Introduction
According to Riker’s (Reference Riker1964) broad definition, one condition for establishing the federal character of a system is the presence of two levels of government over the same territory and population, and at least one of these levels should have an area of action in which its autonomy is recognized. For Elazar (Reference Elazar1990), non-centralization is one of the requisites without which there cannot be a federal system. Autonomy is not guaranteed by simply consolidating certain institutions, but by providing them with material content. That is why when defining the specific federal character of a system, the distribution of competencies acquires enormous importance. It is not a question of the constitution or the highest regulations providing executive, legislative, or judicial capacity to a state or to a part of a federal system, but of the latter being able to exercise these powers in a certain matter. One of the most delicate and controversial questions when formalizing a federal system is the distribution of competencies. Generally, studies of this type on a federal system have had a juridical–formal character and, more concretely, have dealt with the constitutional techniques for distributing competencies.
A great number of these elements are found in the Spanish political system, which makes it into an essential case study on how changes in the sociopolitical and economic context affect institutional design. Firstly, this is due to the open character of the Spanish Constitution on questions concerning the distribution of powers between the Autonomous Communities (henceforth AC) and the State, as we will see further on. Secondly, because the existing regulations leave the door open to subsequent decisions by political actors.
This research, therefore, seeks to fill the gaps that can be found in the detailed description of the different periods in the process of shaping the Spanish state. While this is the main aim, we will also try to provide an answer to several questions that emanate from the theoretical analysis. The first refers to the influence that the political context can have on the decentralization process; the second considers which comes first, fiscal decentralization or the decentralization of competenciesFootnote 1; the third and final question seeks to understand who is in charge of setting the rhythms in the decentralization process and whether this depends on some political actor or, on the contrary, on historical characteristics.
The article consists of four differentiated sections followed by some brief conclusions. The first is concerned with the theoretical framework on decentralization and the influence of contextual elements on the changes, and also with fitting the Spanish case into the analysis. The second explains the general situation of the Spanish system in terms of political decentralization. The third proceeds to test the theory in the case of fiscal decentralization. The fourth and final section analyses the governments existing in the country as a way of understanding the influence of external actors on the decentralization process. As a step prior to the conclusions, we collate the data obtained previously to clarify our research questions.
New institutionalism, shared government and case selection
Our theoretical framework for this analysis first considers the theses on neo-institutionalism proposed by Meyer and Rowan (Reference Meyer and Rowan1977), Zucker (Reference Zucker1977), and Meyer and Scott (Reference Meyer and Scott1983) amongst others; in addition, it also considers the theories of Watts (Reference Watts1999) and Bednar (Reference Bednar2011, Reference Bednar2015) on the decentralization processes in different states.
Focusing on the first part, we must first understand what neo-institutionalism is and why it is essential for understanding the different stages of decentralization in Spain. This approach seeks to understand political phenomena in a holistic way, in which different structures (such as the educational system), public opinion, legislation, and regulations help in analysing voting behavior and citizens’ conduct (Steinmo, Reference Steinmo2013).
In Scott (Reference Scott2001), established the basic pillars of this theory, which he classified as the regulative pillar, the normative pillar, and the cultural or cognitive pillar. This study considers all these areas but, unlike Scott, we are not interested in legitimacy but in which part of the process might have been influenced by the other – whether political decentralization influenced fiscal decentralization or vice versa.
This research, therefore, takes into consideration the rules and regulations of the institutional structures (Streeck and Thelen, Reference Streeck and Thelen2005), given that they are what will determine the process of decentralization in Spain. As the legal steps to be followed in forming an AC are constitutionally established, this means that the Constitution is the fundamental point of the decentralization process, rather than the influence of external actors, in line with the proposals of Immergut (Reference Immergut1992).
To this end, two elements must be considered in order to understand the importance of neo-institutionalism: Critical Junctures and Path Dependence. Starting with the latter term, many authors have defined this as one of the key elements of this theoretical current due to its ideal capacity for explaining social changes in all their complexity. In relation to Critical Junctures, it is important to mention that this term adds the element of institutional change in Path Dependence. It creates a dual model that alternates long periods of inertia with conjunctures of change deriving from the context in which the institutions are found. These events are “critical” as they provoke a break in the inertia and produce a result opposed to how the institution was functioning (Capoccia and Kelemen, Reference Capoccia and Kelemen2007). In this respect, our aim is to detect those Critical Junctures that might explain the reasons for the Spanish system’s evolution since democracy was restored.
However, it is also important to highlight why this case is useful for achieving the above. As Majeed, Watts and Brown (Reference Majeed, Watts and Brown2005) observed, the distribution of competencies determines the classification of the different existing federal models.Footnote 2 In this way, four differentiated models of federalism are created: the dual, integrated, delegated, and cooperative models. However, in spite of the efforts to classify them, all federal systems include variations not only in the distribution of competencies but also in other elements, which means that there are models of decentralization that are changeable rather than static (Marks, Hooghe and Schakel, Reference Marks, Hooghe and Schakel2008). That variability depends, amongst other things, on structural elements such as the structural design of the distribution of competencies (Bednar Reference Bednar2015; Hooghe et al., Reference Hooghe, Marks, Schakel, Chapman Osterkatz, Niedzwiecki and Shair-Rosenfield2016). In this sense, the variation of the systems does not always depend on constitutional elements, but also on political practice and on state and substate institutions.
Nonetheless, one element to bear in mind that is highly relevant in the Spanish case is the weight of concurrent competencies in the system. In countries like Germany or Austria, there are three types of competencies: those of the parts (Länder, province, region, etc.), those of the state and those that are concurrent or shared (La Pergola, Reference La Pergola1984). The same system was adopted when the Spanish Constituent Parliament decided on the type of decentralization for the country. That is why determining the weight of concurrent areas of competency is vital for understanding the system’s flexibility.
Thus, the Constitution is interpreted as a political decision that is based on a concrete reality; that is, it is based on temporary elements that end up becoming structural ones, in line with Path Dependence theory. This element is especially relevant in the case of concurrent competencies, where legislative capacity is divided between the whole and the parts (Beramendi and Máiz, Reference Beramendi and Máiz2004). In our case study, concurrent competencies have great weight, which has resulted in many jurisdictional conflicts being taken before the Constitutional Court. Its jurisprudence when determining or interpreting the limits of legislative capacity is a recognition of the power of the central government at the expense of the AC, forming one of the cornerstones of Path Dependence in the case study.
In spite of the importance of these competencies, the Spanish system does not envisage the creation of a body for promoting the inclusion of regional intentions, interests or points of view when defining the general interest. The Sectorial Conferences were only created when all the AC had reached the limit of their competencies, and have been in regular operation since the start of the 1990s. At present, they are considered to be insufficient for achieving co-governance (León and Ferrín Pereira, Reference León and Ferrín Pereira2011; Mondragón et al., Reference Mondragón, Elizondo, de la Peña, Juaristi, Mokoroa and Novo2015).
For the above reasons, the Spanish system is more complex and two lists of competencies are established in the Constitution: that of article 149 includes those of the state while those of Article 148 represent the potential competencies corresponding to the AC, pending their incorporation in the Autonomy Statutes. Thus, the residual clause of Article 149.3 implies that the competencies not included in Article 149 can be developed by the ACs on condition that they are detailed in their Autonomy Statutes. Otherwise, they will be exercised by the state. This has worked in favor of the system’s adaptability, permitting the distribution of competencies that can be substantially modified without having to alter the Constitution.
The system’s open character, the aggravated need for constitutional reform and the lack of political consensus on the model of territorial organization have led to a situation where once the AC is constituted, no further updating is permitted. In the case of the Spanish political system, some of the advances in strengthening the system’s elements proceed from Path Dependence. The trajectory followed over the last 40 years has made this system into a big laboratory, taking the elements of centralization and decentralization of the system to extremes. The absence of a federal culture has resulted in a greater and unlimited use of the full potential of these elements. It should not be forgotten that the design of the territorial distribution of power is still an important element in interparty rivalry.
As a result, in order to study the Spanish case, it is important to understand that it does not fit neatly into any of the different theoretical frames set out, because it includes elements from different categories, making its categorization difficult. The neo-institutionalist approach is helpful for understanding why these elements are important and how changes have occurred.
Phases of political decentralization
The fact that the Spanish system is a fluctuant one due to its design of shared competencies is a further argument for its study. In this section, we describe the evolution of the decentralization of competencies from 1979 to the present. We examine the extent to which this supports the point of view of Kollman (Reference Kollman2013), who argues that the most common tendency in any system with the characteristics described above is toward centralization. In addition, this will help us to verify the stages into which we can categorize the existence of Critical Junctures that produce alterations in the process.Footnote 3
Asymmetrical stage (1979–1983)
The first stage, which we term asymmetrical, is the initial stage in which the autonomy statutes are approvedFootnote 4 and a process of ceding competencies to intermediate territorial units begins. In this way, from the outset, the Constitution already recognizes an asymmetrical distribution. Two types of menu and some exceptions are established. The most complete menu is that of the fast-track AC (Catalonia, Galicia, and Andalusia) and those where the Foral RegimeFootnote 5 is brought up to date (the Basque Country and Navarre). The only limit placed on these is the list of the exclusive competencies of the state. However, this situation was more complex insofar as the Basque Country, but above all Navarre, had a constitutional mandate to update the Foral Regime through their Autonomy Statutes, which implied the incorporation of another asymmetrical bias into the model. These are the bases of the system that mark the path to be followed.
The more limited menu is for the rest of the AC, which would have to opt for the competencies contained in Article 148 following the agreement between the parties with the most representatives in the Parliament.Footnote 6 The exceptions are Valencia and the Canary Islands that had begun the autonomous process to form fast-track AC. Following long negotiations, it was agreed to redirect these two processes in the same way as the rest, but giving them the maximum level of competencies by means of two organic laws approved at the same as their respective Autonomy Statutes in 1982.
It is important to underscore that not all the AC opt to maximize the number of possible competencies, with the result that there are almost as many lists as there are AC. To this must be added the fact that, once the competencies have been recognized in the Autonomy Statutes, they are not exercised until they have been legislated or transferred.Footnote 7
Stage of centralization (1982–1991)
The process of autonomous decentralization encounters its first pitfall in 1982. There are two events that act to contain the decentralization process and even favor recentralization: the absolute majority of the PSOE (Spanish Socialist Workers’ Party) and Spain’s joining the European Economic Community (EEC) in 1986. Its absolute majority enabled the party in power to control the general political process and, consequently, the decentralization process without any need for political allies. As a result, this led to an increase in the number of cases taken before the Constitutional Court by the AC, given that the support of regional political parties was not required for the political stability of the country. For some authors, this shows how absolute majorities generate an imbalance between the capacities of the whole and the parts (Máiz et al., Reference Máiz, Beramendi and Grau2002). Thus, the strong legitimacy obtained from the PSOE’s electoral support and the flexibility of the regulations produced a strong centralization, leaving the AC in a weak position when it came to negotiating with the central government. This is one of the key elements for understanding the change of course produced in the institutional drift in this and subsequent stages.Footnote 8
Moreover, the second restraint arises due to the start of the process of ceding competencies to a supra-state entity. Joining the EEC entails passing decision making on some matters to the European institutions. The AC consider this process to be a factor of recentralization, given that no mechanism had been included to incorporate their positions on the decisions taken in the Community bodies, and the European system leaves the readjustment of competencies in the hands of its member states, which is why those with a system of co-governance had an advantage over the rest. In the case of Spain, this type of mechanisms did not exist at that time. In this way, the European directives affect the competencies of the AC, limiting their legislative capacity. This would produce a new change in institutional dynamics; although this occurred without any warning as the territorial system had not been prepared to face the community challenge.
Stage of homogenization (1991–2000)
Following the changes that occurred in the system, there is once again a period between 1991 and 2000 in which the institutional system maintains its course in spite of the changes in the government’s political color. From 1989 onwards, the Constitution establishes that those AC with limited competencies could start the process of requesting a widening of their powers. This process had begun in 1987 when some AC expressed a wish to reform their Autonomy Statutes.
The agreements of 1991 sought unification in three aspects: the number of competencies, homogenization of the terms used to designate them, and the functions exercised. Moreover, the creation of co-governance bodies was envisaged in order to pursue greater coordination on certain matters. However, these bodies were perceived as actors for state control of the activities of the AC (Máiz et al., Reference Máiz, Beramendi and Grau2002). It can be deduced from this that although the Autonomy Statutes contained a different list of competencies, their exercise had been rationalized through control of the transfer process.
In 1993, the first minority government of the PSOE took office, followed in 1996 by the first minority government of the PP (Popular Party).Footnote 9 These years saw the strategy of “coffee for everyone,”Footnote 10 which created a framework for homogenizing the distribution of competencies amongst all the types of AC. Thus, in both phases, we can see how the institutional dynamics proceeded in line with the proposals of Path Dependence.
Absolute majorities of the PPFootnote 11 (2000–2005)
From the point of view of regulation, this can be described as a stage of stability. Following the transfer of competencies in Health matters, the autonomous process is considered finalized in terms of the distribution of competencies. Nonetheless, in this stage, there is a recentralization through the extensive use of legislation, as had occurred when the PSOE governed with absolute majorities. Once again, there is a change in dynamics that, while its importance is not felt in this period, lays the basis for the following stage.
Stage of statutory reform (2005–2012)
The year 2004 sees a social change that produces a political change in the country. At the end of this year, the wish of some of the fast-track AC to revise their statutes can be observed. In spite of differences in their aims and aspirations, these can be generalized as three principal conflicts: consolidating concurrent competencies, guaranteeing the interests of the AC in European affairs, and guaranteeing the system of regional funding. There was no coordination on how to put pressure on the government, quite the opposite in fact, and this produced bilateral relations with the state.
Unlike previous processes, the two big parties do not agree on a common position and the issue of the territorial distribution of power becomes a part of interparty rivalry. In spite of that, five of the eight reforms approved after 2004 achieved a high level of consensus. Nonetheless, there is clear disagreement over the statutory reforms of Catalonia, principally, and the Balearic Islands and Aragón.
The AC hold a negative view of the process developed during the previous legislative period, given that it breaks with the autonomous model. This partly explains the recentralizing activity carried out by the PP when in office in the previous stage. These antecedents and the electoral results of 2008 result in a new drive for autonomy in the AC, especially concerning financial matters, but also with respect to competencies (Colino, Reference Colino2009). This drive is materialized in regulatory terms, but also in the reactivation of co-governance mechanisms and the creation of new ones like the Conference of Presidents, which recognize the decision-making capacity of the AC.
Recentralizing stage (from 2012 onward)
From 2012 onward, there is another stage of stability. As happened in the 1980s and in the year 2000, the absolute majority of the party in power is used to reform basic legislation affecting the field of the AC’s competencies. Under the cover of the economic crisis, a rationalization of the Public Administration takes place. This new stage seeks homogenization and control by the state. The AC accuse the state of excessive and extensive regulation on basic questions and of invading their legislative field; two appeals of unconstitutionality are filed.
Phases of fiscal decentralization
Having examined the process of the decentralization of political power, this section considers the process of ceding the respective sources of finances for managing the transferred competencies. In this way, the 17 AC have witnessed changes in their participation in the total nonfinancial expenditure of the public administrations and the revenues the latter receive.
In this sense, the construction of the autonomous fiscal system has been a gradual process, with fiscal competencies functioning as transfers linked to the effective cost of the services. Subsequently, the remaining financing instruments have come into effect, which included the five-yearly negotiation of participation in the state’s revenues or the cession of certain taxes to the AC, amongst other elements.Footnote 12
Stage of creation (1980–1986)
This starts with the approval of the first Autonomy Statutes in 1979. The main goal during this stage is to evaluate the competencies that are transferred, in order to guarantee funding for the services transferred with a sum equal to their effective cost. This adjustment is due to the fact that calculating the financial needs goes hand in hand with calculating those of the decentralized competencies. From this point onward, other elements enter into the equation. Thus, the funding that was being granted to the different regional powers was calculated on the basis of the services that had been transferred (Requejo, Reference Requejo2005).
In 1983, the Laws ceding taxes to each AC are approved, with the aim of concretizing the scope and conditions for carrying out this cession. One year later, the Inter-Territorial Compensation Fund (ITCF) is created in order to correct economic imbalances and make the principle of solidarity effective, which consists in ensuring that a suitable and just economic balance is established amongst the AC.
Therefore, the system only focused on financial sufficiency, seeking solely to cover the real costs of the transferred competencies, without seeking elements like co-responsibility, tax responsibility or financial autonomy, which are of great significance in a genuine process of decentralization (Gómez de la Torre, Reference Gómez de la Torre2010).
Definitive period (1987–1991)
During this stage, the transfer of the taxes assigned to each AC was completed. The guarantee of financing the effective cost of the transferred services ended, and it was necessary to establish the transfers of the percentage of participation in state taxes. In 1990 the ITCF was reformed, separating the goal of new investment from the goal of solidarity, restricting access to the latter by the AC with a lower level of development.
Stage of homogenization (1992–1996)
During this stage, the aim is to undertake the transfer of new competencies. The intention is to try and equate all of the AC by assigning them similar levels of competencies. At the same time, another aim is to guarantee the principle of solidarity among them. For the first time, limits are set on indebtedness. The coordination of indebtedness is addressed in order to meet the convergence conditions established in Maastricht.
Stage of co-responsibility (1997–2001)
In this five-year period, a new model of funding is approved, inspired by the basic principle of assuming a level of fiscal co-responsibility. In this way, a new financial instrument is established: shared taxation under the heading of the IRPF (Impuesto de la Renta para las Personas Físicas; income tax). This consists of the joint use by the central and autonomous administrations of a tax with the same basic structure, with each administration having a certain regulatory power. Two measures are adopted: in the first place, the area of cession of part of the IRPF is widened and, in the second place, certain regulatory powers over the ceded taxes are given to the AC (Ministerio de Hacienda, 2002). With this reform, the AC have delegated the management, inspection, and collection of the ceded taxes. This results in considerable progress in matters of co-responsibility in taxation.
Stage of equalization (2002–2008)
In 2001, a new system of autonomous financing is approved. This seeks to extend the principle of fiscal co-responsibility, which is established in the cession of new state taxes and the assignment of new powers to regulate the taxes already ceded (León, Reference León2015).
In addition, the Sufficiency Fund is introduced to cover the difference between the needs of expenditure calculated for each AC and their fiscal capacity. This Fund seeks to fulfill the principles of sufficiency and solidarity in a way that guarantees the provision of all the transferred services in equivalent conditions. The participation of state revenues is integrated into this fund and funding needs are calculated on the basis that all the AC have the same level of competencies.
The ITCF has reformed again and divided into two: the Compensation Fund and the Complementary Fund. The former is aimed at financing investment expenditure that promotes the creation of income and wealth, and the latter finances the costs of realizing investments or setting them to work, with a maximum of two years (Gómez de la Torre, Reference Gómez de la Torre2010).
Stage of reform (2009–2012)
In 2009, the existing system appears outdated. Diverse circumstances lead to a reform of the previous system of funding: the significant, asymmetrical, and heterogeneous increase of the population, the approval of six new Autonomy Statutes and the provision of public health services (León, Reference León2015).
The goal is to increase Equity and Sufficiency in financing the set of competencies transferred to the AC, in an attempt to guarantee that the basic services of the Welfare State are provided equally to all citizens irrespective of their place of residence (León, Reference León2015). To achieve this final goal a proposal is made to create new funds of autonomous convergence: on one side, the Competitiveness Fund, whose main aim is to reduce the differences in per capita funding amongst AC; on the other, the Cooperation Fund is aimed at increasing the resources of the AC with less revenue.
The third element of the reform in this period is based on an increase of Autonomy and Co-responsibility. Finally, the fourth and final element involves improving the system’s Dynamics and Stability and its capacity of adjustment to the needs of citizens (León, Reference León2015). To avoid instability, an automatic and explicit mechanism is established to update resources annually.
Stage of recentralization (after 2012)
In 2012, the Law of Local Administration Rationalization and Sustainability is approved. Resulting from the global recession that had been underway since 2008, this Law seeks to reconvert the competencies and management capacities of local governments. To this end, the state recovered part of the regional decentralization that had been carried out and ceded in turn to local governments.
During this period, the state’s main objective was economic consolidation. In this respect, big cuts were carried out and public expenditure was contained, which generated problems within the existing limits on the autonomy of the regions. In the same way, a recentralization took place to the degree that, with the introduction of limits on expenditure at all levels in 2012 and the search for budget stability, the central government gave itself the capacity to control, oversee and sanction anything that deviated from the limits the law had established.
From what has been set out above, it can be seen that in general terms tax decentralization is totally dependent on politics. This has meant that critical conjunctures in the system have gone hand in hand with changes of direction in institutional dynamics. Table 1 shows how political decentralization in taxation has been developed and when these Critical Junctures have occurred. It can be seen that the existing centralization–decentralization dynamic has passed through similar stages. In general terms, tax decentralization has always tended to give the regions greater quotas of freedom in terms of legislation, application, and assessment of the different taxes or the parts of the latter that have been ceded, as a means of adapting to certain clear criteria of subsistence and co-responsibility.
Table 1 Spanish federalization phases
Source: Elaborated by the authors.
With respect to the elements of the theoretical frame of neo-institutionalism, it is worth underscoring how big changes in the dynamics of competencies have been brought about by political changes that in general terms are a reflection of social changes in the country. Thus, the first of them occurs in 1982, with the PSOE obtaining an absolute majority and this produces a slowing down of the process of decentralization. To this must be added Spain’s joining the EEC in 1986, an unexpected factor of a different order that also caused the system to be readapted. Subsequently, the year 2000 marks a change whose full significance is not felt in the same period but generates tensions that lead to the following one, from 2004 onwards. In this period changes are indeed found in the system, resulting from the “transgressive” statutory reforms carried out by some AC governed by the PSOE. Lastly, the final stage of change detected corresponds to the economic crisis and the last absolute majority obtained to date. The year 2012 sees recentralization and an increase in the conflicts between the central and territorial powers.
Coalition governments
The political composition of the autonomous governments is the third point of interest in the present analysis, insofar as the latter is a reflection of the Critical Junctures based on neo-institutionalism. Since the publication of Riker’s seminal work, The Theory of Political Coalitions (Reference Riker1962), numerous academics have tried to understand coalition governments (Martin and Vanberg, Reference Martin and Vanberg2003). Thanks to the process of the formation of the AC, new relevant actors – institutions and political parties – come onto the scene and have a direct impact on national politics. In this case, when speaking of coalition governments, we are referring to the inclusion of members of different parties in the cabinet, leaving aside other possible considerations like parliamentary support (Lijphart, Reference Lijphart1999).
Recently, together with the consolidation of the Spanish autonomous system, different studies have been realized in order to provide a theoretical and empirical perspective on the evolution of the composition of regional executives in Spain (Hamann and Mershon, Reference Hamann and Mershon2013; Matas, Reference Matas2015; Field Reference Field2016). As there has been no coalition government at the national level, the political parties in power have to take the regional parties into account due to their significant presence in the Congress of Deputies and the Senate. These formations have played – and still play – a fundamental role in the governance of their regions of origin, as can be seen in Figures 1 and 2. Besides, as Matas explains, ‘sub-state-levels allow for a more comparative analysis among different coalition governments in the same country […], with a greater scientific rigor than we would hardly find in the comparative analysis among countries’ (2000, 9).
Figure 1 Government Composition by Autonomous Communities (AC). Source: Elaborated by the authors. Data from Coalition Government Observatory (http://www.ub.edu/OGC).
Figure 2 Party composition of governments. Source: Elaborated by the authors. Data from Coalition Government Observatory (http://www.ub.edu/OGC).
Thus, for a correct analysis of the coalitions, numerous variables have been proposed, from which we have extracted the different typologies of coalition governments that can be found. Table 2 shows the seven general variables used for such a study, as well as the typologies of governments derived from them. For our analysis, we focus on their size (number of participants), composition (typology of the participating parties), their parliamentary strength or importance (whether or not they have a majority), as well as their temporal duration. This last variable proves to be of special interest for being able to understand the type of temporal development that has occurred and its connection with the statutory and fiscal stages previously set out.
Table 2 Coalition government: analysis of variables
Source: Elaborated by the authors using data in Matas, Reference Matas2000.
The analytical approach to the composition of the autonomous governments reveals multiple points of differentiation. A global review of the data shows that there is a preponderance of governments with a majority, whether single party (38%) or coalition (21%), which points to a global tendency of opting for stability with support in the legislative body. Amongst the minority positions, it is single-party governments that stand out (24%), for which negotiation is a daily question. This tendency toward majority governments in all the AC denotes a relative stability in institutional dynamics.
When the focus is placed on differences amongst regions, a difference is found in the number of governments formed in each of the AC. The data provided by the Observatory of Coalition Governments show that the Basque Country (16), Canary Islands (15), Balearic Islands (15), Catalonia (14), and Galicia (14) are the regions where the greatest number of governments have been formed since their constitution as AC, whether as a result of governmental crises or because legislative periods were not completed. Extremadura (9), Madrid and Castilla-La Mancha (10) are situated at the opposite extreme. This fact points to governmental stability in the latter AC with respect to the previous group. These, together with Murcia (11), are the AC that has only had single-party governments, always led by a state-level party, the PP or the PSOE, irrespective of whether they have a parliamentary majority.
Nonetheless, this characteristic of single-party governmental stability is not found in the Autonomous Communities of the Basque Country, Catalonia or the Canary Islands, where the formation of government coalitions is a constant element. It is interesting that in these communities governments are formed that have differentiated territorial interests, with the influence this can have on autonomous, political and fiscal development.
All of this suggests that the relation between the state executive and those of the different Autonomous Communities is very different between the first and second group of regions described above. Moreover, as a general rule, it can be said that the existence of alternative regionalist or national policies results in a stronger dynamic of coalition forming. This question can also be appreciated in the party composition of the different autonomous governments (see Figure 2).
The composition of regional governments should signify the need for collaboration and negotiation between different parties. Once again, differences amongst the regions can be appreciated. The group formed of Madrid, Castilla-La Mancha, Extremadura, and Murcia, to which we can add Castilla y León, has only had governments formed of parties with a national scope. The extreme opposite occurs in the Canary Islands, Catalonia, the Basque Country, and Navarre. In these AC regionalist governments or governments involving collaboration between parties from both the national and regional levels are the norm. In Catalonia, there has not yet been an autonomous government formed solely by national parties, although governments have been formed in coalition with regionalists (3 out of 14). In both the Canary Islands and the Basque Country there has only been one case of a government formed by a national party alone. The exception to this group is Galicia, which is differentiated by the predominance of single-party governments (50%) formed by a party with a national scope (PP).
It can thus be affirmed that several territorial groups exist. On the one hand, those whose governability is guaranteed by single-party governments with a majority. In contrast to these, there are principally historical communities, or ones with a differentiated statute, where the tendency is for collaboration between different political formations. The third group is situated between these two extremes. Once again, this leads us to think that there is little inclination toward change in the institutions, which are situated on the line of Path Dependency. Even so, these differences amongst regions allow us to think that an in-depth study of their internal dynamics could contribute to a better understanding of these theories when applied to each of the regions.
Data analysis
Thanks to the foregoing descriptions of both decentralization and the coalition governments, we have prepared some figures that jointly explain the combined dynamics of the questions analysed. This will help us to establish whether the dynamics of neo-institutionalization that determined critical moments in the decentralization process are also found in the dynamics of regional government. Therefore, Figures 3 and 4 respectively show a summary of data on the stages of political and fiscal decentralization, with the types of government and their composition.
Figure 3 Power and fiscal decentralization and type of government by year. Source: Elaborated by the authors.
Figure 4 Power and fiscal decentralization and government composition by year. Source: Elaborated by the authors.
As there has not been a single coalition government at the state level, the data shown in the figures will help us to understand whether or not there is some tentative relation between the types of government and their composition, and how the decentralization process evolved. In both figures, the stages of decentralization are marked with a double-line column for political decentralization and a triple-line column for fiscal decentralization.
The goal during the first periods of the decentralization process was to increase the scope of the process itself. In those years, the number of coalition governments is somewhat small, while majority governments are the norm. But this situation changes with the new phase of political decentralization and minority governments become more common. From the start of the process until 2004 there is an increase in fiscal decentralization. In spite of this, it seems clear that the changes in the type of government show no relation at all when they are placed in the framework of this type of decentralization. Carrying out a test of political decentralization underscores how changes in the types of government could have influenced the decentralization process and vice versa. With these exploratory data, we are unable to prove the existence of a direct effect between them, but we can appreciate how changes in the typology of regional governments are added to the changes in the stages of political decentralization. This enables us to affirm the existence of Critical Junctures that produce a change in the institutional dynamics that will later affect both the governments and the dynamics of decentralization.
The relation shown in Figure 3 can also be appreciated in Figure 4. The norm in the first phases of decentralization was to find regional governments formed by state-level parties. But this prevalence declines as the process of decentralization advances. During the 1983–1992 period, when decentralization came to a halt and even went into reverse, the number of state-level parties in regional governments also fell. But they remained in the executives thanks to the increase in coalition governments formed with regional parties. As we advance in time, we can observe a stabilization amongst state-level parties governing alone or in coalition with regional parties, accounting for about 80% of the regional governments.
Although the descriptive analysis of the data does not allow us to establish the bases of the influence between the dynamics of decentralization and governmental dynamics, the data leave the door open for deeper analyses in each of the AC. In this way, and expanding on the above, it would be possible to put the specific context of each AC into the relation and thus broaden the analysis, as well as the conclusions obtained.
Main findings and conclusions
Thanks to the temporal analysis realized we are able to underscore the main characteristics of the decentralization process in Spain. The first of these, which is clear insofar as it was established by the legislation, is the existence of asymmetry. This appears not only in the decentralization process but also in the fiscal part. In spite of that, steps have been taken towards a certain convergence as a way of covering the proposed budget deficit.
In addition, we can stress that there might be some effects on the composition of regional governments, opening up a new line of research due to the relevance of the different types of government. However, we have been unable to determine which side exerts an influence on the other, due to the exploratory character of the data. In spite of that, we have been able to test how the dynamics of change proposed from the perspective of neo-institutionalism also affected the formation and composition of governments.
At the start of the research, we set out three basic objectives to be described and analysed within the theories of neo-institutionalism. The first sought to analyse the influence of the political context on the decentralization process. In this sense, the absence of coalition governments at the national level resulted in a depreciation of the influence of political actors other than those from the ruling party; nonetheless, the regional-level parties were able to exercise a certain influence when their support was needed to maintain governmental stability. In spite of that, we saw how the system has Critical Junctures which determine that counterpoint to the system’s tendency toward stability, thanks to the different changes in government and, above all, the electoral weight of the latter.
The second of the questions that we posed concerned which decentralization process preceded the other. Thanks to the way in which fiscal decentralization was directed and ordered, it took place depending on the decentralization of competencies. The fiscal agreements used for the process of decentralization are mainly dependent on regional needs, which is why they necessarily follow on from political decentralization. Once again, as we have argued throughout the text, fiscal decentralization is affected by the elements of change – mainly social – that influence political decentralization.
Finally, the third of the questions posed sought to determine who was in charge of the decentralization process. We find that it is the central government in a top-down model that decides the rhythms of the process of political decentralization. In spite of this, thanks to the analysis, we can pose a new question referring to the possibility that the composition of regional governments could have some influence on decentralization processes, as a means of putting pressure on state governments that on occasion require support to carry out their work.
In all the cases, we found that neo-institutionalization has helped in understanding the decentralization process in Spain. Thus, following the theoretical lines of the proposed model, this investigation shows how the system moves within a dual system, which combines long periods of stability with other specific periods that produce a differentiated functioning of the institutions and, as a result, of the decentralization process analysed in this text. Thus, the changes that arose following the elections of 1982, 2000, 2004, and 2012 meant that the system had to be adapted and redirected to meet the existing demands. Moreover, these changes took place without involving any institutional ruptures thanks to the open character of the Spanish system of decentralization.
Funding
The research received no grants from public, commercial or non-profit funding agency.
Data
The replication data set is available at http://thedata.harvard.edu/dvn/dv/ipsr-risp
Author ORCID
Ainhoa Novo Arbona https://orcid.org/0000-0001-6184-7916
Author ORCID
Sergio Pérez Castaños https://orcid.org/0000-0002-7775-8632
Author ORCID
Jonatan García Rabadán https://orcid.org/0000-0003-2252-8562
