THE INFLUENCE OF POLITICAL LEADERSHIP IN ELECTORAL BEHAVIOUR: THE CASE OF THE AUTONOMOUS COMMUNITIES IN SPAIN Teresa Mata López (Universidad Complutense de Madrid) Manuela Ortega Ruiz (Universidad de Granada) Introduction Political leadership studies were secondary in the political researches until the 80 s years, when the first publications about this topic that consolidated this line of research, were published. Burns s work, Leadership (1978) was a reference for future studies, such as Kellerman (1984), Blondel (1987) and Tucker (1995). As for the effect of leaders on electoral behaviour, the most important researches were published in the 1980 s as well. Most of them, written by American political scientists, focussed on the presidential system, which made it difficult to extrapolate their findings to the parliamentary systems, where the citizens don t elect directly the President. Some authors as Graetz and McAllister (1987) that have studied this influence in the United Kingdom, reached the conclusion that leaders have a limited impact on voting behaviour. There are other researchers, however, that defend the opposite view (Clarke et al. 2004). The growing importance of the leaders in parliamentary democracies has benefited from the phenomenon known as presidentialization of politics (Biezen and Hopkin, 2005), which gives a greater role to the medias, so that the personal image is more important than the collective image of the political party. In Spain, studies on this issue 1 highlighted the importance of leaders on electoral behaviour, to the point that, for some, the influence of leaders is one of the main features of the Spanish political system (Gunther, 1992). In our research, we jointly review this issue in Galicia, the Basque Country also called Euskadi and Catalonia. Regional elections in Spain are decisive as a result of territorial decentralization of power. Thus, leaders who contend in these elections influence not only on their territories, but also in the rest of the country, because the implementation of many of the policies designed by the Spanish central government depends on they being carried out by the Autonomous Communities. In Spain there are seventeen regional powers Comunidades Autónomas (CC.AA.) Galicia, the Basque Country and Catalonia stand out among them because they are considered to be historical. During the Transition, Spanish politicians opted for the autonomous model of territorial decentralization power mainly to respond to the nationalist aspirations of these regions. As a result they enjoyed special treatment in the 1978 s Spanish Constitution, which recognizes their different identity from that of the rest of Spain. This singularity is also manifested in the sub-party system, especially in the existence of nationalist parties. 1 These works include Gunther (1986, 1991, 1992), Justel (1992), Gunther y Monterio (1994), Barreiro y Sánchez Cuenca (1998), Cínzos y Jiménez (2003), y Rico (2002, 2008, 2009).
The study of the influence of leaders in these communities is interesting, not only for its own characteristics, but also because in the last election, there has been a change of the party in government, in the three cases studied. On the other hand, regional elections in these regions don t match with the rest of the Spanish regions: while in the rest of the communities the regional elections coincide with municipal elections, these three CC.AA and Andalusia 2 hold its elections on a different date. The fact that they don t coincide in time with any other voting makes their study more interesting, as they often reflect a different electoral behaviour. Following the definition developed by Burns (1978:18), we understand leadership as the process by which persons with certain motives and purposes mobilize, institutional, political, psychological and other resources, in competition or conflict with others, in order to stimulate, capture attention and satisfy the motives of followers. In the case of regional elections, the party leaders who stand for elections put in place all these resources to influence the voting decisions of citizens. The purpose of the present study is to prove how leadership affects the voting behaviour in these three territories, and whether the leader s role is critical for the vote decision. To achieve this aim we base our analysis in those made by Guillem Rico (2008) who examines the influence of national leaders in Spain. In addition to checking the real impact of leadership, which is critical to understanding the importance of this phenomenon in the elections, we consider the direction of it. This impact can be of a positive character, when the favourable opinion about a leader leads the voters to vote for this leader s party. But it can also be of a negative character when an unfavourable opinion about a leader determines that citizens choose to the opponent. When we study the impact of leadership in the elections, we must take into account the variables that influence it. Robert Elgie (1995) differentiates between institutional and non institutional factors. Among the former there is the legal frame of the constitutional, the political parties and the electoral systems, that conditions the behaviour of the political leader (Elcock, 2001:62) as well as the structure of resources within the party. The last one is conditional on whether the candidate in the election is or not the leader of his party. If so, the members of his organization will follow unquestioningly their guidelines, which will strengthen its leadership. In the opposite case, his situation is weaker because he has to face also the leader of his own organization. The relationship between the leader and his party is also conditioned by the organizational structure of training. Leadership will be greater when there is unity and internal cohesion in the party (Elgie, 1995:19-20). 2 We have to note that in Andalusia from 1996 to today the regional elections have coincided with national elections. For this reason, we have not included this region in the present study.
Among the non institutional factors there are three that stand out: the historical background, social attitudes, and popular desires (Elgie, 1995:21-23). Regarding these factors, we note that the three CCAA have one feature that makes them special: their strong national identity. In principle, it would determine the behaviour of the citizens in the elections, their actions in the pressure groups as well as the popular wishes which are influenced by the nationalist aspirations of an important number of the population. In consequence we will firstly focus on de description of the political situation in the three communities, with a special emphasis in the characteristics of each territory. Secondly, we briefly analyse some of the features of the autonomic candidates, especially how they were chosen by primaries, by congress or by appointment of the political parties, the degree of knowledge of the candidates, their valuation, and its role in election campaigns. The study of the campaigns constitutes the third part of the research, in which we try to reflect how far it was focused on the leader or on the political parties, if other leaders, both national and regional, participated in the campaign, which might influence voting decisions. With this information, we analyze the influence of leadership in the election results. The study concludes with the exposition of its findings. 1. Political situation in Galicia, the Basque Country and Catalonia One of the characteristics that differentiate these CCAA from other regions is that the three gained access to autonomy since the 1970 s when the Automous State was first created in Spain. That enabled them to create their institutional framework before the other communities 3. The national identity feeling is greater in the three regions, so that the nationalist parties are important actors in these communities, increasing the complexity of the political map. Currently, they are present in all three legislatures, especially in Catalonia and the Basque Country. This sense of identity is manifested even in the national parties. In this sense, the Socialist Party in Catalonia is presented with the initials PSC-PSOE Catalan Socialist Party, in Basque Country with PSE-PSOE Euskadi Socialist Party and in Galicia with PSdG Galician Socialist Party. In the case of Popular Party (PP), this does not occur, although there is some autonomy of regional management over the national one. 3 The Disposión Transitoria Segunda of 1978 s Spanish Constitution establishes these three CCAA have a completed autonomy that the 153 article said (Andalusia has a completed autonomy directly by the 153 article). The rest of the CCAA have a limited autonomy, by the 143 article.
Galicia has enjoyed a great stability in the party system. In this case, PP has always been the largest party and has ruled with an absolute majority from 1990 to 2005. This year, the PSdG and the Galician Nationalist Bloc (BNG) signed a government pact to out Manuel Fraga, the leader of Galician PP and President of the Xunta Galician Government since 1990. He was replaced by the socialist Emilio Pérez Touriño, who could no repeat his charge in 2009. The PP won the elections this year, with an absolute majority and with a new leader, Alberto Núñez Feijoo. Touriño Perez s victory in 2005 was benefited by the running out of steam of the Manuel Fraga s government. The Socialist leader represented a new form of politics, more progressive, and assumed the propositions of Galicians, especially after the agreement with the leader of BNG, Anxo Quintana. Also, from the first time, Perez Touriño renounced to follow the Catalan way, that is the reform of the Statute made in that community, which assumed the inclusion of the PP in all negotiations to reform the Statute of Autonomy of Galicia. Finally, the proposed reform could not take place, which affected the Touriño and Quintana government s image. In the Basque Country, unlike the case of Catalonia and Galicia, not single political party has ever achieved an absolute majority. However, the Basque Nationalist Party (PNV) has been the most voted in all elections. The PNV ruled since 1980 to 2009, although with a different leader. Since 1998, the nationalist candidate was Juan José Ibarretxe, who was President until 2009. In this year Patxi López, the Socialist leader, increased its votes and, with the political support of the Popular Party, he was appointed President. For the first time, since 1980, the Basque government is headed by a non-nationalist politician 4. In the previous legislature (2005-2009), the PSE acted as a constructive opposition, so the party reached specific agreements with the PNV, his closest political rival. After the last regional elections in 2009, the nationalist group proposed to maintain this situation, although the PSE chose to agree with the PP to govern in Euskadi. Thus, a political agreement was signed between both parties, by which the PP gave parliamentary support to the PSE without being part of the socialist government, guaranteeing in that way the stability of government during the current legislature. The circumstances of the Basque Country made it possible that PSOE and PP, the two major parties in national elections, being together in the Basque government. 4 Nevertheless, we note that the PSOE participated in several governments in the PNV, specifically form 1987 to 1998, intermittently. This year, all directors of the PSE resigned their government posts, among other reasons, the approach of the PNV with Batasuna, considered training as the political arm of the terrorist group ETA.
In Catalonia, since the first regional elections in 1980, the coalition Convergencia i Unio (CiU), that is a national and centre-right party, has won most seats in Parliament, although in both 1999 and 2003, the PSC was the most voted party. Since the first regional elections in 1980 until 2003, the President of the Catalan government, Generalitat, was Jordi Pujol 5, who in 2003 ceded the leadership to Artur Mas in CiU. In that year, the Socialist Party negotiated to rule with Esquerra Republicana de Catalunya (ERC) and Iniciativa per Catalunya Verds commonly known as the Tripartito government so that for the first time in 23 years, the Prime Minister of Catalonia, Pasqual Maragall, was not nationalist. This situation was repeated in 2006, with a new socialist leader, José Montilla. In 2010, the three partners of government parties, PSC- ERC-ICV, did not get enough seats to form a parliamentary majority that could prevent a new government of CiU. During the Tripartito, in Julio 2006, they approved the amendment of the Statute of Autonomy, a fact that caused uneasiness in the rest of Spain, and especially in the Popular Party, who filed a constitutional complaint at the Constitutional Court. There was a nationwide campaign to stop the approval of the reform, which gave more attributed more powers to Catalonia. Other regions followed the Catalonian example and began their amendments of their statutes of autonomy. Both the rejection and the intention to match the powers of Catalonia by other Spanish regions resulted in a growing of the nationalist sentiments. Finally, after more than four years of deliberations, the Constitutional Court ruled against some articles of the Catalonian Statute 6, but not against the full text. This judgment was a major upset among the Catalans, who took to the streets to defend the original Statute. And it was a high hit for the Tripartito, which helped the rise of nationalist CiU coalition in the elections held four months later, in November 2010. The erosion of the PSC was also due to the difficulty in making decisions jointly with their partners in government, and more specific causes such as the fire in Horta de Sant Joan. As we have seen, the party in the opposition won in the last regional elections in these three regions. However, the circumstances in each are different. In Catalonia, CiU, the party that had ruled the region until 2003, returned to power with the same leader who lost in both 2003 and 2006. In Galicia, something similar happened but the Popular Party, with a different candidate, was only four years off the Galician government. Finally, in the Basque Country, after 28 years of nationalist government, in this election there has been an alternation in power, this time with a socialist leader who was already a candidate in the elections of 2005. The change of party in government in these communities helps to understand the importance of leadership in the elections. 5 CiU ruled with absolute majority in 1984, 1988 and 1992, and it ruled with relative majority in 1980, 1995 and 1999. 6 Judgment against the Organic Law 6/2006, July 19, of reform of Catalonian Statute of Autonomy (Boletín Oficial del Estado, July 16, 2010).
2. Regional candidates Given the fact that the situation before the elections in each community was different, we decided to examine the candidates in each individual community. These differences are mainly due to the date on which the elections took place in Galicia and the Basque Country were on March 1, 2009, and in Catalonia on November 28, 2010 and the number of parties running to the elections that won parliamentary seats while in Catalonia and the Basque Country there are seven parties with parliamentary representation, in Galicia this number is reduced to three. Thus, in the next paragraph we examine, on an individual basis, the candidates in each community. We follow a chronological order, beginning with Galicia, and then with Euskadi and Catalonia. 2.1. Galicia In this region, the coalition government s leaders, Emilio Pérez Touriño and Anxo Quintana repeated as candidates in the elections of March 2009. Alberto Núñez Feijóo, the Popular Party leader, was running for first time, getting a majority that allowed him to form a government. The three leaders were elected in their respective party congresses, but with different results. In the case of Pérez Touriño, he was appointed secretary general of PSdG at the extraordinary congress held on October 10, 1998, supported by 61, 56% of the delegates. This victory provoked a Galician socialist regeneration which was translated in an increasing electoral support. The positive results reinforced Pérez Touriño, who run for elections in 2009, remaining with the same electoral support. However, the BNG s loss of one seat, won by the PP, made Pérez Touriño to leave the Galician Premiership, relinquishing his post as general secretary of the PSdG as well. BNG chose as national spokesman the top in the party Anxo Quintana in November 2003 with over 70% of the votes of the 11 th Assembly. Quintana was not new in Galician politics, because in 2000 he was appointed Galician senator, after more than ten years as mayor of the Allariz Ourense town. In 2005, despite losing a significant number of votes, BNG reached an agreement with PSdG to form a government. According to this, Quintana was appointed vicepresident of the community. Nevertheless, in the 2009 elections electoral support declined, prompting Anxo Quintana resign his post in the party. In relation to PP, when Manuel Fraga left the presidency of the Galician government in 2005, he returned to Madrid as a senator from Galicia. This led to an internal party mobilization to seek a new leader. Finally, in January 2006, Núñez Feijóo was appointed President of PP in Galicia at the 13 th Congress, with a 96% of the popular delegates. His leadership, therefore, was reinforced by this support.
The fact that these leaders were nominated, more than three years in advance, as candidates for the 2009 elections, meant that the three were well known in Galicia. However, there is one essential difference: the level of knowledge of Pérez Touriño and Quintana is greater before and immediately after these elections, as they occupied the highest government posts in the Xunta. Table 1 also shows that there is little difference between the level of knowledge before and after the 2009 elections, although it is curious that the percentage of people who claim to know the winner of the election, Núñez Feijóo, decreased immediately after the election. Table 1: Level of knowledge Galician leaders a Before 2009 Elections After 2009 Elections Alberto Núñez Feijóo (PP) 79,00 78,10 Emilio Pérez Touriño (PSdG) 84,50 86,60 Anxo Manuel Quintana (BNG) 82,70 84,70 a Numbers shows the percentages of respondents who claim to know the leader and also give an assessment. This is to avoid including those who do not know the leader and simply claim it as a matter of social desirability. Source: CIS surveys ES2783 and ES2796. If we look at the valuation, Pérez Touriño is the most valued leader of the three candidates before the 2009 election, beating his rivals in more than half a point. It is also the only one who gets a score above 5, before the holding of elections (Table 2). However, immediately after, the most valued is Núñez Feijóo and the other two candidates reduced significantly their values, especially the nationalist leader, Quintana. Even as Núñez Feijóo is the most valued after the elections, he cannot overcome the value of Pérez Touriño before the elections. Table 2: Valuation of Galician leaders a Before 2009 Elections After 2009 Elections Alberto Núñez Feijóo (PP) 4,60 5,19 Emilio Pérez Touriño (PSdG) 5,33 4,41 Anxo Manuel Quintana (BNG) 4,60 3,40 a Numbers shows the media valuations from 0 to 10. Source: CIS surveys ES2783 and ES2796.
2.2. Basque Country Basque regional elections in 2009 had a quirk in relations to former elections: the radical left, close to the ETA terrorist world, could not arise. This fact influenced the participation, but especially in the election results. There were seven parties that won representation in the Basque Parliament. Nationalist Basque Party (PNV) was the most voted, followed closely by the PSE, which increased more than eight points over the elections in 2005. Other policy options were, in this order: PP, Aralar, Eusko Alkartasuna (EA), Izquierda Unidad (IU) and Unión, Progreso y Democracia (UPyD), The leaders of these formations have different political backgrounds, and especially they have enjoyed different support within their respective political parties, which undoubtedly has marked their leadership. The way to choose candidates shows the situation of the internal party cohesion. Juan José Ibarretxe was nominated by the PNV by the party s National Assembly in March 1998, in an atmosphere of tension in the Basque Country. His nomination was unopposed and with the support of the executive of PNV, as the appointment of Antonio Basagoiti, then years later, as president of the PP in Euskadi in the 12 th Congress July 2008, where there was no another candidate for the presidency of the regional party. In the case of Patxi López, he was nominated by the 57% of the delegates at the extraordinary congress of the PSE in March 2002. In the cases of minority parties, it notes the choice of Aintzane Ezenarro as the Aralar s candidate. Because of the existence of a considerable number of political parties in Euskadi, the principal leaders, Ibarretxe and López, are the best known, both before and after the elections (Table 3), although there is a candidate who is known by 80% of the Basques: Javier Madrazo. It can be explained because Madrazo was part of the Ibarretxe s government since 2001. The other candidates are better known after the elections, because the active role they had during the election campaign, among other reasons. As in Galicia, the Basque regional government president before, Juan José Ibarretxe, is the best known. Table 3: Level of knowledge of Basque leaders a Before 2009 Elections After 2009 Elections Juan José Ibarretxe (PNV) 92,00 92,90 Patxi López (PSE) 87,40 88,40 Antonio Basagoiti (PP) 58,00 70,10 Aintzane Ezenarro (Aralar) 32,00 54,70 Unai Ziarreta (EA) 37,80 45,30 Javier Madrazo (IU) 80,70 79,30 Gorka Maneiro (UPyD) 22,90 a Numbers show the percentages of respondents who claim to know the leader and also give an assessment. Source: CIS surveys ES2784 and ES2795.
Concerning to the valuation (Table 4), Ibarretxe is the best known both before and after the 2009 elections. There is also a candidate who exceeds the score of 5 after the 2009, Aintzane Ezenarro, the leader of Aralar, which in two months have increased near one point. All other candidates do not pass the score of 5, especially the leaders of the minority formations that are less valued. It should be noted the valuation of Antonio Basagoiti because, although he is the leader of the third political force in Euskadi, he do not exceeds three points before and after the elections. Table 4: Valuation of Basque leaders a Before 2009 Elections After 2009 Elections Juan José Ibarretxe (PNV) 5,41 5,56 Patxi López (PSE) 4,48 4,41 Antonio Basagoiti (PP) 2,71 2,79 Aintzane Ezenarro (Aralar) 4,35 5,28 Unai Ziarreta (EA) 4,20 3,87 Javier Madrazo (IU) 3,80 3,59 Gorka Maneiro (UPyD) 2,82 a Numbers shows the media valuations from 0 to 10. Source: CIS surveys ES2784 and ES2795. 2.3. Catalonia The situation in this community is similar to the Basque Country. The presence of seven parties in the Parliament made difficult for any of them to reach an absolute majority. CiU was the winner, with Artur Mas as its leader, in his third running to regional elections. Mas was nominated for president of Catalonia in January 2002 by the executive of his party, with an special baking by his leader, Jordi Pujol. This formula was also reproduced in the PSC, where the direction elected José Montilla as a candidate for the 2006 and 2010 elections. In both parties, internal divisions were silenced, reinforcing the political leadership of both. In the case of minoritarian nationalist parties ICV and ERC their leaders were chosen in primary elections with different results. Joan Herrera (ICV) was the only candidate presented to the primaries held in November 2009, winning almost all the votes. For Joan Puicercós (ERC) it was more difficult to be designated candidate. There were four candidates, and he only won 37, 22% of the votes. Finally, in relation to PP, Alicia Sánchez-Camacho was designated as candidate in the party s regional conference held in July 2008. It was a contested decision, because her rival, Montserrat Nebrera, won 43,2% of the votes. There were clear internal divisions of the Catalan PP, but Sánchez-Camacho strengthened its position with the support of the national leadership, and especially due to the election results in which the PP was the third party.
As it shown in Table 5, the best known candidate is José Montilla, ), his popularity could be a result of him being the president of the Generalitat government of Catalonia at the time of elections. This situation is similar in the cases of Galicia and Basque Country. As to the winner of the elections, Artur Mas, the percentage of Catalans who claim to know him after the regional elections increases in more than four points. Sánchez-Camacho, meanwhile, is the candidate that that registers a greater increase in the level of knowledge after the elections, among other reasons, due, among other reasons, to the good and unexpected results achieved by the PP in these elections. The lesser known candidate, Joan Herrera, was deputy at the Spanish Parliament from 2004 until shortly before the 2010 elections, as well as the spokesperson for the group in the Parliament, this can explain the low level of knowledge of this leader in Catalonia 7. Table 5: Level of Knowledge of Catalonian leaders Before 2010 Elections After 2010 Elections Artur Mas (CiU) 86,60 90,90 José Montilla (PSC) 92,80 93,60 Alicia Sánchez Camacho (PP) 57,20 81,30 Joan Herrera (ICV) 55,20 65,70 Joan Puigcercós (ERC) 71,60 77,60 Joan Laporta (Solidaritat) 82,30 Albert Rivera (Ciutadans) 67,50 69,30 a Numbers show the percentages of respondents who claim to know the leader and also give an assessment. Source: CIS surveys ES2852 and ES2857. If look at the valuations reflected in Table 6, there are only two leaders who exceed the score of 5, Artur Mas and Joan Herrera, and the latter only after the 2010 elections. With the exception of Sánchez-Camacho and Puigcercós, the other leaders increase their valuations after the elections, although to varying degrees. The one with bigger increase is Mas, president of the Generalitat after these elections. It should be noted that, although Herrera is the least known candidate, he gets the second highest rating both before and after the elections. Table 6: Valuation of the Catalonian leaders Before 2010 Elections After 2010 Elections Artur Mas (CiU) 5,33 6,38 José Montilla (PSC) 4,37 4,47 Alicia Sánchez Camacho (PP) 3,02 2,90 Joan Herrera (ICV) 4,57 5,01 Joan Puigcercós (ERC) 3,89 3,58 Joan Laporta (Solidaritat) 2,65 Albert Rivera (Ciutadans) 2,42 3,04 a Numbers shows the media valuations from 0 to 10. Source: CIS surveys ES2852 and ES2857. 7 The great leader in ICV was Joan Saura, who until 2010 was the Catalonian regional minister of the Interior and Institutional Relations with José Montilla.
3. Electoral campaign According to the Electoral Law, the campaign takes fifteen days, and it ends at midnight of the day immediately preceding the holding of the elections 8. However, acts of propaganda begin much earlier, even if officially there are not election campaign activities. In this research we focus only on the legally stipulated period, for the reason that previous acts may be distorted, for example, because they are institutional acts. In Galicia, the national context greatly influenced the campaign, as the two major parties, the Popular and Socialist, are national parties. Thus, the campaign was focused in part in the corruption scandals of the PP and the management of the economic crisis by the PSOE. This influence can be seen in the participation of national leaders in the campaign. In fact, Rodríguez Zapatero, the Spanish president and leader of national PSOE, and Mariano Rajoy, leader of national PP, were very active in these elections. Nevertheless, Feijóo tried to do a campaign in which his figure would not be tied to the image of national party. Despite the strong presence of the national leaders more in the case of PP than in the PSOE case Galician leaders claimed as their own their parties proposals, accepting the full responsibility for both success and failure of their political parties in elections. So, Touriño assumed all the achievements of the government that he had headed from 2005 to 2009, and asked for a vote to continue the reforms undertaken. In addition, the Socialist leader focused all the criticism on Feijóo, who was blamed for the problems caused by Manuel Fraga, the former leader of the party in Galicia who was in charge of the government before Touriño (PSOE). Touriño, on his part, tried to be presented as separated of his partner in government, Anxo Quintana, since the scandals of his party were hurting the image of the socialists. He said even that he will not head the government at any price, in reference to the possible re-release of PSdG-BNG pact. For its part, Feijóo focused on criticism towards Touriño and sought to distance from the image of national PP, as mentioned above. However, he relied on the support of national leaders, especially in Rajoy, trying to put a clear distance from the previous Popular party s governments in Galicia. Feijóo did not make great electoral proposals, and hoped to reach a success basing it in both the figure of Rajoy and the failings of their opponents, especially Anxo Quintana. The nationalist leader, in fact, was involved in several scandals during the campaign, that significantly hurt him. Although he intended to develop a positive campaign, putting the emphasis on the Galician nationalism and identity, several episodes were known that marred their proposals 9. 8 Organic Law 5/1985, June 19, about General Electoral System (BOE, June 20, 1985; last reform, March 24, 2011). 9 Quintana was accused of misleading a group of seniors who, unknowingly, had to attend one of his speeches. It was caused a major upset in this group and other parties. He was also photographed on the yacht of a building contractor Galician, which demystified the leftist leader, who was accused of selling out the interests of these entrepreneurs.
If we look at the Basque Country, the campaign was marked by the illegalization of Democracia 3 Millones (D3M) 10, which lead to serious clashes between political forces, particularly to street violence against the other political Basque parties in several parts of the community. The economic crisis also influenced the election campaign, although the region had a better position than the rest of Spain. These two questions guided the leaders speeches. While these issues were central to the campaign, the possibility that Patxi López will assume the presidency conditioned all the proposals of the parties. Nationalist forces radicalized his message, saying that if López was lehendakari president of Basque government Euskadi would be ruled from Madrid. For its part, the PSE and the PP were not openly prone to pact between them to create a change of government, but both released messages that could be understood in this sense. For this reason, they avoided direct confrontation, although there were times when both parties discarded, in a symbolic way, the union. In this situation, the interventions of the national leaders were crucial. Both Zapatero and Rajoy accompanied their candidates at the beginning and the end of the campaign 11. The two leaders agreed to introduce the change of government as necessary in Euskadi. Thus, they criticized the work of Ibarretxe and the PNV and identified social stability with a non-nationalist government. But Basagoiti and López were accompanied by other strong national leaders. National parties, meanwhile, devoted his efforts to prevent the López presidency, emphasizing their independence aspirations. Ibarretxe tried to show that PSE was the Zapatero s party, to resemble the policies being carried out by the president of the central government with López s ones. PNV s strategy led to get the radical independence electorate s tactical voting. Also, Ibarretxe was supported by Artur Mas, who criticized the possibility that López and Basagoiti came to an agreement, considering that the pact would result an artificial situation in Basque Country. Meanwhile, Aralar defended more self-government, coming to call, in more than one occasion, the independence for the region. Electoral campaign, therefore, was led by two leaders: Ibarretxe and López, and totally focused in Basques issues. The possibility that the former president left the Basque government and the second got the support of the PP was a defining issue in the days before the election. Thus, the Basque leaders had a very visible role in the campaign. 10 D3M consisted of representatives of the former Batasuna, for which he was illegalized for being the heir to the political environment of ETA. 11 We must attend that the Basque elections coincided with the Galician, so national leaders had to be present in both communities.
Finally, in the Catalan case, the situation is more similar to that of Eukadi, as the majority party is a regional one, and the second is the Socialist Party. The electoral camping was attended by national leaders, but to a less extent than in the other two communities. The presence of Rajoy and Zapatero did not benefit the candidates of their parties, so they assistance in the campaign was minimal. The context in which the campaign was developed was once again marked by the economic crisis and its management by the central government and its impact in Catalonia. At the same time, the declaration of unconstitutionality of certain articles of the Statute was presented as an offense by the nationalist parties. Even PSC proposed the recuperation of the Statute as it was formerly approved in a referendum but without explaining on which type of legal formula it will be based. Yet, as it happened in Galicia and the Basque Country, the campaign focused on the possibility of a change of government. Artur Mas and José Montilla toured the community for support, personally representing the proposals by their parties. Also they were criticized by the opponent 12. The other candidates, except Joan Herrera (ICV), conducted a more aggressive campaign, fiercely attacked the Tripartito, including Puigcercós, leader of one of the parties of government (ERC). Herrera, meanwhile, preferred to focus on those policies pursued by his party in the government, especially all those related to environmentalism. Finally, Sánchez-Camacho (PP) claimed a more centrist policy for Catalonia, leaving aside the nationalist aspirations and highlighting the Spanish role in Catalonia. Therefore, the campaign in Catalonia, like in the Basque Country and, in a lesser extent, Galicia, was focused on the two leaders with a greater chance of electoral success: Mas (CiU) and Montilla (PSC). From the beginning there was a personalization of election campaigns in the three communities, and the top leaders of political parties assumed the role of representing the aspirations of their electorates. 12 Mas blamed to Montilla of the serious economic situation of Catalonia as well as the cuts of the Constitutional Court. For its part, Montilla accused to Mas that he had not proposals and the corruption cases.
4. The influence of the leaders in the voting decision After the analysis of the three regions political context, the features of the leaders who were concurring to the elections, and the electoral campaigns, we are going to study to what extent the leaders evaluation has had any effect in the voting decision. In this work, we focus on an individual perspective, according to the guidelines of most electoral behavior studies. For this we will work with the Spanish Social Research Center (CIS) database, specifically with the data of the post-electoral surveys of Galicia (number ES2796), the Basque Country (ES2795), and Catalonia (ES2857). These three surveys present a double advantage: they have a lot of variables which are essential to our study and they are very similar, allowing us to compare the results of each case with the others. Although the aim of our study is to know the leaders impact in the voting decision, we cannot obviate that there are other variables that have an impact in the voting decision. As Davis (1985) points out, to know the effect of a certain factor, we have to control all the preceding ones. In this case, with the purpose of knowing the leaders impact, we have opted to build a series of increasing complexity models with a multinomial logistic regression, in which the dependent variable is the vote of the last autonomic elections. The simplest model from where we start is the one which has all the control variables. Within these variables we have the demographic ones: gender, age, studies and working situation. The age is measured as the number of years in the survey moment. Studies is an ordinal variable with five leaves: with no finished studies (1), primary school (2), first stage of secondary school (3), second stage of secondary school (4), and university graduate (5). With regard to work situation, we have a variable with five categories: wage earner, self employed, retired, students and houseworkers 13. With regard to the actitudinal variables, we have taken into account the ideology, as the respondent s left-right self placement in a scale from 1(left) to 10 (right), and the party sympathy. The risks of including this variable was already pointed out by Rico (2004:348): it could be acting as an indicator of party preference, which would make it indistinguishable from vote decision. But this same author, basing his argument in Curtice and Holmberg (2005:239), stresses the bigger risk of not taking into account the possibility that the leaders evaluations would be just the evaluations of the parties. Therefore, we have introduced party sympathy to each party as an ordinal variable with four levels: very close (5), close (4), not close but not distant (3), distant (2), and very distant (1). In relation to the last kind of control variables, the contextual ones, we have taken into account the assessment of: the government actions, of the economic situation, and of the political situation. All of them are measured at the regional level, with three ordinal variables with five levels: very good (5), good (4), average (3), bad (2), and very bad (1). 13 There are some Works that have pointed out the relevance of the religiosity (Calvo y Montero 2002). But in this case we lack the appropriate data to include it.
Finally, with regard to our independent variable, the impact of the leaders evaluation, the works in political communication have reduced the leaders image s components to two big dimensions: their own personal features, and their positions in relation to the issues that appear in the political debate. In order to evaluate both, we are going to measure the impact of the leaders through their evaluations, measured in a scale from 0 (very bad) to 10 (very good). Rico (2004: 346 and 356) did point out that the leaders personal and political dimensions potentially contribute to their popularity, and that this popularity influences the vote decision through the leaders evaluations. Despite this fact, due to the lack of appropriate data, we are not going to be able to break down the effect of these dimensions. Nevertheless, and thanks to the inclusion, in the survey of Catalonian elections, of one question referred to the leaders personality, we can say that, in this case, only the 28 per cent considers that the important thing in the autonomic elections is the leaders personality 14. Before we begin with vote models, and again thanks to the inclusion of one question about the vote reasons in the three surveys, we can see to what extend to be the party with the better candidate is one of the main reasons for explaining the vote decision in the autonomic elections 15. Table 7 shows that in the tree regions the percentage of people that declare that this was the main reason for their vote is very low, especially in Catalonia. On the contrary, the main reason appears to be being the party that better defends the interests of the autonomous country, or being the one that better represents their ideas. Table 7: Reasons why they voted for a certain party a GALICIA BASQUE COUNTRY CATALONIA 1º 2ª 1º 2ª 1º 2ª Is the party that better defends the Autonomous Country interests 36,7 16,5 29,4 16,9 24,6 12,8 Is the one that better represents the ideas of people like me 20,8 36,5 27,1 20,7 25,0 20,6 Is the party with the better candidate/with the better leaders b 9,7 8,9 7,0 15,5 2,8 7,3 Is the party more able to govern 8,4 15,8 9,8 15,8 7,5 15,5 Is the most united party / the most trustworthy party b 2,0 3,8 1,5 3,9 19,7 26,9 Always have voted for this party 12,5 9,0 13,5 11,0 11,0 8,5 Others 10,0 9,4 11,7 16,1 9,3 8,4 (N) (1922) (1683) (1793) (1524) (1539) (1280) a The data reflect the percentage of respondents that affirm one of these as the main reason, or the second one, for his/her vote. b The second part of the sentences corresponds to Catalonia, where the answer s categories where slightly different Source: ES2796, ES2795 y ES2857 14 The question was to what extent they agree with this affirmation: In the autonomic elections the most important thing is the personality of the candidate 15 The question was identical in the case of Galicia and the Basque Country, but in the case of Catalonia, two of this answers categories where slightly different. These differences are shown in Table 7.
If we analyze these data in more detail, by voters of each party (Table 8), we realize that in two of the three autonomous communities (Galicia and Catalonia) the PP voters are those which, in a higher percentage, declare having voted the party because it has the best leaders. In the case of the Basque Country the higher percentage is for the PNV voters 16. Table 8: Reasons why they voted for a certain party a GALICIA BASQUE COUNTRY CATALONIA PP PSdG BNG PNV PSE EE PP Aralar CiU PSC PP ICV ERC Is the party that better defends the 35,2 29,3 60,9 52,7 11,2 8,5 18,8 35,5 6,5 2,9 5,9 47,0 Autonomous Country interests Is the one that better represents 15,1 30,9 14,1 8,8 39,1 46,5 49,6 10,7 39,2 35,3 51,0 23,9 the ideas of people like me Is the party with the better 13,7 7,0 4,4 12,4 7,4 1,4 2,6 3,3 1,2 4,9 0,7 2,6 candidate/with the better leaders b Is the party more able to govern 10,7 6,8 6,1 12,4 10,4 12,7 3,4 14,1 1,2 4,9 0,9 Is the most united party / the most 2,1 2,0 0,3 0,5 1,4 2,8 6,0 21,9 15,4 26,5 27,5 7,7 trustworthy party b Always have voted for this party. 12,7 16,3 7,1 10,5 17,2 14,1 4,3 6,4 29,2 14,7 11,1 12,8 Others 10,6 7,7 7,1 2,7 13,4 14,1 15,4 8,1 7,3 10,8 3,9 5,1 (923) (614) (297) (582) (366) (71) (117) (730) (260) (102) (153) (117) a The data reflect the percentage of respondents that affirm one of these as the main reason, or the second one, for his/her vote. The second part of the sentences corresponds to Catalonia, where the answer s categories where slightly different Source: ES2796, ES2795 y ES2857 These are only descriptive data; therefore, although they give us a general idea of the voters reasons, they don t tell us to what extent the leaders evaluation has any influence in the vote decision. We have to work with the models that we have already mentioned in order to know it. We have made two models for each political option that are included in the study. In the first one (control model) we have only included the control variables. In the second one (leaders model) we have added our main independent variable. We have done a multinomial logistic regression, with the memory of vote in the last autonomic elections as our dependent variable. 16 As it is explained in the section of the models analysis, the results of the Basque Country must be carefully seen, above all those related to the PP. the ones about PP.
Since the coefficients of this analysis reflect the contrast between two answer categories, we have fixed the party which has had the bigger percentage of votes as reference category: PP in Galicia, PNV in the Basque Country, and CiU in Catalonia 17. Before we analyze in detail our results, we have to warn that in the case of the Basque Country, we need to carefully interpret the results, especially those referred to the PP. Usually the voters of this party are reluctant to declare that they have voted for the PP. In this case, for example, despite the fact that the PP won more than two times the votes of Aralar, only 73 people admit having voted for them, against the 117 that declare having voted for Aralar. The problems derived from this fact can be seen for example in the PP s typical errors of Table 10. Because of this, and in order to avoid mistaken interpretations, we have decided not to analyze these results in the Basque Country case. With regard to the rest of the cases, Tables 9, 10 and 11 reflect the results of the logistic regression analysis in each region. They show that the variables that are significant are above all the actitudinal and contextual ones, specifically the party sympathy, the ideology, and the retrospective evaluation of government actions. Although it must be noted that the two last ones lose their significance when we add the leaders evaluations. In relation to nationalism, even though we expected a similar effect than to the ideology one, the results show that this is not the case. Its is only significant in Catalonia (with the exception of PP). In the case of the Basque Country it is only significant in the control model. Going back to our question -to what extent do the leaders images have any effect in the vote decision-, in every of the three cases (Galicia, Basque Country and Catalonia) the inclusion of this measure improves in a significant way the model 18. In Galicia and in the Basque Country cases we observe that, in the same way as in the Rico s (2004) study, the influence of the leaders follows a similar pattern to the party sympathy one. In general, as bigger is the closeness to some party and better is the party leaders evaluations, bigger is the probability that they vote for them. In the other hand, we have also observed the inverse relationship, as bigger is the distance between the voter and the party, and worst is the party leaders evaluations, bigger is the probability to vote for other party, although this does not happen in all cases. For example, in Galicia the evaluations of the BNG s leader have no influence in voting or not for PSdG. In the Basque Country, the evaluations of the PSE and its leader have no influence in the vote for Aralar, and vice versa. Finally, in Catalonia the only evaluations that have some effects are the CiU s leader ones, and the ones of the leader of the party which they are voting for. 17 With regard to the other answer categories, we have included the ones that have richened the bigger percentages of votes: PSdG and BNG in Galicia, PNV, PSE, PP and Aralar in the Basque Country, and CiU, PSC, PP, ICV and ERC in Catalonia. 18 Based in the increase of McFadden statistic (Tables 9, 10 and 11)
Table 9: Vote models for the Galician autonomic elections of 2009 a PSdG vs PP BNG vs PP Control Leaders Control Leaders Intersection 0,292 (1,686) 3,014 (2,244) 0,747 (2,083) 6,209 (2,583) Gender (female) 0,193 (0,299) 0,237 (0,419) 0,469 (0,366) 0,411 (0,488) Age 0,034 (0,147) 0,078 (0,191) 0,087 (0,183) 0,092 (0,230) Studies 0,008 (0,015) 0,037+ (0,019) 0,005 (0,018) 0,034 (0,023) Working situation b self employed 0,569 (0,446) 0,656 (0,578) 0,713 (0,534) 0,681 (0,683) Unemployed 0,365 (0,550) 0,678 (0,674) 0,261 (0,625) 0,361 (0,770) Retired 1,103+ (0,584) 2,255** (0,776) 0,010 (0,745) 0,563 (0,983) Students 2,196* (0,890) 1,383 (1,334) 2,242* (0,969) 1,436 (1,455) houseworkers 0,279 (0,640) 1,104 (0,876) 0,841 (0,906) 1,960 (1,448) Party sympathy PP 2,257*** (0,243) 1,899*** (0,280) 2,527*** (0,282) 1,960*** (0,329) PSdG 1,787*** (0,250) 2,056*** (0,310) 1,345*** (0,361) 0,793* (0,398) BNG 0,286 (0,231) 0,104 (0,282) 3,201*** (0,353) 2,135*** (0,388) Nationalism 0,092 (0,074) 0,020 (0,102) 0,193* (0,091) 0,014 (0,125) Ideology 0,992*** (0,128) 0,867*** (0,169) 1,053*** (0,155) 1,042*** (0,190) Economic situation 0,167 (0,215) 0,258 (0,276) 0,134 (0,266) 0,005 (0,333) Political situation 0,010 (2,243) 0,040 (0,341) 0,050 (0,294) 0,323 (0,394) PSdG action 1,240*** (0,271) 0,927* (0,370) 0,303 (0,363) 0,047 (0,483) BNG action 0,067 (0,245) 0,036 (0,319) 1,301*** (0,325) 1,169** (0,431) Leaders evaluations Alberto Núñez Feijóo (PP) 1,078*** (0,151) 1,007*** (0,162) Emilio Pérez Touriño (PSdG) 0,445** (0,154) 0,427* (0,201) Anxo Quintana (BNG) 0,276+ (0,144) 1,276*** (0,197) McFadden: Control model: 0,751 Leaders model: 0,,823 a The data reflect multinomial logistic regression coefficients. Typical errors in brackets. Statistic significance: *** p < 0,001; ** p < 0,01; * p <0,05; + p < 0,1 b The reference category is wage earner Source: CIS post-electoral study: ES2796
Table 10: Vote models for the Basque Country autonomic elections of 2009 a PSE EE PP Aralar Control Liderazgo Control Liderazgo Control Liderazgo Intersection 5,321 (2,025) 4,735 (5,088) 12,45 (7,941) 79.696 (19592) 0,225 (2,643) 3,190 (4,168) Gender (female) 0,037 (0,411) 1,582 (0,991) 0,665 (1,158) 8,026 (4007) 1,075* (0,483) 1,286+ (0,660) Age 0,030 (0,020) 0,010 (0,054) 0,008 (0,060) 0,062 (171) 0,020 (0,024) 0,029 (0,032) Studies 0,014 (0,210) 0,584 (0,647) 0,446 (0,550) 9,071 (2159) 0,130 (0,278) 0,137 (0,388) Working situation b self employed 0,286 (0,768) 4,968** (1,721) 0,414 (2,052) 8,206 (13313) 1,136 (0,919) 2,318 (1,595) Unemployed 0,817 (0,759) 1,929 (3,739) 1,892 (2,642) 10,488 (11483) 1,234 (0,894) 0,207 (1,342) Retired 0,396 (0,700) 0,947 (1,857) 0,573 (2,042) 2,718 (7937) 1,705 (1,192) 0,993 (1,593) Students 0,655 (1,404) 4,141 (8,398) 1,747+ (1,004) 0,641 (1,182) houseworkers 0,704 (0,790) 2,752 (4,263) 1,761 (2,330) 12,600 (9875) Party sympathy PNV 3,436*** (0,427) 4,949*** (1,224) 0,682 (0,971) 4,687 (4119) 3,353*** (0,406) PSE/EE 3,439*** 4,301*** 3,450+ 19,243 0,180 (0,428) (1,037) (1,917) (6296) (0,347) PP 0,525+ 1,915* 7,963*** 27,847 0,632+ (0,269) (0,963) (2,130) (3239) (0,358) Aralar 0,409 3,175* 1,368 6,808 3,356*** (0,275) (1,316) (0,895) (2290) (0,434) Nationalism 0,337** 0,388 0,513+ 2,487 0,271* (0,102) (0,277) (0,311) (602) (0,134) Ideology 0,499** 0,404 0,491 1,201 0,042 (0,163) (0,414) (0,536) (1487) (0,139) Economic situation 0,581* 0,563 3,540** 14,470 0,395 (0,274) (0,684) (1,168) (1570) (0,329) Political situation 0,299 1,128+ 0,933 0,878 0,077 (0,280) (0,640) (0,803) (1798) (0,306) Government action 1,539*** 0,859 2,025* 18,085 1,770*** (0,334) (0,916) (0,803) (2604) (0,432) Leaders evaluations Antonio Basagoiti 0,444+ 3,609 (PP) (0,266) (1125) Aintzane Ezenarro 0,372 4,411 (Aralar) (0,307) (778) J José Ibarretxe 1,863*** 4,417 (PNV) (0,464) (1208) Patxi López 2,612*** 0,819 (PSE) (0,693) (2427) McFadden: Control model: 0,751 Leaders model: 0,823 a The data reflect multinomial logistic regression coefficients. Typical errors in brackets. Statistic significance: *** p < 0,001; ** p < 0,01; * p <0,05; + p < 0,1 b The reference category is wage earner Source: CIS post-electoral study: ES2795. 2,555*** (0,531) 0,108 (0,489) 0,901+ (0,525) 2,680*** (0,601) 0,323+ (0,185) 0,050 (0,196) 0,603 (0,437) 0,056 (0,393) 0,948+ (0,568) 0,255 (0,163) 1,036*** (0,260) 1,274*** (0,281) 0,346+ (0,191)