Intra-party conflict and cabinet dynamics in democratic Spain ( ) José Real-Dato (University of Granada, )

Intra-party conflict and cabinet dynamics in democratic Spain (1977-2008) José Real-Dato (University of Granada, jreal@ual.es ) Paper presented at the 5 th ECPR General Conference. Potsdam, Germany, 10-12 September 2009 Very first draft do not quote without the author s permission. Comments and suggestions are welcome Abstract Managing internal party conflict is one of the rationales behind decisions over cabinet composition and its dynamics along time. A number of factors affect the success in using cabinet appointments and dismissals as an instrument to keep party factionalism under control. They are, basically, constitutional rules regulating government formation and the PM s powers, and the internal distribution of power within the party, and particularly the strength of the organization s leaders vis à vis party factions. In this respect, it is of most importance if the chief executive officer (Prime Minister, President) formally in charge of deciding cabinet composition also occupies the highest leadership position within the party. Focusing on democratic cabinets in Spain since 1977, this paper aims to illustrate the working of all these elements in promoting a successful use of cabinet appointments and reshuffles as instruments to control party factionalism. The conclusion is that, although constitutional rules are important, the key element to explain the success of the PM in using cabinet composition to control intraparty conflict is the degree of control over the party. 1

I. Introduction 1 The degree of intra-party conflict (defined as disputes between groups of party members or factions over policy alternatives and/or control over party resources, including public offices) constitutes one of the factors that may affect cabinet dynamics in democratic parliamentary systems (Laver 1998, 22). In this respect, managing internal party conflict is one of the rationales behind decisions over cabinet composition and its dynamics along time. As a starting point, it can be logically assumed that the main goal both of a government as a whole and its individual members is, on the one hand, within term survival that is, surviving during the present term, at least as it is profitable for one or more of the parties to remain in government and not calling for new elections in case it is constitutionally possible (Lupia and Strom 1995; Laver 2003). In parliamentary systems, within term survival requires maintaining sufficient parliamentary support along the period. Intra-party conflict puts such parliamentary support at risk. On the other hand, both parties and when constitutionally possible individuals in government aim to continue in power during the next term. If not managed, intra-party conflict may threaten the electoral expectations of parties in government. This may be interpreted from the point of view of the incertitude disputes among factions may create in potential electors concerning the final policy alternatives that the party will promote, so they may prefer not to vote or voting for other parties that offer less uncertainty in their policy proposals. (In spatial terms, voters may prefer those parties that, in an ideal one dimensional policy option or ideology scale, have a further mean position but less variance higher probability of delivering their pet policy alternatives than the party experiencing factionalist conflict). Thus, it is in the interest of the party and their members to present a cohesive public image. However, a collective action problem 1 This paper is one of the outputs of the CARRERAS ( Dynamic Analysis of Political Careers in the Spanish Political System ) research project, funded by an Excellence Grant of the Andalusian regional government (Department of Innovation, Business, and Universities) (P08-SEJ-04032). Besides, parts of the paper were also conceived in the context of the extinct ELIPARSIS ( Elites, parties, and party systems, 1868-2004 ) research project, funded between 2003 and 2007 by the Spanish Ministry of Science and Tecnology (BSO-2003-07384). 2

follows here, as the incentive structure party factions confront may present cooperation with other party factions less attractive than pursuing their own interests. 2 If the party is in government, ministerial appointments may be one of the instruments to manage intra-party conflict (as well as it is for inter-party conflict in government coalitions). As Dowding and Dumont point out it is logically in the interest of PM (and consequently, an imperative for her survival in office) to keep the different wings and factions in the party happy. [She] cannot ignore powerful colleagues that command great party or political support (Dowding and Dumont 2009, 8). Ministerial positions are among the highest reputed offices a politician may occupy during her career. Along this reputation incentive and maybe more importantly they are usually associated to the control of public resources that may be used by party notables to reward their followers and, eventually, pursue their policy preferences. Thus, ministerial appointments are a highly attractive payment for faction leaders 3 and an effective way to grant internal party equilibrium (in the sense of absence of conflict or, in any case, controlled conflict among factions). Nevertheless, if the incentives to keep disloyal to the status quo remain favourable that is, if the costs of moving from the present internal equilibrium are lower than the expected payoffs of defying it (and they frequently are) it is logically preferable for party leaders (more likely in the case of those with progressive ambition) 4 and their followers, that the former occupy a higher ranked ministerial position, even becoming PM. In consequence, internal party equilibrium as a result of appointing faction leaders to ministerial positions may be unstable. Besides, appointing faction leaders may result in bringing intra-party conflict inside the cabinet. A number of factors affect the success in using cabinet appointments and dismissals as an instrument to keep party factionalism under control. Firstly, there are the constitutional rules regulating government formation and, in general, the powers of the PM. For instance, such rules may limit the impact of factionalism in government 2 Here party factions are considered to be unitary actors, so it is assumed they have solved their respective internal collective action problems. A more precise specification of the problem of within party coordination would present it as a two-level nested game. 3 However, some faction leaders may not accept any ministerial post, being more selective. As Dowding and Dumont (2009, 9) also point out [o]ften powerful politicians can refuse posts they do not fancy and demand specific portfolios. 4 On the theory of political ambition, see for example Schlesinger (1966) Black (1972) Rohde (1979) Williams and Larscher (1993) Wayne and Kenny (2000) Palmer and Simon (2003) Maestas et al. (2006). 3

behaviour through granting the PM a privileged position against other party leaders in policy formulation or deciding cabinet composition. A second factor is party leadership, conceived as the internal distribution of power within the party, and particularly the strength of those occupying formal leadership positions vis à vis party factions. This seems rather obvious, since a strong leadership would disincentive party factions to break internal equilibrium, making easier the use of ministerial appointments succeed as a mean to control internal conflict. In this respect, it is of most importance that the chief executive officer in charge of formally deciding cabinet composition (Prime Minister, President) also occupies the highest leadership office in the party. In this respect, both positions may mutually reinforce, reassuring the dominance of party leadership over rival factions. This paper aims to illustrate the working of all these elements in promoting a successful use of cabinet appointments and reshuffles as instruments to control party factionalism, by focusing on democratic cabinets in Spain since 1977. It is structured as follows. Following this introduction, the second section offers an overview of cabinet dynamics in Spain during the considered period. This serves as a context for the exploration, in the third section, of the use of decisions over cabinet composition to avoid intra-party conflict. The final section summarises the most important findings. II. Cabinet dynamics in contemporary Spain (1977-2008) In the last 32 years of democratic regime in Spain (between 1977, year of the first democratic election after 38 years of non-democratic rule, and 2009) there have been (Table 1) ten legislative terms (one constituent plus nine ordinary terms), and eleven governments. All of them have been one-party governments, with only three parties having reached that position: the centrist Union of the Democratic Centre (UCD), the Spanish Worker Socialist Party (PSOE) and the conservative Popular Party (PP). The chief executive office in government the President of the Government, equivalent to the Prime Minister (PM) has been occupied by five persons: Adolfo Suárez, Leopoldo Calvo-Sotelo, Felipe González, José M. Aznar, and José L. Rodríguez Zapatero. In four of the legislative terms the governing party enjoyed an absolute majority: the PSOE in 1982, 1986, and 1989; and the PP in 2000. In the other cases, the winning parties obtained a large plurality of seats in the lower chamber allowing the 4

formation of one-party governments whilst leaning on regular or ad hoc support from other parliamentary parties (see Reniu 2002). Table 1. Three Decades of Democratic Cabinets in Spain (1977-2009) Legislative term Party Government in Prime Minister Parliamentary support Government duration Constituent UCD Adolfo Suárez (I) Minority 1977-07-05 >1979-04-05 First UCD Adolfo Suárez (II) Minority 1979-04-06 > 1981-02-25 First UCD Leopoldo Sotelo (I) Calvo- Minority 1981-02-26 > 1982-12-02 Second PSOE Felipe González (I) Majority 1982-12-03 > 1986-07-25 Third PSOE Felipe González (II) Majority 1986-07-26 > 1989-12-06 Fourth PSOE Felipe González (III) Fifth PSOE Felipe González (IV) Sixth PP José M. Aznar (I) Majority 1989-12-07 > 1993-07-13 Minority + parliamentary support (nationalist 1993-07-14 > 1996-05-04 parties) Minority + parliamentary support (nationalist 1996-05-05 > 2000-04-27 parties) Seventh PP José M. Aznar (II) Absolute Majority 2000-04-28 > 2004-04-17 Eighth PSOE Minority + parliamentary José L. Rodríguez support Zapatero (I) (nationalists parties + IU) 2004-04-18 > 2008-04-13 Ninth PSOE José L. Rodríguez Minority ( variable Zapatero (II) architecture ) 2008-04-14 > Source: Adapted from Real-Dato and Jerez-Mir (2009, 102). Note: UCD: Unión de Centro Democrático; PSOE: Partido Socialista Obrero Español; PP: Partido Popular; IU: Izquierda Unida (United Left) Figures 1 to 3 offer a synthetic view of the main organizational features of the Spanish cabinets and their dynamics in terms of global duration and stability in ministerial positions between 1977 and 2008. 5 It also offers the average ministerial duration at each legislative term 6 during the considered period. 7 Excluding the PM, the average size of cabinet is 16.9, ranging from 23 in Suárez s first cabinets and 14 in the 5 Data corresponding to the last legislative term that commencing in April 2008 and still unfinished while this paper is being written are still under construction. So this paper analysis will be limited to the previous terms, where complete information is available from. 6 Duration in ministerial position refers to the time a cabinet member keeps the same portfolio during a prime ministerial term. The term is considered to end the day before the next national election or, in case of government crisis, the election of the new PM. 7 A more complete overview of the main characteristics of contemporary can be found in Real-Dato y Jerez-Mir (2009). 5

first Aznar government. 8 Variations in cabinet size are related mostly to the presence of Vice-Presidents and ministers without portfolio, since the number of departments has remained more stable. In this sense, the average number of departments is 15.6 (with a standard deviation of 1.05). Ten departments appear in every government (sometimes with changing subheadings): Agriculture, Defence, Economy, Education, Foreign Affairs, Health, Interior, Justice, Labour, and Public Works. Most of the governments also included other departments, such as Industry (present in nine out of ten governments), Culture (eight governments) and the Ministry of Presidency (eight governments), the latter being usually in charge of horizontal political coordination among departments, and between the government and the supporting parliamentary group (during the third and fourth legislative terms this latter function was accomplished by the Ministry for Relationships with the Parliament). Figure 1. Cabinet Dynamics: Cabinet Size and Organization (by Events of change) (1977-2008) 25 23 23 23 22 20 15 19 18 19 18 15 15 18 16 16 16 16 16 18 18 18 17 17 17 17 17 17 16 15 15 14 14 14 14 16 16 15 15 15 16 16 16 16 16 10 5 0 Jul 07 (**) Feb 07(*) Sep 06(*) Apr 06(**+) Apr 04 Sep 03(**) Feb 03(*) Jul 02(**) Feb 01(*) Apr 00 Feb 00(*+) Apr 99(*) Jan 99(**) May 96 Dec 95(*) Jul 95(**+) May 94(**+) Nov 93(*+) Jul 93 Jun 92(*+) Jan 92(*+) Mar 91(**) Jan 91(+) May 90(*) Dec 89 Jul 88(**) Jul 86 Jul 85(**+) Dec 82 Sept 82(*) Jul 82(**) Dec 81(**) Sept 81(**+) Feb 81 Sept 80(**+) May 80(**+) Jan 80(**+) Apr 79 Feb 78(**+) Feb 78(*) Sept 77(+) July 77 Constituent Suárez I L1 Suárez II L1 Calvo- Sotelo I L2 Glez I L3 Glez II L4 González III L5 González IV L6 Aznar I L7 Aznar II L8 Zapatero I Cabinet size Ministers without portfolio Vice-Presidents (total) Vice-Presidents with portfolio Total Departments Source: Real-Dato and Jerez-Mir (2009, 106). 8 Through the rest of the chapter, the term government is used in the specific sense of prime ministerial term (that is, the period between two successive investitures, i.e. González I, Aznar II, etc.). Conversely, cabinet refers to the specific ministerial composition of the government between events of change (i.e., during the González II government there were two cabinets, one after government formation, and the second after the July 1988 reshuffle). 6

Note: Data only include ministers and, when appropriate, Vice-presidents. Key: (*) Cabinet adjustment; (**) Cabinet reshuffle; (+) One or more resignations; L = legislative term The President of the Government occupies a central institutional position within the Spanish political system (Heywood 1991, 98; Pérez Francesch 1998, 61). 9 In terms of the composition of the cabinet, this means that the PM has a great autonomy in deciding who enters into the cabinet and who exits, without any authorisation or ratification by parliament or any other State body. 10 This autonomy of the PM is enhanced by the limited formal requirements candidates have to accomplish with in order to be appointed ministers. 11 In consequence, decisions over cabinet composition are, first of all, derived from the PM s calculations regarding her political survival, and, as stated above, this depends on two main sub-goals keeping parliamentary support during the present term and granting re-election in the next term. This means to keep the support of the public, the PM s own party members, and, eventually, other political allies, as well as erode the public and political support of political adversaries. This is achieved through the policies the government implement, as well as other strategies (i.e. distribution of offices, political marketing, or the public perception of the behaviour of party and public officers). Secondly, who is going to be a cabinet member also depends on the PM s beliefs about the suitability of potential appointees. Finally, the PM s formal autonomy is constrained by the particular configuration of the decision situation, which is affected by her party co-members preferences and strategies, those of political allies and adversaries in other parties, as well as the expected impact of decisions over the level of public opinion support. 9 This is developed in detail in Real-Dato y Jerez-Mir (2009, 102-105). In brief, the main features that grant the Spanish PM such central position are: 1) the 1978 Spanish Constitution grants the PM exclusive powers for dissolving parliament and calling for new elections, summoning the cabinet and establishing its agenda, giving instructions to other ministers and solving conflicts among them; 2) the PM is the subject of the vote of investiture and the two mechanisms for withdrawing parliamentary confidence (the motion of censure and the confidence vote ); 3) this institutional centrality transmits into parliamentary elections which are mainly perceived as elections focused on selecting the PM (Van Biezen and Hopkin 2005, 120ff.; see also Heywood and Molina 2000). 10 Although it is the King who appoints or dismisses cabinet members, in practice this is merely ceremonial. 11 They must have Spanish nationality, being entitled with full electoral rights, and not being barred by judicial sentence to occupy public offices. Besides, the only public offices compatible with a ministerial position is that of member of one of the two chambers of the national parliament (either Congreso or Senado). 7

Figure 2. Cabinet Dynamics: Entries, Exits, Portfolio Changes, and Continuity (1977-2008) 25 20 15 10 5 0 Jul 07 (**) Feb 07(*) Sep 06(*) Apr 06(**+) Apr 04 Sep 03(**) Feb 03(*) Jul 02(**) Feb 01(*) Apr 00 Feb 00(*+) Apr 99(*) Jan 99(**) May 96 Dec 95(*) Jul 95(**+) May 94(**+) Nov 93(*+) Jul 93 Jun 92(*+) Jan 92(*+) Mar 91(**) Jan 91(+) May 90(*) Dec 89 Jul 88(**) Jul 86 Jul 85(**+) Dec 82 Sept 82(*) Jul 82(**) Dec 81(**) Sept 81(**+) Feb 81 Sept 80(**+) May 80(**+) Jan 80(**+) Apr 79 Feb 78(**+) Feb 78(*) Sept 77(+) July 77 Constituent Suárez I L1 Suárez II L1 Calvo-Sotelo I L2 Glez I L3 Glez II L4 González III L5 González IV L6 Aznar I L7 Aznar II L8 Zapatero I Entries Exits Portfolio changes Continue from last legislative term Source: Real-Dato and Jerez-Mir (2009, 111). Note: See figure 1. (*) Cabinet adjustment; (**) Cabinet reshuffle; (+) One or more resignations. Key: L = legislative term. Thus, since 1977, we can classify the different prime ministerial periods in Spain in two broad categories: 1) Stable governments. In these cases, the characteristic feature is the PM s wide margin of manoeuvre, which grants him the ability to control cabinet inner dynamics. This is based on a solid parliamentary support and a widely accepted or strong leadership within his party (Suárez I, González I and II, Aznar I and II, and Zapatero). What could be interpreted as cabinet instability (in Aznar II and Zapatero, see Figure 3) largely results from the strategic calculations of the PM. In this category all but Suárez s first democratic government have long global durations and the highest average ministerial durations (Figure 3). 12 12 Suárez I anomaly is basically the result of the de facto constituent nature of the term. During the term the main task the parliament confronted was the elaboration of a new constitutional text. Thus, as soon as it was passed at the end of 1978, the PM called for elections. 8

2) Governments with cabinet instability. Here, the PM s control over cabinet dynamics is highly undermined. These result from a weak or questioned leadership, internal party struggles, and/or the influence of major external shocks (mainly political scandals) undermining the PMs autonomy (Suárez II, Calvo-Sotelo, and González III and IV). As it is evident from Figure 3, these governments have (with exception of González III) lower global durations lower average ministerial durations (again, it must be taken into account the exceptional character of Suárez I). 13 Figure 3: Kaplan-Meier Survival Function Graphs by Prime Ministerial Term (1977-2008) 1.00 0.75 0.50 0.25 0.00 1.00 0.75 0.50 0.25 0.00 L Constituent Suárez I L I Suárez II L I Calvo-Sotelo I L II González I GD = 691 days GD = 603 days GD = 608 days GD = 1296 days AMD = 437.8 days AMD = 341.9 days AMD = 942.3 days AMD = 345.0 days L III González II L IV González III L V González IV L VI Aznar I GD = 1190 days GD = 1276 days GD = 962 days GD = 1406 days AMD = 832.5 days AMD = 736.0 days AMD = 643.2 days AMD = 1035.7 days 0 500 1000 1500 0 500 1000 1500 1.00 0.75 0.50 0.25 0.00 L VII Aznar II GD = 1415 days AMD = 710.0 days L VIII Zapatero GD = 1420 days AMD = 884.8 days 0 500 1000 1500 0 500 1000 1500 Source: Real-Dato and Jerez-Mir (2009, 113). Note: See figure 1. Key: Vertical axis = survival function; Horizontal axis = analysis time (days). GD = Global duration; AMD = Average ministerial duration. It is interesting to point out that cabinet stability is also positively correlated with the success of the respective PM to manage intra-party conflict. However, success in managing intra-party conflict is not a sufficient condition to guarantee cabinet stability. 13 A more extensive and detailed examination of cabinet dynamics in each of these types of governments can be found in Real-Dato y Jerez-Mir (2009). 9

Therefore, while in some governments characterized by cabinet instability Suárez II and Calvo-Sotelo premierships failure in managing party factionalism is one of the main reasons explaining the instability and lower durations, in the other two unstable governments (González III and IV), intra-party conflict was managed quite successfully, and it was not the main reason for governmental instability. The following section deals with the use of decisions over cabinet composition to avoid intra-party conflict and the factors influencing their success or failure. III. Using cabinet dynamics to make intra-party conflict manageable In this section, two different types of cases are considered separately. Firstly, those governments where internal party conflict was successfully managed, in the sense of that the potential destabilizing effect of party factionalism was eliminated and/or controlled under limits unable to affect cabinet stability (defined as above). Secondly, there are the cases where this did not happen, and party factionalism clearly became an element undermining cabinet stability. Figure 4. Cabinet members in party higher executive bodies (percentages) 100 80 60 40 20 0 July 77 Sept 77(+) Feb 78(*) Feb 78(**+) Apr 79 Jan 80(**+) May 80(**+) Sept 80(**+) feb-81 Sept 81(**+) Dec 81(**) Jul 82(**) Sept 82(*) Dec 82 Jul 85(**+) jul-86 Jul 88(**) Dec 89 May 90(*) Jan 91(+) Mar 91(**) Jan 92(*+) Jun 92(*+) jul-93 Nov 93(*+) May 94(**+) Jul 95(**+) Dec 95(*) may-96 Jan 99(**) Apr 99(*) Feb 00(*+) Apr 00 Feb 01(*) Jul 02(**) Feb 03(*) Sep 03(**) Apr 04 Apr 06(**+) Sep 06(*) Feb 07(*) Jul 07 (**) Constituent Suárez I L1 Suárez II L1 Calvo-Sotelo L2 L3 Glez I Glez II L4 González III L5 González IV L6 Aznar I L7 Aznar II L8 Zapatero I 10

Source: Own elaboration from data in Real-Dato y Jerez-Mir (2009, 108-109). Note: See figure 1. (*) Cabinet adjustment; (**) Cabinet reshuffle; (+) One or more resignations. Key: L = legislative term. A. Successful cases Successful management of intra-party conflict has been the main feature in Spanish governments since 1977. In all but Suárez II and Calvo-Sotelo governments (both corresponding to the first regular legislative term) party factionalism was kept under control. As the rest of the UCD governments, Suárez I was a minority single-party government. In its origins, the UCD was a coalition of 15 small liberal, Christiandemocratic, social-democratic, and regionalist parties, formed around a handful of relevant political personalities (barones), together with independents coming from the reformist sector of the Franco regime. 14 At the moment political parties were legalized in March 1977 (just three months before the election date) none of the parties forming UCD had a territorial structure nor were widely known by the general public. At the same time, Adolfo Suárez, who had been appointed PM by the King in July 1976, wanted to continue in office after the 1977 election in order to continue in command of the transition process. However, as Suárez had stood aside from party politics, he needed an electoral platform to run for office (Martín Villa 1984, 76). Such platform would be provided by the UCD. The alliance was a win-win solution for both partners. The UCD benefited from the wide public acceptance of the President of the Government who commanded the transitional process that had made possible the first genuinely democratic elections in 41 years and the support of the state bureaucracy, which helped to flesh out the coalition s territorial organization. In his turn, Suárez got an electoral label to run for office, even though he was not a formal member of any of the parties forming the UCD. The favourable electoral results, along with Suárez wide popularity and the consensual character of parliamentary politics during the constitutional term clearly favoured Suárez s leadership position before the barones of the UCD, who conserved 14 On the history of the UCD, see, for example, Huneeus (1985) and Hopkin (1999). 11

their control over the ideological party families that formed the UCD (Linz 2001). Suárez took this into account in his first democratic cabinet (July 1977), as eight of the the barones and other UCD significant members were appointed ministers. In total, a 63.2 percent of all cabinet members belonged to the main executive body of the UCD in this first cabinet, this figure ascending to a 77.8 after the February 1978 reshuffle (Figure 4). However, Suárez s leadership position against the party barones confirmed in December 1977 by the transformation of the UCD coalition into a single party, and the election of Suárez as its president allowed him to keep a significant degree of autonomy in designing the cabinet. Thus, in his first democratic cabinet, he maintained six out of the 20 ministers of his 1976 government. In addition, the vice-presidencies with portfolio (Defence and Economy) were given to independent personalities (a military and a prestigious economist, respectively) appointed in order to deal with two of the major problems during the constituent period namely, military distrust to the newborn democratic regime and economic crisis. As third vice-president, in charge of political coordination with the party and the parliamentary group, Suárez appointed his right-hand man, Fernando Abril Martorell. The need to combine internal representation of the UCD families with an inner circle of collaborators is the main reason for the comparatively large size of Suárez s cabinets and the higher number of ministers without portfolio (usually charged with a specific task, such as regional devolution, EEC negotiations, or administrative reform) (Figure 1). Moreover, in the February 1978 reshuffle, Suárez took the resignation of the minister of Economy as an opportunity to reinforce his dominant position within the cabinet, the reshuffle ending up with five Suárez s collaborators some of them also members of the UCD executive committee, which explains the increase in the percentage above mentioned entering the cabinet and Abril Martorell, since then Vice-president/Minister of Economy) (Lamelas 2004, 171-172; Abella 2006, 314). Felipe González also exerted an undisputed leadership over the PSOE during his almost 14 years as PM. Between his election as general secretary in 1974 and his appointment as President of the Government in December 1982, the PSOE overcame internal divisions firstly between the exile and the internal organizations, and then between marxists and social-democrats and managed to integrate under its label all the 12

social-democrat family (Méndez 2000; Sánchez Cervelló 2004). González s charisma was also important as a factor reinforcing mass public support the PSOE was the second parliamentary political force since the 1977 election and to consolidate an image of moderation of the party that increased its appeal as a solid alternative of government. González s leadership was complemented by the tight control exerted over the party organization by Alfonso Guerra vice-general secretary of the party with González between 1974 and 1997 and Vice-president of the government between 1982 and 1991. This organizational centralization and González s unquestioned charismatic leadership ensured the absence of any disturbing manifestation of party factionalism, contributing to greatly increase González s formal autonomy in designing the cabinet and, thus, reducing the importance of controlling intra-party conflict as a rationale behind cabinet composition. This also meant that the PM and the government had considerably more leeway in policy-making with respect to the party organization, which worked more as a highly disciplined electoral machine than as a source of policy input (López Garrido 1985; Heywood 1991, 102). In this sense, the lower percentage of members of the Executive Committee of the PSOE in González s first cabinet compared with those of UCD in Suárez and Calvo Sotelo s governments (Figure 4) may be interpreted as a manifestation of the former. Besides, the absence of other members of the main party body in the cabinet would reinforce González s and, particularly, Guerra s leadership over the party, as it reduced for hypothetical alternative candidates to party leadership positions the possibility of using cabinet as a platform to promote their personal images. This state of affairs was soon institutionalised, as in the 1984 30 th PSOE Conference it was introduced in the party statutes the incompatibility of being member of the federal Executive Committee with any executive public office (including cabinet positions), a constraint not appearing in other parties with the exceptions of the president of the party, its the general secretary, and the executive secretaries (Maravall 2003). This explains the continuity of low percentages of members of the party executive in cabinet in the following PSOE governments. The internal equilibrium within the PSOE was broken in 1991. In January, Vicepresident Guerra resigned after a political scandal involving his brother. Although he conserved his position within the party as vice-general secretary, Guerra s exit was seen by some cabinet and PSOE members the so called renovadores aiming to renew the part organization and questioning Guerra s tight rule (Prego 2002, 137). The March 13

1991 reshuffle was interpreted in terms of the struggle between guerristas (the alleged Guerra s supporters) and renovadores. González whose charismatic leadership within the party remained undisputed at every moment, and who also had become personally distanced from Guerra along years opted for keeping internal equilibrium within the cabinet between both factions, although clearly favouring the renovador sector. Thus, one of its main personalities, the head of the Economy department, Carlos Solchaga saw how some men of his confidence were appointed as ministers in major economic portfolios (Public Works and Agriculture). In compensation, some of Guerra s supporters were promoted into the cabinet and some of his main detractors (particularly the ministers of Public Administration and Culture) were dismissed. In the early 1990s the function of the PSOE as an electoral machine also started to debilitate, and in the 1993 election it lost its absolute majority. Along with the economic crisis or the rising of the Popular Party as a real alternative of government, there were other endogenous factors linked to the party that had damaged the its public image and could explain this electoral retreat, such as the internal conflict between guerristas and renovadores or a number of corruption cases that affected the party organization (particularly the Filesa affair in May 1991, related to the illegal funding of the party) and the government itself. 15 With the aim to stick to his promises of ethical and political renewal made during the electoral campaign, and to isolate the government from the contagion of party conflict and corruption cases, González formed a cabinet with an eminently technical profile, with a lower presence of the party. Thus, six portfolios were occupied by individuals not affiliated to the PSOE, less than a half were members of the socialist parliamentary group both the lowest proportions since 1982 and just two members belonging to the party s Executive Committee (also the lower proportion ever) (Figure 4). In addition, the ministers from the guerrista faction were all dismissed, while the main representative of the renovadores faction, Carlos Solchaga, was appointed spokesman of the socialist parliamentary group at the Congreso. González managed to keep intra-party conflict under control, isolating the government from internal disputes until they finally resolved in the 1994 33 rd party conference, where the guerrista sector was relegated from the party s executive although Alfonso Guerra remained as vice-general secretary until the next conference, in 1997. However, 15 A year after Guerra s resignation, the minister of Health also had to resign due to irregularities while he was president of the national company of railways (RENFE). 14

a number of political scandals during the term (Real-Dato y Jerez-Mir 2009, 120 n.15-17) ended by exhausting the party s chances to renew its plurality in 1996. Intra-party conflict was not an issue during the Aznar governments. Similar to González, in the years passed since the re-foundation of the PP in 1989 from the former Alianza Popular (Popular Alliance, AP) Aznar consolidated his leadership over the party organization confirmed by the high degree of presidentialization conferred by its statutes by managing to transform the PP into a winning party 16. Aznar rebuilt the entire party organization and the PP acquired a more centrist and liberal flavour through integrating into the PP the Partido Liberal (Liberal Party, PL) and the Partido Demócrata Popular (Popular Democrat Party, PDP) 17 in 1989, and the progressive incorporation of other members from the extinct UCD 18, and the party created by Suárez in 1982, the Centro Democrático y Social (Democratic and Social Centre, CDS). In sum, Aznar managed to unite under just a partisan label the whole Spanish ideological spectrum from the centre-right to the extreme-right (García-Guereta 2001). Despite his unquestioned leadership, Aznar took into account these varied sensibilities within the PP in designing the composition of his first government Tusell 2005, 90), as well as the need to transmit both to the public and his parliamentary allies an image of centrism and moderation. Thus, Aznar s first cabinet included members from the AP, the PDP, the PL or the extinct UCD, as well as independents. Besides, a general characteristic of the PP governments, in contrast to the PSOE, was the overlap between the core of the cabinet and that of the party (Figure 4). In example, in his first government, Aznar appointed the PP general secretary (Álvarez-Cascos) as first Vicepresident/Minister of Presidency; and the three vice-general secretaries (Rato, Mayor Oreja, and Rajoy) as second Vice-president/Minister of Economy, Minister of Interior and Minister for Public Administration, respectively. In 2000, with an absolute parliamentary majority, and an increased charismatic leadership exerted over the 16 Thus, following the unexpected defeat in the 1993 general election, the PP initiated an unstoppable race towards the power, winning the European Parliament elections in 1994, the local and regional elections in 1995 and, finally, the victory in the 1996 general elections. 17 Both the PL and the PDP had formed an electoral coalition with the AP in the 1986 general election. The PDP also formed coalition with the AP in 1982. 18 Actually, the PDP was formed in 1982 as a scission from the UCD, more specifically from members of its Christian-democrat sector. 15

party 19, Aznar s personal and political preferences prevailed over internal partisan equilibrium criteria in deciding the composition of the government. However, the overlap between the cabinet and the party main executive offices remained. The case of the first Zapatero government parallels to some extent to that of Aznar s. As his predecessor as PM, Zapatero had progressively consolidated his control over the PSOE after his election as general secretary in July 2000 (Campillo 2001). Despite some initial scepticism on his leadership skills (Rodríguez Aizpeolea 2004; Jaén and Escudier 2008), he was able to finish the internal crisis the PSOE had experienced since the early 1990s, exacerbated by the succession of electoral defeats, and integrate the party organization. The favourable electoral results obtained by the PSOE in the 2003 local and regional elections and, the unexpected victory at the 2004 general election 20, confirmed the view of Zapatero as the leader the PSOE had been looking for since González s resignation as general secretary in 1997. As it happened in the case of Aznar, the composition of the first Zapatero cabinet was a manifestation of both electoral pledges (i.e. its gender parity nature) (see Real-Dato and Jerez-Mir 2009, 107ff.) and the aim of reflecting the PSOE s internal diversity. He appointed several members of the parliamentary group (10 out of 16), as well as ministers representing the most important territorial branches of the PSOE from those regions with a higher electoral weight Catalonia, Andalusia or where the PSOE had been in office since the first regional elections as it was also the case of Andalusia, and that of Castilla-La Mancha and Extremadura. Concerning those cabinets members belonging to the PSOE s main executive body, their proportion was again much lower (Figure 4) than that of the PP cabinets for the reasons above mentioned. Another difference between Zapatero s and his predecessor s cabinets was in the case of the former none of the two vice-presidencies was occupied by a higher official of the party. Finally, an interesting feature was the appointment as Minister of Defence of José Bono, Zapatero s main rival in his election as general secretary and the candidate supported by the official sector of the party 19 A clear example of Aznar s complete control over the party was that he, and not the party s main bodies, selected his successor as candidate for the presidency of the government in the 2004 general election. 20 See Torcal and Rico (2004). 16

during the XXXV Conference in 2000. This appointment immediately calls upon the advise Dowding and Dumont (2009, 8) remind from Vito Corleone s to his son in the film The Godfather, about keeping one s enemies closer than friends, a form to control an potential internal opponent, as well as to guarantee leadership renovation in the PSOE branch and the regional government of Castilla-La Mancha, that Bono had presided since 1983. However, as Zapatero s leadership at that moment can be considered strong enough, Bono s appointment could also be better interpreted as a way for Zapatero to honour and reconcile with an important sector of the PSOE. B. Unsuccessful cases All the cases in the former section are examples of successful management of intra-party conflict, where the strategic use of cabinet dynamics by the PM played a central role. In contrast, Suárez in his second democratic government, and Calvo-Sotelo later, were not able to control party factionalism, which greatly contributed to government instability during the first democratic legislative term (Figure 3). The main factor explaining the failure to manage intra-party conflict in the first democratic legislative term was a weakening of Suárez s leadership over the UCD. One of the main reasons was Suárez s own leadership style. Despite becoming the President of the UCD in 1977, he always privileged his role as PM, leaving organizational issues to his collaborators (particularly his right-hand man in the government, Abril Martorell, and the UCD general secretary, Arias Salgado) (Lamelas 2004). Besides, he tried to impose an even greater unitary character on the party, bypassing the different ideological families that still were active. Thus, after the 1979 election, Suárez excluded the most important barones from the new government. This was a decision that caused great uneasiness within the UCD, and that would have devastating consequences on the future stability of the government. In the March 1979 election the UCD obtained in the Congreso only two more seats than in 1977 (168, eight MPs below the absolute majority). Once the Constitution had been passed, the new political conditions inevitably led to a period of normal politics, putting to an end the consensual manners that had characterized the previous term. Since parliamentary parties policies left no margin either for a coalition 17

government or post-electoral agreements (Reniu 2002, 204-214), Suárez finally formed a minority government that depended exclusively on the support of the UCD parliamentary group, its composition still reflecting the internal equilibrium among the foundational parties (Powell 2004, 164). Thus, with his decision of excluding the barones from the cabinet, Suárez was curtailing the future expectations of his government, as well as undermining his own leadership over the UCD. Besides, the external circumstances were not favourable at all. On the one hand, during the first ordinary legislative term, Suárez s popularity experienced a continuous decay (Gunther 1986, 471) exacerbated by the difficulties of the economic and the political situation (Hopkin 1999, 127ff.), the latter mainly associated to the problems derived from the process of regional devolution and ETA terrorist attacks, which fed military unrest and the threat of a coup d état. On the other hand, Suárez behaviour in the parliamentary arena did not help much to improve his situation. Suárez always felt uncomfortable with the requirements of parliamentary politics (Martín Villa 1984, 90) for example, he refused to intervene both in the investiture session in March 1979, and in the motion of censure presented by the PSOE in May 1980 (Real-Dato and Jerez-Mir 2009, 118, n. 12) creating an image of political passivity, skilfully exploited by the opposition, and particularly by the PSOE, by then the second parliamentary party. Suarez s inability to act as a factor of internal cohesion (Powell 2004, 160) set in motion struggles among party factions pervading both the cabinet and the parliamentary group. These internal conflicts regarding ideological and organizational issues later evolved into centrifugal forces that caused the disintegration of the UCD (Hopkin 1999; Gunther 1986). This progressive lack of control over the government was reflected by the three reshuffles during Suárez II. The one in January 1980 was provoked by the resignation of the minister of Culture who disagreed with his party s position against granting the Andalusian region similar treatment to that obtained by Basque Country, Catalonia, and Galicia. The devolution issue also brought the first electoral defeats to the UCD (in the Andalusian referendum 21, and the regional elections in the Basque 21 Held in February 1980, this referendum was aimed at deciding whether Andalusia would constitute its autonomy institutions in a similar manner that the so called historical regions. The UCD asked its voters to abstain, while the PSOE, the Communist Party, and the Andalusian Party asked for an affirmative vote. 18

Country and Catalonia) during the first months of 1980 (Abella 2006, 364-366), provoking a crisis within the cabinet and party that led to the following reshuffle. In May 1980, Suárez s intention to create a new vice-presidency in charge of regional policy triggered dissensions within the PM s inner circle, adding to the prior disputes between the Vice-president Abril and two ministers in the social-democrat family. To solve the crisis, Suárez proposed some of the barones to join the cabinet, but they refused as long as Abril still occupied a central position. In the end, the cabinet resulting from the reshuffle left almost everyone unsatisfied. Internal opposition and the motion of censure presented by the PSOE eroded Suárez s popularity. In addition, his relationship with Abril worsened, ending up with the resignation of the Vice-president in August. The September 1980 reshuffle was Suárez s last attempt to reverse the UCD crisis, bringing back the main barones into the cabinet. The reshuffle however did not stop internal criticisms but, on the contrary, fuelled the opposition within UCD parliamentary group, where some relevant members felt excluded. Unable to cope with both growing internal and external pressures (mainly, economic crisis, terrorism, and military unrest) Suárez resigned in January 1981 (Huneeus 1985, 323-327; Oneto 1981). Despite Suárez successor, Calvo-Sotelo was the only candidate with an acceptable profile for the different families in UCD (Christian-democrats, liberals, social-democrats, and suaristas or Suárez s supporters) (Calvo-Sotelo 1990), he lacked the public charisma and support to construct an effective leadership that could restore both party and governmental stability. At the moment of his investiture, in February 1981, the separation between cabinet, parliamentary group, and party leadership (the latter still controlled by Suarez s supporters) made even more difficult the coordination between the two and the control of centrifugal forces within the UCD (Huneeus 1985, 352ff.). An early manifestation of these forces was the resignation of the socialdemocrat baron Fernández Ordóñez to form his own party, which provoked a first reshuffle in September 1981. Three months later after the UCD s fiasco in the October Galicia regional election a new reshuffle was designed to save the critical situation. However, neither Calvo-Sotelo s election as President of the UCD (Calvo-Sotelo 1990, 71-75) nor the appointment of the baron Martín Villa as Vice-president in order to increase political coordination between cabinet and party were enough to reverse the situation (Martín Villa 1984, 117). The last reshuffle, in July 1982 (also after a major electoral defeat in the Andalusian regional election) was a desperate attempt to rescue 19

the UCD from the crisis before the unavoidable call for elections. The party s electoral strategy prevailed over the cabinet. Some ministers left it in order to occupy relevant positions in the organizational structure to prepare the upcoming national election, and Calvo-Sotelo ceded the UCD presidency to the party candidate, the Christian-democrat baron and Speaker of the Congreso, Landelino Lavilla. This could not stop the final disaster, confirmed in the October 1982 election, when the UCD obtained only 11 seats in the Congreso. IV. Conclusions This paper has illustrated the use of cabinet dynamics as a tool to manage intraparty conflict. As it has been stated, in the case of Spain, where all governments have been single-party ones, the effectiveness if this instrument in terms of avoiding party factionalism affects government stability is mainly related with the degree of centralization in party leadership and the reinforcement between this leadership and the formal powers the Constitution assigns to the PM. In the case of González, his leadership over the PSOE was so undisputed thanks to his charisma and the control of the organization exerted through his number two, Alfonso Guerra that party factionalism was not a main issue in his governments until Guerra left the cabinet and his opponents saw an opportunity to take the party s control from his hands. However, this internal conflict did not affect González leadership nor cabinet stability instability during González III and IV was mainly due to external shocks produced by scandals involving ministers and the party. In the cases of Suárez I, Aznar I, and Zapatero I, the PMs clearly took into account in designing the cabinet the different factions existing within the party. These three governments have some common elements. They are the first mandates (in the case of Suárez I, his first democratic mandate) and in the three cases the party in government has not absolute majority in parliament. So compared with González (who obtained a landslide in 1982 and revalidated an absolute majority in 1986) it can be argued that they considered their leadership over the party despite it was quite undisputed in the case of Aznar and Zapatero was not still consolidated enough to not taking into account the existence of different party factions or sectors as a main variable in the designing of government. In this respect, it is interesting the differences in their 20

second term in office between these three PMs. In the case of Aznar, as it has been shown, he did not pay as much attention as they did previously to representing different party factions in cabinet. Following the previous argument, it may be argued that circumstances changed the perception of his leadership over their parties, clearly associated to the absolute majority obtained by the PP in 2000. In the case of Zapatero in 2008, analysts agree that he still pays attention to the two main territorial branches of the party (those in Catalonia and Andalusia). In this case, the PSOE improved its results compared to those in 2004 although it did not obtained absolute majority. In the case of Suárez, his failure in during his second democratic mandate may be explained by an error of judgement, as these circumstances were absent the UCD hardly improved its electoral results in 1979, and the previous coalitional and coalitional organizational and leadership structure still prevailed in the party, despite Suárez and his collaborators efforts to transform it into a unitary party and the success obtained in commanding the transition process during the constituent term. So Suárez s decision of not taking into account party factionalism in designing his second government was a clear calculation error. In the case of Calvo-Sotelo, he lacked the leadership legitimacy over the UCD to reverse the centrifugal forces started while his predecessor was in office. Finally, the observed autonomy of the PMs decisions over cabinet regarding party factionalism may be seen as a manifestation of the executive face (Poguntke and Webb 2005, 5) of the presidentialization in the Spanish political system. However, as it has been acknowledged before (Real-Dato and Jerez-Mir 2009; Van Biezen and Hopkin 2005; Heywood 1991) the autonomy of the Spanish PM does not only depend on the formal autonomy granted to him by constitutional rules, although that clearly contribute to reinforce the central position of the PM and his leadership over the party. The evidence in this paper ratifies these assertions. In this sense, formal rules regulating government formation and the powers of the PM do not guarantee a successful management of intra-party conflict. In contrast, this is particularly dependent on contingent political factors, such as the PM s ability to impose his leadership over his party and gathering parliamentary support. 21