ANA MARIA EVANS MAARTEN PETER VINK Measuring group switching in the European Parliament: Methodology, data and trends (1979-2009) Anälise Social, 202, xlvii (i.°), 2012 issn online 2182-2999 EDigÄO E PROPRIEDADE Instituto de Ciéncias Sociais da Universidade de Lisboa. Av. Professor Anibal de Bettencourt, 9 1600-189 Lisboa Portugal — analise.social@ics.ul.pt Andlise Social, 202, xlvii (i.°), 2012, 92-112 Measuring group switching in the European Parliament: Methodology, data and trends (1979-2009). Party group switching in the European Parliament (ep), where parliamentarians individually or collectively switch from one party group to the other, is a well-known contributor to the volatility of the EP party system. We present a new dataset that contains party group information on all MEPs from 1979 to 2009. As a first step to a more comprehensive analysis of the phenomenon of party group switching in the EP we describe characteristics of all switches that have occurred in these six legislatures, with a focus on the trends across time, variety between member states, party groups, and ideological party families. Keywords: comparative politics; European Parliament; political parties; party switching. Migracao partidaria no Parlamento Europeu: metodologia, dados e tendencias (1979-2009). A migracao partidaria no Parlamento Europeu (pe), fenomeno que define a mudanca de um deputado ou de um conjunto de deputados para um grupo politico diferente daquele em que inicia o mandato, contribui para a volatilidade do sistema partidario da unica institui-cao europeia diretamente eleita pelos cidadaos. Neste estudo apresentamos os resultados de uma base de dados original e exaustiva sobre a evolucao da filiacao parlamentar de todos os deputados eleitos ao pe desde as primeiras eleicoes diretas em 1979 ate ao final da sexta legislatura, em 2009. O estudo examina as caracteristicas das migracoes que ocorreram neste periodo, analisando em particular a direcao dos fluxos, a sua variacao no tempo, por nacionalidade, grupos parlamentares e familias ideologico-partidarias. Palavras-chave: politica comparada; Parlamento Europeu; par-tidos politicos; migracao partidaria. Ana Maria Evans » Institute of Social Sciences, University of Lisbon; ana.maria.evans@ics.ul.pt Maarten Peter Vink » Department of Political Science, Maastricht University. Email: m.vink@maastrichtuniversity.nl ANA MARIA EVANS MAARTEN PETER VINK Measuring group switching in the European Parliament: Methodology, data and trends (1979-2009)1 introduction This paper examines party group switching in the European Parliament (ep). A core set of political groups have operated in the ep since the first direct elections in 1979, namely the Christian Democrats, the Socialists, the Liberals, and the Communists. But in the course of ep legislatures, there have been frequent movements of individuals and national party delegations in and out of these groups. Moreover, the composition of these groups has varied significantly between legislative mandates. Furthermore, a significant number of new parliamentary groups have emerged, merged, and disappeared during and between ep legislatures. The fluidity of the ep party system contrasts with the stability of parliamentary parties during the course of legislative mandates in most West European democracies. The existing literature on party groups and political coalitions in the ep is dominated by analyses on voting cohesion, drawing on roll call votes to examine the compliance of national party delegations with the voting line of 1 Earlier versions of this paper were presented at the joint meeting of the Maastricht University research group on Politics and Culture of Europe and the ics-ul research program Citizenship and Democratic Institutions (Lisbon, June 2010), the meeting of the ecpr Standing Group on European Union (Porto, June 2010), and the biennial conference of the European Union Studies Association (Boston, March 2011). We thank Thomas Christiansen, Peter A. Hall, Bjorn Hoyland, and Steven Van Hecke for their comments. We are grateful for research assistance by Vera Henriques and Ana Guardiao. Part of the research has been funded by the fct project "Switching Behavior in the European Parliament: Voting Power and Ideology", ref. pocti/cpo/6ioi2/2004. ana maria evans and maarten peter vink Political Groups in the ep (see Attina, 1990; Bardi, 1994; Faas, 2002, 2003; Hix, 2002, 2005; Hix, Noury and Roland, 2006, 2007; Kreppel, 2002, 2004; Marsh and Norris, 1997). There have been few analyses of political affiliation in the ep specifically geared toward an understanding of the phenomenon of party group switching (for exceptions, on specific cases and periods, see Verzichelli, 1999; Maurer, Parkes, and Wagner, 2008; Evans, 2009; McElroy, 2008; McElroy and Benoit, 2007, 2009). This paper seeks to contribute to a better understanding of this phenomenon. We trace all instances when members of the ep (meps) have changed between ep groups (epgs) in the period between the first directly elected legislature in 1979 and the end of the sixth legislature in 2009. In order to do so, and drawing on earlier databases containing comprehensive information on the parliamentary activities of meps (Hix et al, 2006; Hoyland et al., 2009), as well as on data from the ep archives, we have built a new comprehensive dataset with extensive information on the ep curriculum of the 4163 person legislature combinations between 1979 and 2009. The results of a first descriptive analysis of these data highlight the significant extent of group switching in the ep, with on average 9% of all meps, and up to 18% of meps during the 1989-1994 legislature switching party group, with some of them switching up to four times per legislature. These findings constitute a building block toward a more systematic understanding of party system fluidity in the European Parliament and of the more general phenomenon of legislative switching. Before proceeding to the presentation of our dataset and core findings, we give a brief account of the role and functions of European Party Groups (epgs) and their importance in ep parliamentary life. In particular, we address the specificities that characterize these party groups, when compared to the typical parliamentary groups of West European parliamentary settings. The goal is to provide a summarized account of the institutional context in which choices of ep political affiliation are made and how such decisions relate to the primary goals of elected politicians in the ep. We relate this discussion to available literature on the topic of legislative switching. Next we address the methodology adopted for the compilation of the data, in particular coding procedures and the cross-checking techniques we have adopted in order to guarantee the reliability of the information. We also provide an account of the core features of the dataset. In the section on the evolution of switching in the European Parliament between 1979 and 2009 we focus on the variation across legislatures, countries, and political families, the timing of switches, directions of the flows, the percentage of collective vs. individual switching, and the relationship between switching at the national and supranational levels. We conclude with a brief discussion of these findings. measuring group switching in the european parliament (1979-2009) the character and evolution of parliamentary groups in the ep European Political Groups (epgs) differ from national parliamentary parties, and some fluidity of membership should be expected as a result: epgs are not the "parliamentary arms" of enduring and effectively competing party organizations. Although one can argue that the oldest groups in the ep have built up some organizational strength and mechanisms for continuity through their links with European party federations (see Maurer, Parkes and Wagner, 2008), in general epgs can be described as post-electoral coalitions between national party delegations and individuals who broadly identify with a set of programmatic principles. Each parliamentary group in the ep is made of a set of party delegations and meps who come from different cultural settings, with entrenched national perceptions and specific historical legacies, and which represent more diverse social, economic, and political interests and organizational traditions than any national parliamentary party, epgs do not perform the typical gate keeping function of national political parties because it is the latter that have a determinant voice in the selection of candidates for the ep elections. Furthermore, there is no direct link between European political groups and voters. Although the largest and oldest epgs issue policy platforms and election manifestos before ep elections, the campaigns for these elections are run by national parties and often revolve around national issues (Reif and Schmitt, 1980; Kreppel, 2002, 2004; McElroy and Benoit, 2009). This leaves a wide scope of choice for ep grouping by elected representatives. National parties that compete for votes at the domestic level may choose to belong to one same parliamentary group in the ep. A case in pointare the Portuguese Left Bloc and the Portuguese Communist Party, two parties that contest on the left in Portugal but sit together in the Confederal Group of the European United Left/Nordic Green Left (gue/ngl) in the European Parliament. Nevertheless, the choice of political groups in the ep is critical, epgs play a key role in important aspects of organizational and political life in the ep. They play a key role in agenda setting and question time (see Rule 115) and in appointing meps to leadership positions in the ep and to important functions in committees and delegations (Rules 13, 24 and 186). Each epg is entitled to a share of the ep budget, which is calculated according to a complex formula that takes into account the number of members and the number of languages represented in the group (Rule 31). The Groups are also entitled to support from the ep in terms of staff, office space, and technical facilities (see Corbett et al., 2005, p. 95 and ff). All these perks are important for national party ana maria evans and maarten peter vink delegations, including the financial appropriations from which they share to help finance staff and to conduct "information" activities (ibid, p. 102). Furthermore, group membership is of great significance for the capability to bargain the outcomes of voting in the ep (see Settembri, 2004). This means that, as the ep has acquired more powers through treaty revisions, the role of its political groups in influencing policy has also increased. Hence, the first and most important choice national party leaders and meps make when elected to the ep is whether to form a new group, or join an existing group, or to become non-attached, or to maintain membership in their previous group - in the case of those who held seats in the previous ep legislature (see Maurer, Parkes, and Wagner, 2008, p. 250). The two largest groups - the Socialists and the Christian Democrats - have dominated party politics in the ep, and have held a large percentage of leadership functions in the ep. During the first five ep legislatures, between 1979 and 1999, the weight of these two groups increased steadily, from 53.6% of seats to 66% of seats. The combined weight of the two groups declined somewhat to 62.6% of the seats at the beginning of the sixth legislature, in July 2004. At different stages of the institutional development of the ep, there have been important changes of its party group composition, and a tendency for fragmentation, with the emergence of new political groups on the left and right - but more accentuated on the right - and with the formation of groups voicing regional interests, green movements, and eurosceptic parties as well. Until 1972, only three political families, namely the Socialists, the Liberals, and the Christian Democrats, were represented in the ep. With the 1979 direct elections, the ideological spectrum of the ep widened out of the center, toward both left and right, with the Communist and Allies group monopolizing the left and the European Democratic Group and the European Progressive Democrats, splitting the right-hand side of the ep ideological spectrum. The most significant additions to existing ep groups at the start of the second legislature in i984were the formation of the Rainbow group, representing regionalist and green parties, and of the new European Right group. The beginning of the third legislature, in 1989, was marked by further fragmentation, following the split of the Communist Group into the Group for the European United Left and the Left Unity group, and the formation of the Green group, which stole seats from the Rainbow group and eventually led to the demise of the latter after the subsequent ep elections, in i994.The fourth legislature was marked by further fragmentation as a consequence of the formation of a new regionalist party - the Group of the European Radical Alliance - and of the formation of the Europe of the Nations group, formally representing Euro-sceptic/anti-eu parties in the ep, for the first time. Although the Group of the measuring group switching in the european parliament (1979-2009) European Right disappeared, a new group formed on the right, with the entry of the Forza Italia party delegation, which formed the Forza Europa group on its own (in 1995 the party transitioned to the Union for Europe group). In 1999, the number of groups in the ep decreased again, with the merger of the national party delegations representing green and regionalist parties into one group. The opening of the sixth legislature in 2004 saw no significant change to the ideological distribution of epgs. the institutional foundations of party group switching in the ep Several features of parliamentary life in the ep make it more complex and more subject to fragmentation than is the case in the standard institutional settings of West European parliaments. The entry in the ep of new national party delegations, bringing in new political views and new national interests when new Member States join the eu, induces political and institutional complexity and even invites unstable political behavior. Changes of procedural rules associated with alterations in the decision-making powers of the ep also contribute to complexity. Moreover, switching is facilitated by the lack of penalties in the Rules of Procedure (see Rule 17 and 186) and by the loose nature of parliamentary groups in this institution. The indifference and low levels of knowledge among European voters with regard to ep activities and surveys on ep elections (see European Election Studies) reflected in the high levels of abstention for these electoral contests, also encourage a flexible approach to membership and open the door to the possibility of national party delegations and meps choosing to affiliate with different epgs. While these political and institutional elements facilitate switching behavior, they are insufficient to explain how this phenomenon is triggered, the evolutionary patterns that characterize it, and how it relates to party system development in the eu. Research on legislative switching in national settings relates the institutional context in which choices of political affiliation are made with the general assumption established by earlier studies on party politics that political competition is driven by the basic goals of maximizing votes, office, and policy (Strom and Muller, 1999). The basic premise that inspires this approach is that institutions condition the resources and opportunities through which politicians fulfill their ambitions (Hall and Taylor, 1996). Political institutions that grant politicians a great deal of autonomy vis-á-vis parties, namely electoral rules that weaken party leadership control over candidate selection, weak party structures, weak ties between electorates, and party labels (i.e. weak party identification), allow politicians to manipulate political id for short-term ana maria evans and maarten peter vink electoral goals and for the distribution of pork-barrel benefits among local clienteles (see Mainwaring, 1991 and Mainwaring and Scully, 1995). In these institutional settings, one should expect legislators to calculate their expected eligibility and distributional revenues in any party that may be generally compatible with their policy goals and to decide on switching party id if the electoral benefits are greater than the transaction cost of moving (see Desposato, 2006). The logical deduction is that one should expect legislators to remain in, or have incentives to move to, governing parties or large parties - because the latter give privileged access to state resources, and thus more pork, than opposition or smaller parties (see Heller and Mershon2005; 2009). In turn, parties should be willing to accept defectors when their current members reason that the arrival of new members increases their own expected payoff (see Laver and Benoit, 2003). These studies shed light on the factors one would expect to influence the propensity to change political affiliation in the ep. As we have seen above, epgs do not control candidate selection, they lack party structures comparable to those of national parties, and voters can hardly identify an epg label. Electoral rules matter very little in this context. All Member States - except for the uk until 1999- have adopted proportional methods for the ep elections. Although there are differences in proportionality related to the number and size of constituencies and to the thresholds for representation, as well as to the possibility that voters choose and/or rank their favorite candidates, such variation is insufficient to establish direct ties between the legislative behavior in the ep and re-election or re-nomination for party lists (see Corbett et al, 2005, p. 14). In this loose ideological context, Evans (2009) and Maurer et al. (2008) claim that party leaders that do not represent extremist views and/or single issues have strong incentives to opt for one of the largest epgs. Based on examination of the ep affiliation of Portuguese parties since 1986, Evans argues that parties that rotate in national government tend to prefer (and try to move to) the epp and the pes, in order to enjoy the inter-institutional policy networks that are sponsored by these groups and that provide bargaining and informational capability that are unmatched by smaller groups, including the Liberals. After long stints in the opposition, party leaders may switch their discourse and ep membership toward Euro-scepticism, with the goal of enhancing their electoral prospects, as the cases of the British Conservatives and the Portuguese Popular Party illustrate. But as soon as these parties are elected back to government they start again to approach the largest ep groups (see Evans, 2009; Maurer et al., 2008). Research on legislative party switching in national settings provides insights on other factors that have not been explored in the few studies conducted so measuring group switching in the european parliament (1979-2009) far on ep switching. According to Shabad and Slomczynski (2004), one should expect to find a correlation between the direction of flows and the degree of party system development. Accordingly, party mergers and intense movements to older (vs. new) parties are an indicator of party system consolidation. Flows within (rather than across) political families indicate the consolidation of ideological blocs, which is accordingly, a general trait of developed party systems. We thus expect the direction of flows to point toward the largest and oldest groups, namely the epp and the pes, because these are the epgs most endowed in terms of office and financial perks, and the more seats these groups obtain, the more resources they are entitled to. Earlier studies have also found that different stages of the electoral cycle provide different kinds of incentives for switching, including the expectation to obtain parliamentary offices and privileges in the early period of legislatures, the goal to improve electoral positioning in the late stages of a legislative mandate, and the aim to achieve policy goals at the time of major policy decisions (see Mershon and Shvetsova, 2005). Drawing on these findings, we expect switching in the ep to be concentrated in three periods of the legislative cycle: at the very beginning of each legislature, when national party delegations negotiate the prerogatives of parliamentary group membership; before the legislative mid-term, when important leadership positions in the ep rotate (see Rule 17); and toward the end of the mandate, when meps position themselves for re-nomination to national party lists. ep group switching dataset Although much information is available on parliamentary affiliation and parliamentary behavior of meps (see in particular Hix et al, 2006 and Hoyland et al., 2009), none of the existing databases is specifically geared toward the analysis of group switching in the ep. As a result, and somewhat surprisingly, considering that the phenomenon of group switching in the ep is well-known (see e.g., Brzinski, 1995 for an early discussion; and Corbett et al, 2005), little comprehensive and systematic analysis of this phenomenon is currently available. In this section we discuss the methodology we have followed for the construction of our new, comprehensive dataset on group switching in the European Parliament, and present an account of its core features. In the next section we draw on the information provided by this new dataset to provide an account of the character and trends of group switching in the ep. The dataset contains data on all 4163 individuals who have held a seat in the ep since the first direct ep elections in 1979 and until the end of the sixth legislature in June 2009. As a starting point, we have drawn information from 100 ana maria evans and m a arten peter vink the Hix et al. dataset for the first five legislatures (Hix et al, 2006)2 and from Hoyland et al. (2009) for the sixth legislature.3 Although the Hoyland et al. database encompasses all meps since 1979. we started from the Hix et al. dataset because it includes more accurate (though not fully updated) data on national party affiliation.4 The reason for this is that the online ep archive, from which the Hoyland et al. database derives its information, for a long while included only information on the latest available national party position. The ep archive was updated in early 2010 and now includes information on the party platform on the basis of which an mep was elected into the ep at the start of each legislature. However, it does not mention changes of national party affiliation during legislatures.5 With the goal of enhancing the reliability of the dataset, we have cross-checked the data from both sources with information on the parliamentary curriculum of each mep in the updated ep archive, as well as with ep publications on the composition of political groups at the beginning and conclusion of each legislature. In case of doubt we consulted additional sources to verify, in particular, information on national party affiliation. After compiling all the relevant information, the next step was to merge the data into a single data row for each mep per legislature. For example, if an mep switches between party groups within one legislature, this would lead to multiple entries in an output file from the Hoyland et al. database, in line with the way in which this is noted at the ep archive. The Hix et al. dataset contains only one ep party group affiliation (per mep per legislature) and we used the ep archive and the Hoyland et al. database as sources for information on group switching during the first five legislatures. Our dataset is structured in such a way that all relevant information on a single mep per legislature is contained in one data row. The whole dataset thus has 4163 data rows, for all meps per legislature. Note that the total number of meps is significantly higher than the total seats in this period (3507) due to replacement of meps in the course of legislative mandates. The ep Group Switching Dataset includes a complete overview 2 The dataset is available at http://personal.lse.ac.uk/hix/HixNouryRolandEPdata.htm. 3 The automated database is available at http://folk.uio.no/bjornkho/MEp/default.htm. 4 For example, Enrico Ferri, an Italian mep, was elected into the ep in 1989 and 1994 on the list of the Partito Socialista Democratico Italiano, and - after a period in the Centro Cristiano Democratico between 1996 and 1999 - was elected in 1999 on the list of Forza Italia. In the Hoyland et al. database, Ferri is described as affiliated to Forza Italia in 1989 and in 1994 as well. In the Hix et al. dataset, Ferri is described as affiliated to the ccd in the fourth legislature (1994-1999), which is also partly incorrect if one takes into consideration this was not the national party that got him elected into the ep. In our dataset and in the analysis developed here, we consider only the national party platform on which each mep was elected to each given ep legislature. 5 The archive is available online at http://www.europarl.europa.eu/members/archive/. measuring group switching in the european parliament (1979-2009) 101 of all national parties and ep party groups, including a coding scheme.6 In the remainder of this section we discuss the coding of the different variables. With regard to party group switching, we consider as a switch any change in ep party group affiliation that is not merely the consequence of a change in the name of a party group. As long as one group can be seen as the natural successor' of another, for example, when the Progressive Democrats were renamed as Group of the European Democratic Alliance, in 1988, we code both groups equally and do not consider a move from one group to the other a party group switch.7 We count as switch any move between party groups with different party group codes both in between legislatures (switches taking place at the start of a new legislature) and during a legislature. Including switches at the start of a legislature may be less intuitive, for those who are familiar with the international literature on party switching, which normally focuses only on switches that occur during legislative terms. However, whereas in national legislatures the electoral platform and the parliamentary organization normally come together in the organization of a political party, in the ep these two dimensions of political competition unfold separately, as we have seen above, meps are elected on the basis of a national party platform, and not on a European party group platform, but once they enter the ep, the strategic and ideological choice that has to be made with regard to the organization of their legislative activity relates primarily to the (European) party group, and not to the (national) party. Hence, national party delegations and meps confront the choice of party group affiliation at the start of each new legislature and strategic group switches often occur here (see next section). In a significant number of cases meps switch both at the start of a term and also once, twice, or even three times during a term. We code all these switches separately including the timing of each switch (measured by week into the legislature, thus ranging from 0 to 60). We also code the group of departure and the new group for each switching occurrence. Furthermore, we code not only the ep party groups involved, but also the broader ideological party family to which each 6 The dataset can be obtained on request. Please contact Ana Evans (ana.maria.evans@ics. ul.pt). 7 Even when a name remains the same, this does not mean that there is no ideological re-direction of a group. The Rainbow Group in 1989 retained its original name despite a fundamental ideological redirection due to the exit of the Greens. The regionalists and radicals who were left behind would join the new Green/European Free Alliance group in 1999. In the latter case our coding captures both the strategic choice of the Greens to exit the Rainbow Group and to start an independent Green party group in 1989 as well as the regionalists and radicals to join in 1999. 102 ana maria evans and maarten peter vink group belongs, in order to capture if and how specific trends are associated with specific ideological groups. For each switch we also code whether it involves an individual or a collective switch. We code switches as collective where at least two persons switch at the same time (in the same week) from one same party group to another same party group. Frequently, this means whole national party delegations switching from one ep group to another (as the Conservatives did in 1992 when they joined the European Peoples Party). Sometimes collective switches involve a faction of a national party delegation. When two or three people make a move together from one same group and into another same group, we also code this as a collective switch. All other switches are coded as individual switches. With regard to national party affiliation, and as mentioned above, we include information on the party platform on which a given mep has been elected for each given legislature where s/he holds a seat. We code national party switches that occur in-between legislatures by comparing national party affiliation at the re-election time. In other words, if an mep is elected to the ep for a second or subsequent time, but on the basis of a different national platform, we code this as a national party switch'. We do not code national party switches during an ep legislative term due to lack of systematically available data. As for ep leadership, we include information on what kind of positions were held by meps during a legislative term, including ep (vice-) President, Quaestor, Committee Chair, Delegation Chair, and President of a party group (=member of the Conference of Presidents). This information is available on the website of the ep, via the Archive. We have used the variable "Leadership" from the Hoyland et al. database to read this information from the website and have subsequently added this to our dataset. At this stage we have not yet included information on the allocation of rapporteurships, crucial positions within the political system of the ep (see e. g., Kaeding, 2004). trends When we look at the data on the evolution of group switching in the European Parliament between 1979 and 2009, at least two things are immediately striking. First, whereas group switching is largely acknowledged as a significant phenomenon, and has been estimated by some to range between 2% and 11% of members changing party group affiliation (Hoyland et al, 2009, p. 147), our data suggest that these estimates are conservative. Table 1 shows that a total of 377 out of all 4163 meps switched affiliation between 1979 and 2009, amounting to an average "switch rate" of around 9%. Nearly a third of those meps switched more than once, leading to a total of 519 switches (12.5% of measuring group switching in the european parliament (1979-2009) 103 TABLE 1 Number of group switches per legislature (1979-2009) Legislature Number of switches per mep per legislature meps ching Total n.° of switches Switchers as % of meps Switches as % of meps meps Seats Ox lx 2X 3X 4X Total swit Total Total 1 1979-1984 542 8 1 0 0 9 10 1.63 1.81 551 434 II 1984-1989 614 10 11 3 0 24 41 3.76 6.43 638 518 III 1989-1994 498 104 4 0 0 108 112 17.82 18.48 606 518 IV 1994-1999 636 49 39 6 1 95 149 13.00 20.38 731 626 V 1999-2004 633 36 22 5 1 64 99 9.18 14.20 697 626 VI 2004-2009 863 49 25 3 0 77 108 8.19 11.49 940 785 Total 3785 256 102 17 2 377 519 9.06 12.47 4163 3507 total meps). We see a clear peak of switch behavior in the 1989-1994 and 1994-1999 legislatures. In the third legislature 108 meps, or nearly 18% of all meps, switched group affiliation at least once. Important episodes were the Greens splitting off from the regionalists and radicals and forming their own group, as well as the Conservatives joining the epp. In the fourth legislature, however, due to the fact that a large number of meps (46) switched two or more times, the total number of switches peaks at 149 (20.4% of total meps in that period). Whereas the number of meps switching (64) and the number of switches (99) decreased substantially in the fifth legislature, these numbers increased slightly in the sixth term (77 and 108, respectively). When we look at the variation across member states, we first see significant differences between countries and, second, two countries that stand out: France and Italy. These two countries account for more than half of all switches (283 out of 520). This should be related to the fragmentation of the French and Italian party systems, for example, when compared to the German party system. Between those two, the Italian case stands out with the 592 Italian meps in the whole twenty-year period, switching party group affiliation 181 times (30.7%). We come back to this when discussing party switching at the national level (Table 6, below). For some of the new member states, especially Poland and Romania, party group switching is also very significant, while in other new member states such as Cyprus, the Czech Republic, Estonia, Hungary, Latvia, Malta, Slovakia, and Slovenia no meps had switched party group in the first legislature after accession. 104 ana maria evans and ma arten peter vink TABLE 2 Number of group switches per member state (1979-2009) Member Number of MEPs per legislature who switch x times meps hing 1 n.° tches :hers f meps :hes f meps meps Seats state Ox lx 2X 3x 4x Total switc Total of swi Switc as % o Switi as % o Total AT 76 2 1 0 0 3 4 3.80 5.06 79 60 BE 164 8 6 0 0 14 20 7.87 11.24 178 146 BU 26 2 1 0 0 3 4 10.34 13.79 29 18 CY 6 0 0 0 0 0 0 0.00 0.00 6 6 CZ 24 0 0 0 0 0 0 0.00 0.00 24 24 DK 95 9 4 1 0 14 20 12.84 18.35 109 94 ET 7 0 0 0 0 0 0 0.00 0.00 7 6 Fl 64 0 0 0 0 0 0 0.00 0.00 64 46 FR 574 42 27 2 0 71 102 11.01 15.81 645 495 DE 571 12 0 0 0 12 12 2.06 2.06 583 540 GR 187 6 1 0 0 7 8 3.61 4.12 194 146 HI) 26 0 0 0 0 0 0 0.00 0.00 26 24 IT 479 61 38 12 2 113 181 19.09 30.74 592 495 IE 100 1 0 0 0 1 1 0.99 0.99 101 88 LA 10 0 0 0 0 0 0 0.00 0.00 10 9 LI 12 2 0 0 0 2 2 14.29 14.29 14 13 LU 49 1 0 0 0 1 1 2.00 2.00 50 36 MA 6 0 0 0 0 0 0 0.00 0.00 6 5 NL 176 4 2 1 0 7 11 3.83 6.01 183 164 PL 43 13 5 0 0 18 23 29.51 37-70 61 54 PT 141 13 1 0 0 14 15 9.03 9.68 155 122 RO 59 4 5 0 0 9 14 13.24 20.59 68 35 SK 14 0 0 0 0 0 0 0.00 0.00 14 14 SN 8 0 0 0 0 0 0 0.00 0.00 8 7 ES 329 28 8 1 0 37 47 10.11 12.84 366 302 SW 81 1 0 0 0 1 1 1.22 1.22 82 63 UK 459 47 3 0 0 50 53 9.82 10.41 509 495 Total 3786 260 98 16 3 377 519 9.06 12.49 4163 3507 measuring group switching in the european parliament (1979-2009) 105 Table 3 presents the information on collective versus individual switches, by party group. In line with the available literature, we find that in general collective switches clearly outweigh individual switches. However, our data also demonstrate that this partly depends on the party group context. For the party groups on the fringes of the ideological landscape the collective switches significantly outweigh the individual switches. For the party groups in the center this is much less the case, or even the opposite (as within the pes and epp). Figure 1 shows the distribution of switching over the legislative term. What we see here is that around a quarter of all switches take place at the start of the legislature. These are generally meps who have been re-elected into the ep and switch party right at the start of the legislature. We also find some meps who switch twice or even three times within the first month (15 in the period under examination). Furthermore, we observe high switch activities in the third year TABLE 3 Collective versus individual switches per party group (total 1979-2009) Party Group* Collective Individual Total COM/EUL/NGL 30 4 34 LU 0 1 1 RBW/ERA 8 3 11 G/Greens/EFA 18 12 30 S/PES 4 17 21 L/LDR/ELDR/ALDE 21 13 34 EPP/EPP-ED 12 20 32 ED 43 2 45 FE 27 0 27 EPD/EDA/UFE/UEN 27 8 35 ER/DR 7 6 13 ITS 21 0 21 EN/I-EN/EDD/IND/DEM 34 11 45 TDI/CTDI 30 1 31 NA 110 29 139 392 127 519 * Reference point is the party group from which an MEP switches. o ON FIGURE 1 Number of group switches per month of legislature (total 1979-2009) 140 120 100 80 60 40 20 u 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 □ VI 20 0 0 0 1 2 0 0 0 0 0 2 0 0 0 0 0 5 0 0 0 4 4 0 0 0 1 1 4 10 18 0 2 0 0 0 0 0 0 2 21 4 0 1 0 1 1 2 0 0 0 0 0 1 0 0 0 0 1 0 ■ V 42 0 0 1 0 0 0 0 0 0 0 0 0 1 0 2 2 0 6 0 0 0 0 0 1 0 1 20 0 3 0 2 1 0 0 4 6 0 1 0 0 1 0 1 1 0 0 1 1 0 0 0 0 0 1 1 0 0 0 0 □ IV 26 0 0 0 0 3 0 0 0 0 0 0 28 0 2 1 0 0 0 0 0 1 0 0 2 0 2 1 25 3 16 7 0 0 2 0 0 0 0 0 0 0 2 0 0 0 2 20 1 0 0 1 1 3 0 0 0 0 0 0 □ III 21 0 3 0 0 0 0 0 0 0 0 0 1 0 0 0 0 2 0 0 0 1 1 0 1 0 0 0 2 5 0 0 0 0 33 1 1 0 0 0 0 1 27 0 0 1 0 1 1 0 0 0 0 1 1 1 0 4 2 0 ■ 11 7 0 0 1 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 0 0 12 2 12 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 □ 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 3 1 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 > z > 2 > x Z > z c > > x H W Z w H W X < z measuring group switching in the european parliament (1979-2009) 10J of each legislature, before but also after the important mid-term moment when leadership positions are re-allocated. Looking closer into the direction of party group switching in the ep, Table 4 shows the overall movement between ideological groups between the start and the end of each legislature. The cells on the diagonal, highlighted in bold, represent those meps who stay within their ideological group or "party family" (we refer here to party families in order to highlight the broader strategic movement between ideological groups). First, when comparing row totals (start) and column totals (end) we see that in general there is a movement toward the two largest ideological groups, the Socialists and the Christian Democrats. Second, we see that while the Socialists mainly profit from competition with the orthodox Left, the Christian Democrats receive mostly from the Conservatives and the populist Right. Third, it should be noted that the movement is not unidirectional, and we also see (smaller) groups of people moving away from the ideological center. In Table 5 we look at office incentives and focus on the leadership functions that meps may hold in the ep, such as President or Vice-President of the ep, Quaestor, Committee Chair, or Delegation Chair. The results presented TABLE 4 EP Group Switches by Party Family (total 1979-2009) End legislature 4-> 4-> Start legislature* j-> t-i— 100.0%, Pearson Chi-Square: .771; Sig (2-sided): .380. in Table 5 are ambiguous. The expected direction of the relationship between the two variables ep Leadership and Party Switch holds only to the extent that the percentage of group switching is lower for people with a leadership position in one of the six legislatures (7.7%) than for those who have not held such a position (9.2%). However, the relationship between the two variables displayed in the crosstab is not statistically significant. Since there are relatively few leadership positions available within the ep, the effects of leadership may be expected to work more at a group level than at the individual level. TABLE 6 EP Party Group switch and national party switch EP party group switch Total No Yes No National No 3734 91.4%, 352 8.6% 4095 200.0% party switch Yes 43 63.2% 25 36.8% 68 200.0% Total 3786 90.9% 377 9.1% 4163 200.0% Pearson Chi-Square: 64-444, Sig (2-sided): .000 measuring group switching in the european parliament (1979-2009) 109 A factor with more intuitive explanatory force is party switching at the national level. If meps have switched party affiliation at the national level there are good reasons to expect that this may have consequences for their party group affiliation at the ep level. Table 6 shows that, indeed, meps who have switched party at the national level, and are re-elected in the European Parliament, are more prone to switching than meps who have not switched at the national level. National party switching is an important phenomenon in countries with volatile party systems, such as France and Italy, and we see that 40 out of the 69 meps who are re-elected into the ep on the basis of a new party list are from either France or Italy. discussion It is a well-known fact among observers of party politics in the ep that electoral losses at the time of European elections contribute to the disappearance of parliamentary groups at the start of a new legislative mandate in the ep. This occurs when such electoral losses mean that the parties that held together in an ep parliamentary group fail to assemble sufficient seats to form an ep group at the start of a new ep legislature. On the other hand, the election of members from national parties previously not represented in the ep contributes to the formation of new coalitions - and therefore to the emergence of new ep parliamentary groups, often representing new sets of political interests. National party realignments during the campaigns for ep elections are also well-known factors for altering the choices for coalescing in the ep - hence, changing the composition of ep parliamentary groups at the start of a new legislature (see Corbett et al. 2005).Yet the literatures on legislative switching and on ep party politics offer little analysis of the phenomenon of group switching in the ep. Our research seeks to fill this gap. The phenomenon of ep party group switching discussed in this paper highlights, to a greater extent than has been observed before, the fluidity that characterizes the composition of European parliamentary groups. On average 9% of all meps switch during legislative terms. Party group switching is a phenomenon that gained force especially in the legislatures during the 1990s, up to a maximum of 18% for the 1989-1994 term, with strong prevalence among representatives from France and Italy, though by no means limited to those two countries. With regard to our expectations derived from the literature, we indeed see a clear tendency of party group switches from the ideological extremes, both left and right, toward the center. We also observe that most switching takes place at the outset of legislative terms, with another peak around the half-term moment, when responsibilities rotate within the ep hierarchy. 110 ana maria evans and maarten peter vink Second, when discussed in relation to the degree of consolidation of the European party system, this phenomenon also raises questions with regard to the quality of democratic representation in the eu. At a normative level, legislative switching is regarded as a reflection of instable party identity, which raises questions on all the essential dimensions of an operating democratic system, namely on the accountability of elected representatives, on the strength and performance of party representation, on the meaning of party labels to voters, on the reliability of policy-making coalitions - meaning stable government-opposition alignments that allow politicians to make reasonable calculations when negotiating legislative reform. The ep is hardly a nascent democratic institution: it has been operating for over 50 years, and its members have been directly elected by the citizens of Member States for the past three decades. Yet its unique mix of electoral contests based on national party list, internal legislative organization, and political dynamics dominated by international party groups provides an institutional context that is not conducive to party system stability. From that perspective, party group switching may be seen more as a natural element, than as an aberration of the political system of the ep. In this paper we have provided a first comprehensive descriptive account of the phenomenon of party group switching in the ep. Explanatory work starts with a solid understanding of the nature of the phenomenon that needs to be explained and the extent to which it occurs. In that sense the evidence presented here complements existing studies that have focused more on ideological coherence of groups. These studies, albeit within limited time periods, emphasize the importance of ideological congruence between national party platforms and the policy positions of European parties (McElroy and Benoit 2007, 2009). Our dataset is geared toward explaining group switching and can be expanded with additional information, for example on seniority of meps, voting behavior on roll call votes, or rapporteurships. Now that comprehensive data about elected representatives in the ep become increasingly accessible, the way is open to a better and more comprehensive understanding of the phenomenon of party group switching. This paper takes a step in that direction. measuring group switching in the european parliament (1979-2009) 111 bibliography attina, F. (1990), "The voting behaviour of the European Parliament members and the problem of Europarties". European Journal of Political Research, 18 (5), pp. 557-579. bardi, L. (1994), "Transnational party federations, European parliamentary groups, and the building of Euro-Parties". In R. S. Katz and P. Mair (eds.), How Parties Organize: Adaptation and Change in Party Organizations in Western Democracies, London, Sage, pp. 357-372. brzinski, J. B. (1995), "Political group cohesion in the European Parliament, 1989-1994". In C. Rhodes and S. Mazey (eds.) The State of the European Union. Building a European Polity7. London, Longman, pp. 135-158. corbett, R., jacobs, E and shackleton, M. (2005), The European Parliament, London, John Harper Publishing. desposato, S. (2006), "Parties for rent? Ambition, ideology and party switching in Brazil's Chamber of Deputies". American Journal of Political Science, 50 (1), pp. 62-80. evans, A. M. (2009), "Bringing back a larger pie from Brussels: The adjustment of national party strategy through party switching in the European Parliament", cies e-Working Paper, available at http://www.cies.iscte.pt/destaques/documents/cms-wp62_Evans.pdf faas, T. (2002), "Why do meps defect? An analysis of party group cohesion in the 5th European Parliament". European Integration online Papers (eiop), 6 (2), available at http://eiop. or.at/eiop/texte/2oo2-oo2a.htm faas, T. (2003), "To defect or not to defect? National, institutional and party group pressures on meps and their consequences for party group cohesion in the European Parliament". European Journal of Political Research, 42 (6), pp. 841-866. hall, P. A. and taylor, R. (1996), "Political Science and the three new institutionalisms". Political Studies, 44 (5), pp. 936-957. heller, W. B. and mershon, C. (2005), "Party switching in the Italian Chamber of Deputies 1996-2001". The Journal of Politics, 67 (2), pp. 536-559. heller, W. B. and mershon, C. (2009) Political Parties and Legislative Party Switching, London, Palgrave Macmillan. hix, S. (2002), "Parliamentary behaviour with two principals: Preferences, parties and voting in the European Parliament". American Journal of Political Science, 46 (3), pp. 688-698. hix, 8.(2005), The Political System of the European Union, London, Palgrave Macmillan. hix, S., noury, A. and roland, G. (2006), "Dimensions of politics in the European Parliament". American Journal of Political Science, 50 (2), pp. 494-511. hix, S., noury, A. and roland, G. (2007), Democratic Politics in the European Parliament, Cambridge, Cambridge University Press. hoyland, B., sircar, I. and hix, S. (2009) "An automated database of the European Parliament". European Union Politics, 10 (1), pp. 143-152. kaeding, M. (2004), "Rapporteurship allocation in the European Parliament: Information or distribution?" European Union Politics 5 (3), pp. 353-378. kreppel, A. (2002), The European Parliament and Supranational Party System: A Study in Institutional Development, New York, Cambridge University Press. kreppel, A. (2004), "Moving in the other direction? The impact of domestic party system change on Italian meps". Journal of European Public Policy, 11 (6), 975-999. mainwaring, S. (1991), "Politicians, electoral systems, and parties: Brazil in comparative perspective". Comparative Politics, 24 (1), pp. 21-43. 112 ana maria evans and maarten peter vink MAiNWARiNG, S. and scully, T. R. (eds.) (1995), Building Democratic Institutions in Latin America, Stanford, Stanford University Press. marsh, M. and norris, P. (1997) "Political representation in the European Parliament". European Journal of Political Research, 32 (2), pp. 153-164. Maurer, A., Parkes, R. and Wagner, M. (2008), "Explaining group membership in the European Parliament: The British conservatives and the Movement for European Reform". Journal of European Public Policy, 15 (2), pp. 246-262. Mcelroy, G. (2008), "Intra-Party politics at the trans-national level: Party switching in the European Parliament". In D. Giannetti and K. Benoit (eds.), Intra-Party Politics and Coalition Governments in Parliamentary Democracies, London, Routledge, pp. 205-226. Mcelroy, G. and benoit, K. (2007), "Party groups and policy positions in the European Parliament". Party Politics, 13 (1), pp. 5-28. Mcelroy, G. and benoit, K. (2009), "Party group switching in the European Parliament". In W. Heller and C. Mershon (eds.) Political Parties and Legislative Party Switching, Basingstoke, Palgrave Macmillan, pp. 147-172. mershon, C. and shvetsova, O. (2005) "Electoral cycles and party switching: Opportunistic partisan realignment in legislatures". Paper prepared for Presentation at the 2005 Workshop of the Party Switching Research Group, University of Virginia, Charlottesville, July 10-14. reif, K. and Schmitt, H. (1980), "Nine second-order national elections: A conceptual framework for the analysis of European election results". European Journal of Political Research, 8, pp. 3-44. settembri, P. (2004), "When is a group not a political group? The dissolution of the tdi group in the European Parliament". Journal of Legislative Studies, 10 (1), pp. 150-174. shabad, G. and slomczynski, K. M. (2004), "Inter-Party mobility among Parliamentary candidates in post-communist East Central Europe". Party Politics, 10 (2), pp. 151-176. strom, K. and muller, W. C. (1999), "Political parties and hard choices". In W. C. Muller and K. Stroom, (eds.), Policy, Office, or Votes? How Political Parties in Western Europe Make Hard Decisions, ny, Cambridge University Press. verzichelli, L. (1999), "I gruppi parlamentári dopo il 1994. Fluiditá e riaggregazioni". Rivista Italiana di Scienza Politica, 26 (2), pp. 391-413. Received 05-06-2011. Accepted for publication 26-11-2011. Evans, A. M. and vink, M. P. (2012), "Measuring group switching in the European Parliament: Methodology, data and trends (1979-2009)". Anälise Social, 202, xlvii (i.°), pp. 92-112. Copyright of Analise Social is the property of Institute de Ciencias Sociais and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.