. Party convent. stands for party convention. Sources: Sandri et al. (2013); Pasquino (2014: 561), Vercesi (2015), and own elaboration and updates based on party statutes. The nomination process proved to be less complicated for the League, where Salvini and his deputy secretary Giorgetti had total control of lists. Such centralization was fa- cilitated by the transformation of the League from regional to national party.4 2 See: https://www.ilfattoquotidiano.it/2018/01/30/elezioni-berlusconi-schiera-indagati-e-fedelissimi- le-liste-cambiate-nella-notte-de-girolamo-litiga-con-carfagna/4126607/ (first access: 10 May 2018). 3 See: http://www.napolitoday.it/politica/forza-italia-liste-candidati-degirolamo-laboccetta.html (first access: 10 May 2018). 4 See: https://www.ilfoglio.it/politica/2018/01/24/news/lega-matteo-salvini-candidati-collegi-uninom- inali-sud-174968/ (first access: 11 May 2018); http://www.ilgiornale.it/news/politica/bongiorno- bagnai-tutti-i-candidati-lega-1488411.html (first access: 10 May 2018). WHO’S THE WINNER? AN ANALYSIS OF THE 2018 ITALIAN GENERAL ELECTION 91 PD’s path was more tortuous: for the first time, the definition of party lists was cen- tralized in the hands of the party leadership (i.e., Renzi and his inner circle). The internal opposition was almost deleted from party lists (or, however, from the most se- cure positions), in order to guarantee the formation of parliamentary groups loyal to the party leader.5 Finally, the only party that in the 2013 election requested some specific require- ments for candidacy – the M5S – not only followed in other parties’ footsteps by opening candidacy to all citizens (thus enhancing the possibility of the renewal of the parliamen- tary class), but it also gave to the central party leadership (i.e., Di Maio and Grillo) the opportunity to personally select the candidates collocated in the top positions of the closed party lists.6 5. Discussion and Conclusion We have seen that the 2018 Italian election has not caused an earthquake in the format of the party system. Thus, party system deinstitutionalization cannot be a viable expla- nation for turnover rates. Beyond electoral volatility and party performances, data show that it is fruitful to investigate intra-party factors and selection procedures to improve our knowledge about turnover determinants. Albeit with different nuances, the empirical evidence points in the direction of the strengthening of leaders’ control over nominations. Overall, this seems to highlight a certain systemic transformation of selection criteria within parties for the 2018 election. How parties work and organize, selection criteria, and leadership styles seem to in- fluence the degree of renewal in the parties’ rank-and-file. Considering that all the four main parties have shown significant centralization (with the partial exception of the M5S with regard to selection), it seems that the combination of highly personalized lead- erships and a strong control over nominations is a valuable clue (among others) to solve the turnover enigma. With the exception of FI, all the main parties moved, where possible, towards more inclusive leadership selection methods between 2013 and 2018: this kind of legitimation can account for the enforcement of more leader-centred forms of candidate selection. It is fundamental to remember that three out of four of the main parties changed leader- ship after 2013 (the M5S defined its own leadership); the high turnover could be explained by the leaders’ will to form reliable parliamentary groups in the Chamber of Deputies. To conclude, we can argue that a large part of the turnover is likely to be determined by what occurs before the elections and that the new frontier of turnover studies should probably be found in the analysis of intra-party organizational factors (leadership, or- ganization size, party territorial dimension, party role in institutions, etc.), party 5 See: https://www.panorama.it/news/politica/le-liste-del-pd-annunciano-una-nuova-scissione/ (first access: 9 May 2018); and also: http://www.repubblica.it/politica/2018/01/27/news/pd_trattativa_a_ol- tranza_nella_notte_per_le_liste_renzi_non_ci_sara_condivisione_totale_-187372869/ (first access: 9 May 2018). 6 See: https://www.panorama.it/news/politica/nuove-regole-per-il-movimento-5-stelle-addio-agli-in- cendiari/ (first access: 10 May 2018). SALVATI and VERCESI, Party Organizations and Legislative Turnover 92 electoral performances, and nomination of party candidates (Gouglas and Maddens 2017; Gouglas et al. 2017). It is worth bearing in mind that these three aspects are strictly intertwined. This means that there is a reciprocal influence behind the shaping of the empirical effects. References Almagisti, M., L. Lanzalaco and L. Verzichelli (eds.) (2014). La transizione politica italiana. Da Tangentopoli a oggi. Roma: Carocci. Bartolini, S. and R. D’Alimonte (1996). ‘Plurality Competition and Party Realignment in Italy: The 1994 Parliamentary Elections. ’European Journal of Political Research 29(1): 105- 142. Blondel, J. and J.-L. Thiébault (2010). Political Leadership, Parties and Citizens: The Person- alisation of Leadership. New York: Routledge. Chiaramonte, A. (2015). ‘The Unfinished Story of Electoral Reforms in Italy.’ Contemporary Italian Politics 7(1): 10-26. Chiaramonte, A. and V. Emanuele (2015). ‘Party System Volatility, Regeneration and De-In- stitutionalization in Western Europe (1945–2015).’ Party Politics 23(4): 376-388. Cotta, M. and L. Verzichelli (2016). Il sistema politico italiano. Bologna: IlMulino. D’Alimonte, R. (2013). ‘The Italian Elections of February 2013: The End of the Second Repub- lic?’ Contemporary Italian Politics 5(2): 113-129. Emanuele, V. (2015). ‘Dataset of Electoral Volatility and its internal components in Western Europe (1945-2015).’ Rome: Italian Center for Electoral Studies, http://dx.doi.org/10.7802/1112. Francois, A. and Grossman, E. (2015). ‘How to Define Legislative Turnover? The Incidence of Measures of Renewal and Levels of Analysis.’ The Journal of Legislative Studies 21(4): 457-475. Gouglas, A. and B. Maddens (2017). ‘Legislative Turnover and Its Sources: It’s the Selection.’ Politics Online First, doi: 10.1177/0263395717701161. Gouglas, A., B. Maddens and M. Brans (2017). ‘Determinants of Legislative Turnover in West- ern Europe, 1945-2015.’ European Journal of Political Research Early View, doi: 10.1111/1475-6765.12246. Grimaldi, S. and M. Vercesi (2018). ‘Political Careers in Multi-Level Systems: Regional Chief Executives in Italy, 1970–2015.’ Regional & Federal Studies 28(2): 125-149. Hall, R. and P. Van Houweling (1994). ‘Avarice and Ambition in Congress: Representatives’ Decisions to Run or Retire from the U.S. House.’ American Political Science Review 89 (1): 121-136. Hazan, R.Y. and G. Rahat (2010). Democracy within Parties. Candidate Selection and Methods and Their Political Consequences. Oxford: Oxford University Press. Heinsohn, T. (2014). ‘Institutional determinants of legislative turnover in the German state parliaments.’ The Journal of Legislative Studies 20(4): 473-494. Heinsohn, T. and M. Freitag (2012). ‘Institutional Foundations of Legislative Turnover: A Comparative Analysis of the Swiss Cantons.’ Swiss Political Science Review 18(3): 352- 370. Hibbing, J.R. (1999). ‘Legislative careers: why and how we should study them?’ Legislative Studies Quarterly 24(2): 149-171. WHO’S THE WINNER? AN ANALYSIS OF THE 2018 ITALIAN GENERAL ELECTION 93 Huntington, S. P. (1968). Political Order in Changing Societies. New Haven and London: Yale University Press. Istituto Cattaneo (2018a). Come sarà la prossima classe parlamentare? Bologna: Fondazione di ricerca Istituto Carlo Cattaneo. Istituto Cattaneo (2018b). Elezioni politiche 2018 – I flussi di voto – 2^ parte: altre 4 città sotto analisi. Le tendenze principali. Bologna: Fondazione di ricerca Istituto Carlo Cattaneo. Kerby, M. and K. Blidook (2011). ‘It’s Not You, It’s Me: Determinants of Voluntary Legislative Turnover in Canada.’ Legislative Studies Quarterly 36(4): 621-643. Kostadinova, T. and B. Levitt (2014). ‘Toward a Theory of Personalist Parties: Concept For- mation and Theory Building.’ Politics & Policy 42(4): 490-512. Kuklys, M. (2013). ‘Legislative Turnover in the Baltics after 1990: Why Is It So High and What Are Its Implications?’ Baltic Journal of Political Science 2(2): 30-49. Lanza, O. and G. Piazza (2002). ‘Il ricambio dei parlamentari.’ In G. Pasquino (ed.), Dall'Ulivo al Governo Berlusconi: le elezioni del 13 maggio 2001 e il sistema politico italiano. Bologna: Il Mulino, pp. 239-273. Manow, P. (2007). ‘Electoral Rules and Legislative Turnover: Evidence from Germany’s Mixed Electoral System.’ West European Politics 30(1): 195-207. Marino, B. and N. Martocchia Diodati (2017). ‘Masters of Their Fate? Explaining MPs’ Re- Candidacy in the Long-Run: The Case of Italy (1987-2013).’ Electoral Studies 48: 1-9. Matland, R.E. and D.T. Studlar (2004). ‘Determinants of Legislative Turnover: A Cross-Na- tional Analysis.’ British Journal of Political Science 34(1): 87-108. Matthews, D.R. (1984). ‘Legislative Recruitment and Legislative Careers.’ Legislative Studies Quarterly 9(4): 547-585. Moncrief, G.F. (1998). ‘Terminating the Provincial Career: Retirement and Electoral Defeat in Canadian Provincial Legislatures, 1960-1997.’ Canadian Journal of Political Science 31(2): 359-372. Moncrief, G.F., R. Niemiand L. Powell (2004). ‘Time, Term Limits and Turnover. Trends in Membership Stability in U.S. State Legislatures.’ Legislative Studies Quarterly 29(3): 357-381. Musella, F. (2015). ‘Personal Leaders and Party Change: Italy in Comparative Perspective.’ Italian Political Science Review 45(3): 227-247. Musella, F. (2018). Political Leaders Beyond Party Politics. Basingstoke: Palgrave Macmillan. Openpolis (2018). Tre poli contrapposti 2018. Un’analisi della composizione di camera e senato all’avvio della XVIII legislatura. Roma: Associazione Openpolis. Pasquino, G. (2014). ‘Italy: The Triumph of Personalist Parties.’ Politics & Policy 42(4): 548- 566. Pedersen, M.N. (1979). ‘The Dynamics of European Party Systems: Changing Patterns of Elec- toral Volatility.’ European Journal of Political Research 7(1): 1-26. Poguntke, T., S.E. Scarrow and P.D. Webb (2016). ‘Party Rules, Party Resources and the Poli- tics of Parliamentary Democracies: How Parties Organize in the 21st Century.’ Party Politics 22(6): 661-678. Polsby, N. (1968). ‘The Institutionalization of the U.S. House of Representatives.’ American Political Science Review 62(1): 144-168. Rosenthal, A. (1974). ‘Turnover in State Legislatures.’ American Journal of Political Science 18(3): 609-616. SALVATI and VERCESI, Party Organizations and Legislative Turnover 94 Russo, F., F. Tronconi and L. Verzichelli (2014). ‘Snipers and Switchers. The Difficulties of Parliamentary Representation in the Italian XVII Legislature.’ Polis 28(2): 85-106. Salvati, E. (2016a). ‘Towards a European Parliamentary Class? A Proposal for a Typology of the MEPs.’ Journal of Comparative Politics 9(1): 59-74. Salvati, E. (2016b). ‘Matteo Renzi: A New Leadership Style for the Italian Democratic Party and Italian politics.’ Modern Italy 21(1): 7-18. Samuels, D. (2000). ‘Ambition and Competition: Explaining Legislative Turnover in Brazil.’ Legislative Studies Quarterly 25(3): 481-497. Sandri, G., A. Seddone and F. Venturino (2013). ‘The Selection of Party Leaders in Italy, 1989- 2012.’ In J.B. Pilet and W.P. Cross (eds.), The Selection of Political Party Leaders in Con- temporary Parliamentary Democracies. A Comparative Study. London: Routledge, pp. 93-107. Sisson R. (1973). ‘Comparative Institutionalization: A Theoretical Explanation.’ In A. Korn- berg (ed.), Legislatures in Comparative Perspective. New York: David McKay Company, pp. 17-38. Somit, A. (1994). The Victorious Incumbent: A Threat to Democracy? Dartmouth: Dartmouth Pub Co. Squire, P. (1988). ‘Legislative Professionalization and Membership Diversity in State Legisla- tures.’ Legislative Studies Quarterly 17(1): 69-79. Squire, P. (1992). ‘Career Opportunities and Membership Stability in Legislatures.’ Legisla- tive Studies Quarterly 13(1): 65-82. Squire, P. (1992). ‘Membership Turnover and the Efficient Processing of Legislation.’ Legisla- tive Studies Quarterly 23(1): 23-32. Swain, J.W., S.A. Borrelli, B.C. Reed and S.F. Evans (2000). ‘A New Look at Turnover in the U.S. House of Representatives, 1789-1998.’ American Politics Quarterly 28(1): 435-457. Vercesi, M. (2015). ‘Owner Parties and Party Institutionalisation in Italy: Is the Northern League Exceptional?’ Modern Italy 20(4): 395-410. Vercesi, M. (2017). ‘What Kind of Veto Player Is the Italian Senate? A Comparative Analysis of European Second Chambers.’ Journal of Modern Italian Studies 22(5): 604-623. Verzichelli, L. (2006). ‘La classe politica italiana, dalla crisi all’adattamento: accesso, circola- zione e carriere dal 1994 al 2006.’Rivista italiana di scienza politica 36(3): 455-478. Verzichelli, L. (2009). ‘Mutamento e continuità nell’élite parlamentare repubblicana.’Ventu- nesimo Secolo 8(19): 51-72. Verzichelli, L. (2010). Vivere di politica. Come (non) cambiano le carriere politiche in Italia. Bologna: Il Mulino.