122     Advanced Industrial Democracies
Taylor, Charles. 1991. Shared and Divergent Values. In Options for a New Canada, ed. Ronald L.
Watts and Douglas Brown. Toronto: University of Toronto Press. Trudeau, Pierre-Elliot. 1968. Federalism and the French-Canadians. Toronto: Macmillan of
Canada.
Tully, James. 1999. Aboriginal Peoples: Negotiating Reconciliation. In Canadian Politics 3, ed.
James Bickerton and Alain-G. Gagnon, 413-42. Peterborough: Broadview. Watts, Ronald L. 1999. Comparing Federal Systems. 2d ed. Kingston: McGill-Queen's University
Press.
Wheare, K. C. 1946. Federal Government. Oxford: Oxford University Press Young, Robert. 1999. Quebec's Constitutional Futures. In Canadian Politics3, ed. James Bickerton and Alain-G. Gagnon, 301-21. Peterborough: Broadview.
CHAPTER FOUR
Spain
Unfulfilled Federalism (1978-1996) Pablo Beramendi and Ramon Maiz
The Spanish constitution is often celebrated as a model of how successful transitions can, or even should, be made. Paradoxically enough, this picture is blurred precisely in the arena in which the results of the constitution-making process were most daring and innovative: the design of a new territorial distribution of power. The almost unanimous assumption that the constitution can provide the present and future skeleton for Spanish democracy turns out to be problematic when one takes into consideration its actual performance in the management of territorial cleavages in Spain. The aim of this chapter is to understand the achievements and shortcomings of the Estado de las Autonomias (EA)1 in the broader, comparative context of the capacities of federalism to deal appropriately with territorial conflicts.
Spam's federal institutions have been generally successful in managing territorial conflicts. In fact, their very design reflects the commitments that were integral to the new constitution's initial success during the transition. However, as time passed, these same features evolved in such a way that, while still contributing to effective management, became problematic. While the overall record of the EA in the management of territorial conflicts is a positive one, we will argue that the existing debates about the EA are mainly a consequence of those features that facilitated its initial success. The ambiguous nature of the constitutional contract was a necessary condition to complete
124     Advanced Industrial Democracies
the transition. Ambiguity in 1978 nurtured a short-term solution for problems among the different actors involved. However, some major institutional features agreed upon in the early stages of the process and the very dynamics generated by the EA itself are causing a change in the preferences of major actors within the system. As a result, several proposals to reform the system have emerged, pointing both to the weakness of an otherwise rather successful institutional design and to the possible directions of its evolution. \
The essay has four sections. The first presents the major features of the historical evolution of the relevant actors and institutions. The second section offers a brief description of the constitution of 1978. Section 3 focuses on two indicators of the performance of the EA: its institutional stability and the evolution of territorial identities. The final section analyzes the capacity of the EA to deal with territorial conflicts, based on the interplay between the intended and the unintended consequences of its original institutional design.
Actors and Institutions in Historical Perspective
The 1978 constitution brought far-reaching historical change to Spain. Centralism had been a dominant characteristic of the institutional organization of Spanish politics since 1810. Thus, the process of decentralization initiated with the constitution must be seen as a path-breaking moment, especially after nearly forty years of strongly centralized dictatorship. In fact, the EA represents the first enduring noncentralized polity in Spanish history and a direct political response to the existence of several competing identities in Spain. In what follows we account very briefly for the key factors that led to their emergence.2
Though our limited space leads us to focus almost exclusively on the nineteenth and twentieth centuries, we must take into account a number of elements inherited from previous periods. One such element was the survival of other identities, linked to the former territories of the Crowns of Castilla and Aragon, due to socioeconomic, ethnic, political, and institutional factors. Another factor accounting for the multiple identities in Spain derived from the fact that the notion of a Spanish identity was constructed differently by Liberals and Absolutists.
Thus, Spanish nationalism was doomed to be weakened from its very early stages. On top of this, the performance of the different governments between 1845 and 1876 did not contribute very much to the expansion and consolidation of a strong, widely accepted Spanish national identity.3 The lack of adniinistrative efficiency, increasing political conflict, and, last but not least, the slow and unbalanced pattern of economic development of the country also failed to make positive contributions to that process.
Spain 125
Despite the state's inability to forge a single national identity, a first step toward democratization was taken with the constitution of 1869. Four years later, another constitutional text proclaimed the First Federal Republic, linking, quite remarkably from our perspective, the processes of federalization and democratization. The constitution of 1873 established a federation with seventeen states, including Cuba and Puerto Rico. Each member state had its own constitution, parliament, and executive and judicial powers, guaranteed in the original constitutional contract. A senate, in this case a second chamber of territorial representation, was also considered. The localism of the left and the militarism of the fight doomed the First Republic to an almost immediate crisis, with two important consequences. First, the Spanish conservatives came to strongly associate any notion of decentralization with chaos, disorder, and anarchy. Second, regionalist movements were boosted by this experience, because, like significant sectors of the Spanish left, they saw in the federal formula an institutional solution to the need for democratization and the integration of opposed regional identities.
From i860 on, a change occurred in the pace of the formation of other competing national identities. The three main instances—the Basque Country, Catalonia, and Galicia—shared two major similarities. Between 1850 and 1890, all three regions witnessed similar developments. On the basis of historical and ethnic differences, cultural movements, driven mainly by literature and historiography, aimed at the recovery of some common roots of identity. Parallel to this cultural awakening, some early ideological formulations and political movements emerged. This was the moment of regionalism. Finally, after the disaster of 1898, the consolidation of nationalist platforms, organizations, and parties took place. The three regions also shared concerns about the inadequacy of Spanish governmental institutions, which were ultimately an underlying factor in any claim in favor of an increase of self-government. Nevertheless, despite these common factors, the experiences of the three regions differ substantially (Beramendi 1999, 83-88).
It may very well be argued that the Basque Country was the only region with a well-developed sense of identity during the first two-thirds of the nineteenth century. The Basque elites, composed mainly of rural nobility and clergy, had special privileges in their relations with the Church as well as their own institutional framework, the Diputaciones Forales. The Liberal Revolution erased the former and threatened the latter. As a consequence, these elites massively supported the Absolutist reaction and Carlism and were defeated in three wars (1833-39, 1846-48, and 1872-75). Progressively, as a result of these defeats, the Basque elites followed the path of Fuerism, an ideologically diverse regionalist movement, quite successful in expanding a sense of territorial identity among significant sectors of the population. The ultimate failure of Carlism forced the Basque traditionalist elites to seek alternative ways to accommodate
126     Advanced Industrial Democracies
Spain 127
their political preferences (De Pablo, Mess, and Rodriguez 1999). Switching their national identity from Spanish to Basque would eventually be their way ahead. Therefore, this traditionalist and very religious sector of the Basque Country would play a crucial role in the ultimate development of the Basque nationalist ideology and movement, conditioning quite heavily their further development (Aranzadi 1981; Azurmendi 2001; Juaristi 1987; Elorza 2001).
The experience in Catalonia was far less dramatic. The movement of cultural recovery, the Renaixenca, was quite successful in boosting the use of the regional language. The processes of agrarian modernization and early industrialization transformed the social structure of the region, launching more effective means for socialization and transmission of political ideas. However, despite its cultural expansion and ideological pluralism, Catalanism had hardly any political or electoral support at this stage.4
Finally, and in contrast with the other two cases, Galicia represents a case of continuity. The lack of significant economic changes facilitated the continued survival of structures belonging to a mainly rural society. The fidalguia, or lower nobility, were at the same time profiting from their land rents (foros) and were well positioned in the thick network of political clienteles around Madrid. Thus, they had no incentive at all to switch to another ethnic or territorial identity and, therefore, the regional language and culture became a drawback for those aiming at upward social mobility. For Galicia, regionalism was an issue only for a few isolated groups, incapable of attracting more significant sectors of society.
The final decade of the nineteenth century as well as the first three decades of the twentieth century witnessed the final stages in the articulation of the territorial conflict in Spain. In 1898, a new, more intense phase in the development of such conflict began. The loss of Cuba and the Philippines was the ultimate indicator of the ongoing crisis of the political machine of the Restauracion, built upon an enormous patronage system disguised as a parliamentary regime, administratively inefficient, economically ominous, and, not surprisingly, with a declining legitimacy as a territorial reference point. The so-called Disaster had two major consequences. It provoked a remarkable reaction and spawned changes in the formulation of Spanish nationalism, which in turn triggered the transformation of many regionalist movements into nationalist platforms and organizations. Eventually, the basic structure and participants in the territorial conflict in Spain would emerge from the convergence of these two processes.
Spanish nationalism was once again ideologically divided (Alvarez-Junco 2001). Though all the factions of the Regeneracionismo shared an organic conception of Spain, the two major branches trying to boost the pride of a nation humiliated by its failing political system had opposing strategies. On one side stood a clearly right-wing authori-
tarian vision, radically opposed to the claims of an increasingly organized working class and unwilling even to consider any regionalist or protonationalist idea. The defense of Spanish unity, at any cost, was their leit-motif, which achieved an initial success during the dictatorship of Primo de Rivera (1923-29) and would eventually be supported ad pedem literae by Franco and the Church and all the forces behind the coup d'etat in 1936. On the other hand, a more liberal and democratic vision of the Regeneracionismo, closer to the left side of the ideological spectrum, tried to promote intellectual renewal through the modernization of the educational system at all levels. It supported the social and economic demands of the working class and was open to interaction with regionalist and nationalist movements, which were increasing in salience.
The differences between the three major regional experiences continued to be significant, affecting both the scope and the ideological profile of each of the three nationalist movements. In the case of the Basque Country, the aforementioned switch from Carlism was coupled with nationalism and rapid industrialization. This process, driven by the production of capital goods (Diez-Medrano 1995), proved to be an enormous shock to the social structure of the region. Foremost among its consequences were a massive inflow of immigrants coming from other regions, especially Galicia, Extremadura, and Andalusia, and the emergence of the first working-class organizations. Over time, the rapid industrialization in the Basque Country generated a level of economic development well above the Spanish average. Preindustrial elites with rural lifestyles and values faced the erosive power of modernization and urbanization while extremely dynamic capitalist elites integrated thoroughly with the dominant Spanish elite. At the same time, a left-wing working class drawn from the underdeveloped regions of Spain maintained a popular culture that differed from that of the Basque.
The social base of Basque nationalism was principally composed of the traditional segments of society that had been hurt by the region's economic modernization and the liberal policies of the homogenizing Spanish state. These policies involved a range of political and cultural changes, including organizing the country into provinces following the French model and establishing a national educational system and linguistic uniformity policies built around Castilian history and language. Other social groups joined the Basque nationalist movement later on, but the founding principles remained the same. The racist, antiliberal, and military character of the nationalist myths, symbols, and narratives elaborated by the movement's founding intellectuals gave Basque nationalism a radical, separatist, anticapitalist, and traditional Catholic discourse. Not surprisingly, the majority of the industrial bourgeoisie distanced itself from the Basque nationalist phenomenon. Basque nationalism achieved its support in the old rural and urban strata. Nationalism expanded, but did not achieve an overwhelming acceptance in society at large. The final incarnation of Basque nationalism can be understood as a
128     ňavanced Industrial Democracies
Spain 129
reaction to both industrialization and immigration by the traditional groups linked to Carlism and Fuerism. Major aspects of this discourse still underpin the message of many Basque Nationalist Parry leaders and members, even though democratic principles, moderation, and pragmatism have been incorporated for the purpose of electoral competition.
The Catalan nationalist movement experienced an impressive political takeoff during the first decades of the twentieth century, reflected by the growing political and electoral presence of the Lliga Regionalista (Riquer 1977). Between 1914 and 1923, the expansion of Catalan nationalism achieved institutional recognition in the first experience of self-government worthy of that name: the Mancomunidad, led by Prat de la Riba. In fact, Catalonia would be the only region in which the institutionalization of political autonomy foreseen'in the constitution of the Second Republic was to be implemented, though very briefly (1932-34). So, what was the engine behind that process? Quite simply, the answer lies in the fact that the Catalan bourgeoisie, unlike the Basque, built upon a process of industrialization based on consumer goods (Diez Medrano 1995) The loss of the colonies in 1898 reduced the markets for their products and made it very risky to leave the regulation of their affairs to be managed exclusively from Madrid. They could not ally with Madrid as their counterparts did in the Basque country. So, they opted for an alternative national identity regarded as more appropriate for the promotion of their material interests.
As in the Basque Country, Catalonia's industrialization created a level of economic development far above the Spanish average. However, its social and political consequences were rather different. The economic transformation in Catalonia generated an increasingly modern social structure that gradually incorporated preindustrial segments of society into the capitalist context. Catalan nationalism was designed and directed by a dynamic, modern bourgeoisie that found the parties and representative mechanisms of a centralist national state unresponsive to Catalan attempts to influence the Spanish political system. This bourgeoisie led and encouraged the economic modernization process, but perceived a lack of any real representation of its interests in the policies of Madrid's government. Its political autonomy as well as the development of the Catalan language and culture were hindered by the centralist political and cultural homogenization imposed by the Spanish liberal state. This gave rise to a civic, moderate, negotiating, and nonsecessionist nationalism, which from the former political party, the Lliga, to the present-day Convergencia i Unio (CiU) has always sought to combine its demands for self-government with a significant participation in Spanish national politics. Most sectors of the Catalan society followed that pattern and, as a result, the ideological and organizational profile of Catalan nationalism became more and more plural, as shown by the convergence of several parties and organizations into
the Esquerra Republkana de Catalunya, a left-wing organization which, before the civil war (1936-39), would manage to surpass the Lliga in terms of political and electoral support.
Galicia presents a strong contrast with the Basque and Catalan cases. The absence of modernization, industrialization, and urbanization resulted in territorial disarticulation, a heavily agrarian social structure, and a level of development below the Spanish average. Galician nationalism was led by intellectuals, the local petit bourgeoisie, and the liberal professions. These groups felt both sidelined within the Spanish state and hindered by economic backwardness and lack of opportunities, which caused a massive migration toward more developed areas of Spain, Latin America, and eventually other European countries. Galician society was also negatively affected by the cultural and linguistic homogenization instituted by the Spanish state, which forced the Galician language out of the schools, reducing it to a language "of peasants," regarded as unacceptable for the purposes of national modernization. Nationalism flourished culturally and intellectually along traditional lines, evolving into moderate liberalism and then, after some internal fragmentation, into democratic republicanism. But its limited social support resulted in a precarious organizational and electoral base that, for a long time, left Galician nationalism a second-tier force within the Galician party system. Only after 1931, in the midst of the political mobilization generated by the Second Republic, would the Partido Galeguista become a representative force, though far less successful than its Basque or Catalan equivalents.5
When the Second Republic came into existence in April 1931, the demands for decentralization were strong enough to require constitutional recognition.6 Thus the Estado Integral as designed in the constitution can be understood as an intermediate solution between the traditional centralist polity, in a never-ending crisis, and a full federal solution, which was considered far too hazardous at the time for nearly everyone (Beramendi 1999,79-100). In this respect, the Estado Integral can be considered the direct precursor of the solutions adopted in 1978, as it was the first institutional attempt to accommodate the tense relationships between different national identities. It also shared the 1978 constitution's open and flexible design, its recognition of non-Spanish political identities, and its subnational self-government institutions. The collapse of the Republic and the civil war prevented the Estado Integral from being fully developed. However, it was on the basis of this precedent that (despite the efforts by the Franco regime) the demands for autonomy and decentralization would reemerge after 1975-
After 1978, the evolution of nationalist parties is characterized by two different experiences. From the very beginning they dominated the Basque and Catalan regional elections. In the Basque country, the Partido Nacionalista Vasco/Basque Nationalist
130      Advanced Industrial Democracies
Party (PNV) received over 30 percent of the vote in regional elections, while, in Catalonia, the CiU has averaged 40 percent. The Galician nationalist party had an almost insignificant record at the beginning of the period, but gained votes progressively, ending the 1990s with 25 percent of the vote. Table 4.1 presents the evolution of nationalist voting in the regional elections in the Basque Country, Catalonia, and Galicia.
In Catalonia, nationalist parties consistently rank first in regional elections while non-nationalist parties are systematically ahead in national elections. In the internal competition between nationalist parties, CiU's moderation and bargaining style representing center-right nationalism consistently triumphs over the radical secessionist nationalism of Esquerra Republicana de Catalunya (ERC), which attracts 8 percent of the vote in contrast to CiU's 40 percent. Under Jordi Pujol's leadership, the CiU has ruled the region since the first regional elections and, thus, has been able to organize a solid nationalism, identifying itself as the true defender of regional interests. In terms of center-periphery relations, Catalan nationalists have followed, according to their previous historical experiences and ideological background, a very pragmatic and moderate strategy in their relations with the central government and the parties supporting it (Socialist Party, Popular Party). Bargaining, collaboration, and exchange have dominated open conflict in the pattern of relations between the Catalan and the central government since 1978.
The Basque Nationalist Party has also been hegemonic since its beginning, establishing a solid organizational, electoral, and clientelist base in the early 1980s and controlling the regional institutions either alone or in coalition with Socialists or other nationalist parties. In this highly competitive pluralist electoral system, polarized between statewide parties (Popular Party and Socialist Party) and nationalist ones (PNV, Herri Batasuna, and Eusko Alkartasuna), each of the two groups occupies roughly 50 percent of the political space, with the regionalist parties slightly stronger. Whenever polarization is intensified, Euskadi suffers the threat of two potentially irreconcilable communities. The weakness of the right (Popular Party) until the 1990s and the decline in support for the Socialist Party helped the regionalist PNV to remain in government. Popular support for Euskadi ta Askatasuna's (ETA—literally, Basque Country in Freedom) activities has been waning: the regional vote for its political organization, Herri Batasuna, declined from 18 to 10 percent in the last elections. The PNV's political stance had been increasingly moderate since the 1930s due to its Christian Democratic agenda. However, in a recent attempt to end ETA violence, it changed course in 1998 and began to reject the EA framework, demanding the right to self-determination, and seeking the construction of a European Union (EU)-based Basque state separate from Spain and France. In this sense, the PNV reached an agreement with Herri Batasuna and also with ETA to stop terrorism as part of a broader joint political
^9
csi C\
w O
Oh W
&S
w
O 00
Oh
o
•S S
GO Z*
Oh ea
^ \C N
sO ©s
o\ \q 00
K  CO tN
O co
Ph. (J 6
Oh _. Oh Oh
GO
U Q U CJ
3 «? Pi c^ ^h co in
W O
GO
o w
t3 C5
Oh CO CO
Oh Oh Oh
\p N (O M
o
00
+
u
Oh Z
co
W Oh
a
so cn
fs, CO \q oq (N ts N IO 0\ -h
(N ~h ^h
Oh Oh
53 (J 0h <
Oh  Oh < P
r> co >-* o 00. © c>- o\ o sd \6 a\
*       C*fl ""H
Z < « CO CJ Oh < Oh  W  X Oh Oh < P
inHH/foooaN^t^cN^
so    ; .
H\    CO rH
^ rsi        ^H -H
►2 w -
P3  «  Oh  Oh  P <
CO
D
y Oh
w R "f
U U Oh
CO 1;
,C?s   CO   in K
cs] *-t .—i .—i rsi
.8
Z < S W to O y
Oh W X « Oh Oh P
r-. o o ^ -st< tj< co co »-i c\
Z £2 W to CJ Oh Oh  SB   B,   Oh  Oh <
h in co N o co co id o\ ^ ^ co ti*
ES a
a
Oh 53 W  Oh Oh P «!
in co
CS]    ^h    CO tjh
Spain 133
j-j m (J «
U
D O
M O
Oh PL,
owe
w
Dog
W
o
D U S
o s a
o
O 0<
O Oh
o
in o> in m
Oh
W  Q ro
■ o s j y „
a, % T W Oh
<     a, w
Q
O
CN  CN        i-h ,-h
LO 00 O 00
en a, to o a< an
■a
Q
o
^o H «
O W
o
5
3
«3 ^
o -a
1 (3 J2
•s i
H CO O
> (2 o ri. ' S & 1 J2 ■ •If *l
o;wo<M
a
o
S
2 5
3 u
!2 w
—
m g
S w Q ■A < 05 0
5 5 x d
c &0
■0 & w IBM
-3
=3
n -2 w
•S.SO
Oh CJ Oh
"3 S
U CJ
=3 ."2 s
c c 'u Si D o
'If" w 3 fro
5
&
•3 £ w
■ago
K « co
3 O Oh Oh CU
V 'J3
Si ~° 13
0 I ? ■
to — T" to CO Oh
0 -3 ,P 0
T3   g Oh T3
■I =s i
^ £ -2 ^
o y J 5
J a
« to
u 1 a,
.£ 1 IS c
.2 o
□ ■£]
'5 .2
■3 Z
agenda. This strategy failed badly and the agreement was subsequently rescinded due to the continuation of terrorist acts by ETA, which has been responsible for more than 800 murders since democracy was ushered in. As a result of this development, the level of polarization in the Basque Country has increased both between parties within the region and between the regional and the central governments, in what constitutes a much more unsettled pattern of center-periphery relations. Meanwhile, terrorism remains at the center of the political scene.
During the democratic transition, Galician nationalism was organized in small groups with radical Marxist-Leninist and secessionist agendas. These positions became increasingly moderate during the 1980s and 1990s. This process of moderation includes the acceptance of the constitution and the EA as the vehicle with which to pursue their political platforms and the participation in the regional parliament and local governments. The demand remains for greater self-government and defense of new Galician interests based in the modernization and urbanization processes of the 1980s and 1990s. The Galician Nationalist Block (BNG) has gradually incorporated a number of nationalist groups and parties that developed in the 1970s, forming a broad nationalist front spanning from extreme left wing to neo-liberal positions in which social-democratic positions seem to be the dominant ones. This catch-all strategy increased the percentage of the vote in regional elections to a high of 25 percent. The BNG earned a favorable position within the structure of political competition at the regional level by becoming the only Galician nationalist option, moderating its ideological position, leaving behind secessionist ideas, and accepting the EA, though occasionally demanding self-determination to please its supporters. The BNG also took advantage of the Galician Socialist Party crisis in the 1990s and increased its impact due to excellent organization, a catch-all discourse, and charismatic leadership. Still, the BNG had no impact on the center-periphery relations. Galicia has been ruled almost continuously by the conservative Popular Party (PP) since the beginning of the process of decentralization, with the exception of a brief period between 1987 and 1989. As a result, relations between Galicia and the central government had a very low profile until the PP took over the central government in 1996. They intensified after that date, but largely as an internal affair for the PP.
The Institutional Setting of 1978
The features of the 1978 constitution that bear most on the territorial cleavages described above are:7 (1) an openness and flexibility in the distribution of political powers, (2) a multidimensional asymmetry, and (3) an absence of institutionalized cooperation in center-periphery relations.
134      Advanced Industrial Democracies
Spain 135
Openness and Flexibility in the Distribution of Political Capacities
Perhaps the most important way die-Spanish constitution of 1978 organized the territorial distribution of power is the fact that the final text did not reflect any coherent model of state. Actors at the time were much more aware of what they did not want to be reflected in the constitution than anything else. The very notion of the Estado de las Autonomias is not even recorded in the constitutional text, but was coined later. Thus the constitutional design of 1978 remained open, indeterminate, and ambiguous. Like most other constitutions, the Spanish one was an incomplete contract with lacunae to be closed by normal politics (Dixit 1996; Rodden and Rose-Ackermann 1997,1527). But in this case, the degree of incompleteness was remarkably high by comparative standards. *
Rather than stating clearly who does what, the constitution lists those subjects and fields over which the state has exclusive jurisdiction (art. 149) and those subjects and fields over which regions (autonomous communities [AC]) are "entitled to assume exclusive jurisdiction" (art. 148). It does not say that the political capabilities and jurisdictions listed in that article are in fact allocated to the AC—thus, the initial level of power to be adopted by each AC is not precisely defined.8 Each of the seventeen AC had to write an Estatuto deAutonomia, a sort of regional basic law. The concrete contents of each of those Estatutos was an open issue, subject to political contention. They were not fixed a priori (art. 148.3). Second, the system contemplates some "catching-up" mechanisms, allowing the revision and expansion of the levels of decentralization initially achieved. So, for instance, the state has the capacity to decentralize those domains that happen to be susceptible to delegation (art. 150.2). Moreover, every AC has the right to expand its level of power five years after the approval of the Estatuto (art. 148.2). But the constitution also includes some clauses that could well allow the state to slow down, or even undo, part of the decentralization process on behalf of the general interest. The constitution is silent as to which path to pursue. This silence made the initial distribution of political capacities particularly significant, which leads us to the subject of asymmetry.
Multidimensional Asymmetry
Asymmetry implies an uneven, unequal distribution of resources or capacities within a given set of actors. One can argue that the Spanish federalization process has nurtured, from the very beginning, a highly asymmetrical political-system involving electoral asymmetry, asymmetry of access, and fiscal asymmetry.
Electoral asymmetry refers to the differences in the electoral success of political forces representing exclusively regional interests. It depicts differences in the presence
and political capacity of different nationalist parties in the national political institutions. The Spanish electoral system is one of proportional representation, built upon the D'Hondt Law. It was originally aimed to overrepresent rural provinces, with lower population density and a higher likelihood to opt for conservative forces. It also aimed to ensure the generation and stabilization of a two-party political system at the national level (Montero 1998,53-80). In both respects it must be regarded as successful. For instance, the Communists (Partido Communista de Espana/Spanish Communist Party [PCE]/Izquierda Unida/United Left [IU]) have never been able to obtain any representation in thirty-four out of the fifty-two provinces of Spain, despite being a fairly stable political force in die national parliament. The examples of the Centro Democrdtico y Social (CDS) and the notorious failure of the Partido Reformista Democrdtico (PRD) also support this argument.9 This electoral system has allowed for those nationalist parties with a high presence in their regions, first the PNV and CiU, and later on the Coalicion Canada (CC), to have an almost constant and growing presence in the central parliament. Between 1977 and 1996, the CiU jumped from 2 seats to 16, the PNV jumped from 2 to 6 seats, and the regionalist .parties in the Canary Islands and Galicia jumped from o seats to 4 and 2 seats, respectively. From 1993 to 1996 and from 1996 to 2000, the political capacity of certain nationalisms has increased remarkably. Due to a combination of several factors, which we will discuss later, region-based nationalist parties have become key factors in the achievement of any stable government when neither the Socialist Party nor the conservative party has been able to achieve an absolute majority. As Linz and Montero (1999) have pointed out recently, PNV, CC, and CiU share the double condition of having a strong coalition potential in die central parliament and holding office in their regions.
Asymmetry of power refers to the existence of substantial variations in the degrees of autonomy enjoyed by different provincial governments. Every regional government's Estatutos enjoyed the same level of constitutional protection, but their content and scope varied considerably. The Basque Country, Catalonia, Galicia, and Andalusia achieved much higher levels of autonomy than the rest of the regions. As we will see below, the consequences of this gap were to be far-reaching for both the political dynamics of the system and its capacity to deal with territorial conflicts.
Finally, fiscal asymmetry refers to the constitutional facilitation and further political development of a twofold system of fiscal federalism in Spain. Two fiscal regimes are at work: the general or common one, applied to fifteen out of seventeen AC, and a special one, only applied in the Basque Country and Navarra (Aja 1999,172-98; Monasterio et al. 1995). These two autonomous communities collect their own income, corporate, and value-added taxes and then contribute a fixed amount to the central government. The central government received over 95 percent of the revenues collected by the
136     Advanced Industrial Democracies
other fifteen AC,10 until the situation was partially amended, allowing the AC to collect and manage a greater share of the income tax (15 percent in 1994, 30 percent in 1996).
Together with the presence of three nationalist movements, each with varying levels of political mobilization and party organization, the presence of a violent terrorist organization claiming independence in the Basque Country must be regarded as a fourth and special source of asymmetry. Violence has changed the parameters of the political process, altered the structure of the bargaining process, and modified the relative position that one major actor, the PNV, would have otherwise had. In fact, the evolution of both the political profile and the strategies of this party illustrate our case clearly." The ETA's violence puts the PNV in the position of a gatekeeper, with one foot in each world. It generates both costs and constraints (the impossibility of full acceptance for the constitution, the possibility of blame from both sides) and further bargaining resources (a monopoly in the management of any possible solution, in the relations with Madrid, in the relations to the social network of ETA) which have allowed the Basque Country to achieve and maintain the highest levels of political autonomy.
Absence of Institutionalized Cooperation in Center-Periphery Relations
Traditional and well-established definitions of federalism present it as a mix of self-rule and shared-rule (Elazar 1987). After twenty years of permanent decentralization, Spain qualifies as "federal" in all respects related to self-rule. But, in terms of shared rule, Spain does not qualify as federalist at all. Shared rule implies a coordinated and institutionalized means for making all those decisions affecting both realms of power in such a way that both nationwide and regionally bounded interests are represented and protected. Yet, there is no institution performing such a function in the Spanish political system. There are only two regular forums that constitute arenas for cooperation: the fiscal and financial policy council and the regular meetings of all the national and regional ministers for education (Colomer 1998,49-50). However, they are limited to their specific policy areas. There is also a senate. However, while the constitution presents it as "the Chamber for territorial representation" (art. 69.1), neither the way it is elected nor the array of its functions is characteristic of a federal second chamber. The vast majority of its members are elected at the same time and on the same territorial basis as the members of the national parliament (arts. 69.2-69.4). Each province is allocated four seats, combining to provide 208 of the 256 seats. The remaining seats are appointed by the different AC regional parliaments. Each AC must appoint one senator by default plus one more per million inhabitants within its boundaries (art. 69.5). This adds up to between forty-four and forty-eight seats, depending on the year. As a
Spain
137
result of this method of election, the senate's composition resembles that of the national parliament. This, in turn, makes its real institutional influence very small and its capacity as a realm to represent territorial interests nonexistent. In the event of disagreement, the lower house always prevails (art. 90.2).
Some minor devices to promote cooperation and coordination among the different levels of government have been launched. But they are far from being a real alternative to the absence of a proper institution. Among all of the devices, two deserve attention: the Conferencias Sectoriales (sector conferences) and the Acuerdos de Cooperation entre el Estado y las Comunidades Autdnomas (cooperation agreements between the states and the AC). The former are basically policy-specific realms for the exchange of information and carry out consultative functions (Cruz Villalon 1990; STC 76/1983). The latter have been established as a joint effort in specific fields or projects (such as public works or joint programs; see Alberti Rovira 1996, 616-36). But neither can be considered an alternative form of institutionalized cooperation.
The absence of institutionalized cooperation was bound to bring two major consequences. First, actors engaged in an endless process of bilateral negotiations. Second, the constitutional court remained the final venue to resolve conflicts over policy domains, jurisdictions, and duties. The openly incomplete nature of the constitutional contract, together with the absence of an institution in which territorial contentions could be politically solved, forced the court to assume increasing importance in determining the evolution of the system.
Spain is, dius, a rather peculiar case. From the perspective of the political capacities of its constituent units we can hardly deny its federal and highly asymmetrical character. However, federalism in Spain is also incomplete. The dual character of its fiscal dimension and, most important, the lack of an arena for territorial representation and cooperation are the major sources of this incompleteness.
The Evolution and Performance of Federalism in Spain
Before exploring how Spain's unique federal structures affect its capacity to deal with territorial conflicts, a discussion of our criteria for evaluation is in order. Identifying criteria to evaluate federal systems is not an easy task, for every choice is susceptible to bias. The optimal choice for criteria would be to include all the significant dimensions along which there has been contention over federal policies. These would include what Simeon and Mintz (1982) call conflicts of taste, which regard identity-related issues, and conflicts of claim, which regard distributive issues. Since this project is concerned primarily with the capacity of federal systems to accommodate conflicts of taste, we chose two indicators related to these. Our first is institutional
138     Advanced Industrial Democratíes
Spain 139
stability. By looking at institutional stability, we aim to capture the extent to which Spanish federal institutions have managed to be self-enforced. We assume that the lower the levels of contention and contestation in the institutional design, the greater the success of Spanish federal institutions in promoting their own self-enforcement and the greater their success as an integrative device. A similar logic underlies our second criterion for evaluation, namely, the evolution of identities. As long as we observe a robust association between the development of the EA, the expansion of dual identities, and the shrinking of exclusive ones, the Spanish federal institutions may be regarded as successful (Stepan 1997)."
The Endogenous Character of Institutions: Patterns of Political Decentralization
The evolution of the Spanish Estado de las Autonomias took place in three major stages: 1978-83,1983-93, and 1993 until the present. All share a similar feature: institutions are at the root of the bulk of the political contention among major political actors.
1978-1983
The period from 1978 to 1983 is the most intense of the institutional game. After the initial steps of the EA, the first big inflection point came with the launching of the I Acuerdos Autonómkos (autonomous agreements), signed by both the Partido Socialista Obrero Espaňol (PSOE) and Union de Centra Democrático (UCD) in 1981. These were the very first attempts to control the emerging dynamics of the system, since the early stages of Spanish democracy could not afford any avoidable instability. The failed coup d'etat in February 1981 provides us with a good indicator of the complexity of the environment at the time. The objectives pursued by the two major parties are easily traceable through the contents of the agreements (MAP-1982): the clarification and generalization of the institutional designs of all Spain's autonomous communities, the adoption of a common position regarding the fiscal arrangements to be implemented, and, last, but certainly not least, the creation of a law aimed at harmonizing the decentralization process, namely, through the Ley Organica de Armonización del Proceso Autonómico (LOAPA). This law was the first big step in what we interpret below as a permanent tension between harmonization arid asymmetries. It created a conflict about the real meaning of the process of decentralization. As a response to this conflict, the constitutional court formulated two far-reaching rulings on the subject of autonomy. In its ruling 37/1981, the court legitimated differences in the way citizens are treated by the different realms of power. Civil rights are regarded as the only field in which absolute equality must be guaranteed. Subsequently, in its ruling 73/1983, it overrode most of the LOAPA and established that the Estatutos de Autonómia enjoyed a
constitutional status well above ordinary legislation. Autonomy was constitutionally protected thereafter (Aja 1999,65). The I Acuerdos Autonomicos ended the first phase of the federalization and became embodied in the newly elected institutions. Following the elections on May 8,1983, thirteen new regions (AC) were finally created.
1983-1993
These years were marked by the expansion and consolidation of the Estado de las Autonomias, driven mainly by the three consecutive socialist governments of Felipe Gonzalez, who benefited from an absolute majority in the parliament. Table 4.2 presents a general picture of the evolution of the transfer of powers from the central government to the AC between 1978 and 1997.
As Table 4.2 illustrates, the process of decentralization developed rapidly between 1983 and 1986. Most of the transfer of powers involved the construction of regional political institutions and the transfer of some elementary services formerly provided by the central government. After 1986, the process slows down significantly. In the particular case of the Basque Country, this is very much related to the issue of violence. For the rest of the AC, it just indicates an intensification of the institutional conflict between several levels of government. These involved complaints about the constitutionality of central and state governments lodged by both administrations against each other and conflicts of jurisdiction. There are two kinds of such conflicts: the positive (when two administrations claim control over any given policy field) and the negative (when both levels of governments claim not to have the responsibility to provide some particular good or service). As mentioned in the first section of this essay, the constitutional court had to assume the key political role of resolving many of the open aspects of Spanish federalism during these years. The levels of tension over who had the right to do what, as measured by the evolution of the positive conflicts of jurisdictions, rose rapidly during the mid-1980s. The Socialists' strong parliamentary majority during this period facilitated bringing issues to the constitutional court, if bilateral negotiations failed. If the disagreement was about legislation approved by either level of government and the other disagreed, the expected outcome was an appeal regarding the constitutionality of the law. Alternatively, if the conflict regarded the sharing of powers, conflicts of jurisdiction were expected. In this case, most of the conflicts had their origins in the definition of the meaning of "basic" jurisdictional areas to be guaranteed by the central government in the fields of concurrent legislation. These involved some of the most important policy fields, such as education, environment, and health. Not surprisingly, the high number of conflicts between 1983 and 1991 introduced significant delays in the development of such programs and, in turn, in the overall pace of the process of decentralization (Table 4.2).
Spain 141
^   ^   ^   ^   ^   ^ 00 W00O\MO\O\NC0\O
c^i oj m O vo 00 m      in mcoo-HO\r>.oo-\m
1
^ 2 !ľ: ^ 0 ^ ^      <w moomoootNOLno
Os vo VO O O C** VO
O (NI  Tfl  o  (N cn
O  VO 1—   N  'í [s O
O vo o o m ■—1 ro
N  CO  N  N  N  \£) in
CO  fO  CO  CO  O  t-O  rf r-<
tN m m      (N (N m
<n  t*N  (N  CN  O —<  O r-l
N M h S o\ >£l O CN ■—< cN        h (N
tS   fN   CO   (N   O   O O
OO0OO--hooo
■^VOOVOOCNmocN
00   (N t—1   fo   0"\   O O
»—1 m r-4 •—1 cn <n
(N        (N  m  N  (N  (N  tn h
mo^t}*Tŕrocnoo
3
2 E
T3
52   *š< 10
•2 >s -s*
; « u S B -s
5ž
-
a
!l
O
6 B 3 §
.n 0
p 1 G
a Si
I 8 S.| b u*
II
In this context, Spain's entry into the European Economic Community (EEC) (1986) became one more source of political conflict. Before Spain's entrance into the EEC, the central government was the only acknowledged realm of power with full political capacities. Nonetheless, many of the choices made during the integration process affected both the interests and jurisdictions of all the AC. This change in the external political environment created the problem of how the AC could represent their interests before the EU. As long as representation was not possible, nationalist parties would perceive Europeanization as a hidden means of centralization within a constantly open institutional agenda.
At the end of this period, Socialists (PSOE) and Conservatives (PP) shared the view that the level of instability, the increasing demands for further decentralization, and the practice of bilateral exchanges demanded some sort of coordinated reaction by the two countrywide parties. As a result, the IIAcuerdos Autonómicos were adopted.
Retrospectively, these pacts can also be seen as a temporary commitment by these two national parties not to use issues concerning the institutional design of the country as a partisan tool. The major effects of this agreement were twofold: on the one hand, they meant a great step forward in the overall levels of decentralization. The reform of the Estatutos de las Autonómia gave the AC with slower access to autonomy (art. 143) the potential to receive most of the transfers enjoyed by those that achieved autonomy earlier and faster, namely, the Basque Country, Catalonia, Galicia, and Andalusia. This explains the increase in the transfer of power over the past few years represented in Table 4.2. By 1994, the process was almost concluded except for health transfers. Somewhat paradoxically, the center-rooted parties provided, in the early 1990s, an average increase in the levels of decentralization. As we will see in the following section, this very fact contradicts some conventional expectations about the processes of decentralization. But implicit in this expansion was the pursuit of the horizontal equalization of power. When the nationalist parties reacted against this harmonizing change, the system polarized further.
1993-1999
The results of the general elections in 1993 opened a new stage in the development of Spanish federalism. The structure of political competition at the central level prevented both the Socialists (1993) and the Conservatives (1996) from achieving enough seats to rule on their own. This change implied significant shifts in the relationship between the central and the nationalist state governments.13 The political case against higher levels of harmonization was now much stronger. Therefore, important policy changes were to be expected. The evolution of Spanish fiscal arrangements illustrates this point well. In 1994, tax policy was finally reformed to allow AC to collect and
142      Advanced Industrial Democracies
Spain 143
manage 15 percent of the general income taxes. The Conservatives at the time accused the socialist government of "selling" Spain. After 1996, when the conservative government needed nationalist support, the reform was expanded to include 30 percent of general income taxes. It is hard to think of any of these reforms being implemented under the political conditions existing prior to 1993.14
However, harmonization is still perceived as a risk by nationalist parties. They have repeatedly argued that blanket decentralization is no solution. In addition, the issue of terrorist violence and the aforementioned complex position of the PNV (who decided to link any step toward the achievement of peace with deep constitutional changes) increased the levels of polarization. During these years the number of appeals before the constitutional court increased. In fact, it was during this period that some of the major ideas for the substantial reform of the EA were proposed.
So, even though "self-determination," "sovereignty," and even independence have been present in the discourses of nationalist parties throughout the period, it is only recently that the BNG, PNV and CiU have adopted a common platform in order to pursue a constitutional reform toward some sort of indeterminate "confederal" arrangements.15 Leaving aside the vagueness of their statements, these three parties have recently agreed upon a set of minima for the constitutional reform. These can be summarized as follows: (1) explicit acknowledgment of the plurinational character of Spain, with the only subjects of sovereignty to be the three historical AC and the rest of Spain; (2) the transformation of the senate into an asymmetrical, territorial chamber; (3) the establishment of the special fiscal regime currently enjoyed by the Basque Country and Navarra in Galicia and Catalonia; (4) a reform of the constitutional court, incorporating both regional representation and input on the appointment of its members; (5) the achievement of full external representation before the EU; and finally (6) the decentralization of the social security system.
These positions provoked two different responses: while the Conservatives (PP) sought to defend the constitution in its current state, the Socialists (PSOE) sought to reform the constitution along a federalist path.16 The contents of the PSOE proposal shared very little with the nationalist proposals. The only real points of agreement concerned the representation of the AC in the EU. But even these were dubious, given their fundamental disagreement as to who the major collective actors should be. The nationalists argued that not all the AC should be entitled to EU representation. The Socialists argued that each of the seventeen existing AC must have representation. Moreover, the rest of the Socialists' proposals had even less in common with the nationalists' ideas. They argued that solidarity, as opposed to asymmetry, should guide a general reform of fiscal federalism in Spain. Cooperation and intergovernmental coordination were the principles needed for a stable development of the system. The reform
of the senate should be as symmetrical as possible. At first sight, then, the prospects for an agreement between the federalist proposals of the PSOE and the nationalist claims were rather low.
2000-2003
In March 2000 the PP won the national election by a margin that allowed it to rule alone, without the support of the nationalist parties (CiU, PNV). Meanwhile, the reactivation of terrorism after the 1998-99 truce and the growing disagreements with the Basque Nationalist Party as to how to pursue peace in the Basque Country, exacerbated debates about both the EA and the path toward its reform. As a result, positions became more extreme. While the Basque Nationalists (PNV and Eusko Alkartasuna) openly advocated a radically new institutional framework that would situate the Basque Country outside the EA framework, albeit still "associated" with the Crown,17 the PP declared the constitution untouchable and warned against any discourse proposing its reform, regardless of its content. After 2000, such discourse became the axis of a neo-centrist approach to institutional reform. In this framework, the federalist solution proposed by the PSOE was considered a dangerous game that objectively facilitated the ultimate goals of those aiming to destroy the EA. The neo-centralist approach to the EA has not stopped the plea for a reform of the Estatuto de Autonómia in Catalonia. In fact, both the Catalan nationalists (CiU) and the Catalan Socialists (PSC-PSOE) have put forward proposals that advocate direct representation at the EU, adoption of a fiscal regime similar to the one in the Basque Country and Navarra, and recognition of Catalonia as a nation within a plurinational state.18 In brief, due to the neo-centralist twist of the PP in the last legislative term, the disagreements regarding the EA have become more bitter and more visible.
Identities
As mentioned earlier in this section, we take identities as one more indicator of the evolution of the tension between different nationalisms in Spain (Stepan 2001,315—62). Our empirical strategy to cope with them is twofold: first we map their change over time using several surveys collected by the CiS in 1979,1985-86, and 1996. For each of the years we map identities using the standard scale of the ordered answers to the question: "How do you consider yourself? Exclusively Spanish, more Spanish than from your AC, as Spanish as from your AC, more from your AC than Spanish, and exclusively from your AC." It must be noted from the very beginning that the data for 1979 are not stricdy comparable, since they are answers to the question: "When you are abroad, how do you consider yourself?" In our view, the most useful information from the 1979 survey concerns exclusive AC identities. The underlying assumption is that only those
144     Advanced Industrial Democracies
people with extremely strong regional identities would declare themselves to be Basque, Catalan, or Galician if approached, for instance, in Picadilly Circus.
As a second means of assessing identities, we carry out a binomial logistic regression model of the determinants of political identities in 1996, checking the extent to which identities are actually driven by institutional and discursive factors (namely, the effects of the presence of nationalist parties) when controlling for education and labor market position. The dependent variable takes the value 1 if the individual shows dual identities in either order and 0 if she or he presents any of the exclusive identities. Our major concern is with the degree of development of dual identities as opposed to either of the two exclusive ones.
Table 4.3 shows that the levels of non-Spanish exclusive identities have remained fairly constant in Catalonia and Galicia. We observe also a significant drop in the Basque Country (from 29 percent in 1985 to 20 percent in 1996). This appears to be related to the growing rejection of terrorist violence among the nationalists. It is also related to the growing electoral presence of the Socialists and, more recently, the Conservatives (PP) in the Basque area. The relative success of nationalist mobilization strategies can also be traced through the changes in the internal distribution of dual identities: in Galicia and Catalonia there is a significant shift from those who see themselves as both Spanish and either Catalan or Galician toward a more intense identification with the regional nationality associated with their AC. In the Basque Country, however, the distribution remains practically unchanged.
The rest of the table shows that the development of the Estado de las Autonomias has also had important consequences for the evolution of territorial identities in the rest of Spain. In spite of their internal changes, the levels of dual identity have remained significantly high in Galicia, the Basque Country, and Catalonia.19 Moreover, in the rest of the country, the EA has diminished the weight of exclusively Spanish identities. Even in Madrid and La Rioja, two of the AC where there was no competing reference in terms of political identities, the dual ones represent the biggest share of the population in 1996. In the case of Valencia, dual identities seem to be quite consistent, despite the political expansion of regionalist/nationalist parties. On the contrary, in Aragon, we observe an increase in the proportion of people who see themselves more identified with their AC than with Spain (from 11 to 17 percent between 1985 and 1996). Finally, Andalusia and Canarias constitute two interesting cases. Andalusia experienced a centripetal movement from a more strongly AC identity toward an identity equally Spanish and Andalusian. In Canarias, the centripetal shift comes from both sides. Strong preexisting exclusive identities have given way to an increase in the proportion of groups showing dual identities (both equal and skewed toward the AC). t
Spain
145
o '—i K
o o m o —1 o o
CN  01  CN  CO >—< >—' ■—I
IR: N ffi VO Ov        lO O
cn^co^MS^in^co
in 00 o ^ t<
.-H    i-H   00   ^ O"*
(N o o ^ o in h h h rj
^HTjHininincdo^-^
i-H CO
O  O  O  CJ\  O  O O   cd cd
000^000 ca ca
OOO^OOO   cd cd
p o o o
o\ n ^ ri K
o p in 06
oppiqppp cd cd f4 tfi d (o rs ifl ps c
cd   cd O   cd cd cn'oid-^CCOCC
cd IN C cn
cd cd O cd cd K   C   C 00   C c
. in o        . ro
Ocdr^rMcdcd^cdcd
cncoNincej^occ
c
3
0 U
cd   « %
1 §
: cd
•>-> *—i   <st   G   y RJ^=!
.5
"cd C
<u <l>
3 'C
o £
■si
.■ti c S .2
•: ca
£ e 53
£ § 5
|Q 2
*r-t o\ a
« °*
cr\ >~ .5 x>
-«•S & §
sa'? y
CO
J1-5.S
in |S J-co „TJm
2-Sin^
£2 a-? S RI b S
2SSg
u -a ji s§ « u o =
§sa. 1
146     Advanced Industrial Democracies
Spain 147
Are these results really linked to the development of nationalist parties and the Estado de las Autonomias? Could they not result from other factors? In order to find out, we have estimated a simple binomial logistic regression model whose results are presented in Table 4.4. _________
According to the model, three variables have a significant effect on the development of dual identities. The scale of nationalism20 and the presence/absence of a strong nationalist party in the region have a strong, negative impact on the development of dual identities. The variable tracking the degree of development of the process of decentralization in each AC in 1996 shows the expected results. The implications of the model reinforce the interpretation we drew from Table 4.3. Controlling for nationalism, education, and labor market position, the EA has a net positive, though not very strong, effect on the development of dual identities.
In conclusion, it may well be said that Spanish federalism has been relatively successful in maintaining and nurturing dual, compatible identities. However, nationalist parties have also succeeded in skewing the distributions in Catalonia and Galicia in favor of positions closer to the AC, while essentially holding constant the levels of
Table 4.4 Modeling Dual Identities			
Variable Exp (B)	B	S.E.	Wald
LEVEL 1.0252	.0249*	.0060	17.4636
NATION .2601	-1.3467*	.2748	24.0115
NATSC .8451	-.1683*	.0465	13.0988
UNEMPL 1.1330	.1249	.2995	.1738
EDUCAT 1.0782	.0753	.1051	.5123
Constant	.7591	.8667	.7671
N= 4322 R-squared: 0.439
Source: CIS (Center for Sociological Research) survey no. 2228 on National and Regional Consciousness 1996.
Notes: Binomial logistic regression model. The variables included in the model have been defined as follows. LEVEL: continuous variable that represents the degree of the development of the process of decentralization in a particular AC in 1996. Data from the Ministry of Public Administration. NATSC: is a scale of nationalism that goes from 0, self-perception as fully non-nationalist, to 10, self-perception as fully nationalist. NATION: dummy variable capturing the presence(l)/absence(0) of strong nationalist parties in each of the seventeen AC. UNEMPL: dummy variable capturing the labor market position of the interviewee (1: unemployed; 0: at work). EDUCAT: continuous variable that measures the number of years in education of the interviewee. *Significant at p < 0.001 level.
exclusive identities. Finally, a quick look at these levels in Catalonia and in the Basque Country shows that neither are a priori high enough to allow any credible push toward secession at present.
Nonetheless, our model also shows that identities are shaped by the strategies of actors, rather than the other way around. This suggests that this threat may become more credible as long as hegemonic nationalist parties push their discourses and strategies in that direction. This very possibility is, in itself, an extremely powerful political source of instability (backed in the Basque Country by the activities of ETA) which influences to a great extent the present and the future of Spanish politics.
Institutional Choices and Unintended Consequences: Federalism and the Management of Territorial Conflicts in Spain
How can the empirical evidence provided in the previous sections shed some light on the more general questions addressed by this project?
In line with other contributions to the project, it must be first noted that the Spanish experience points to the interactive relationship between the design and performance of federal institutions and the specific articulation of historical, economic, and identity-related differences within Spain. The institutional design of 1978 is not independent of the existing structure of territorial cleavages and the subsequent pattern of territorial conflicts was not independent of the consequences of the EA.
The EA's ability to manage territorial conflicts can be assessed by combining two different perspectives. Relative to experiences in other countries and to Spanish history, the capacity of the EA to integrate different clusters of economic and identity-related differences has been impressive. From this perspective, the EA must be regarded as a success. Conflicts about linguistic rights, the control of policy domains, health care funding, the design of curricula, or foreign policy decision making, to mention only a few, will continue to exist as long as there are several layers of government. The major success of EA lies in that, however contested at the ideological/discursive level, all actors (but ETA) act within the rules of the system. Additionally, the EA itself has managed to modify the structure and depth of identity-related differences, by reducing mutually exclusive differences and nurturing dual ones in Catalonia, Galicia, and, to a lesser extent, the Basque Country. The practice of decentralization had important effects in the rest of Spain as well, where dual identities have increased in relation to Spanish exclusive ones.
148      Advanced Industrial Democracies
However, while regional and national incumbents, and all other salient political actors, follow the system's rules, they have been trying to adjust the scope and nature of decentralization from 1978 on. Spain presents us with an open-ended process in which a general picture of success hides complex internal dynamics, causally linked to the intended and unintended consequences of the original institutional design.
It is fairly well known that the conditions under which the EA was designed were not optimal. In the context of the transition to democracy from a heavily centralized dictatorship, any maneuvering room was strictly limited. Retrospectively speaking, the sources of success can be found in the combination of a twofold strategy: the acknowledgment and satisfaction of demands affecting mainly symbolic and identity-related issues in the constitution itself,21 and the deliberate omission of major issues on which full agreement was impossible. In this latter respect, to quote the current president of the constitutional court, important aspects of the central state's boundaries were de facto "deinstitutionalized" from the very beginning (Cruz-Villal6n 1990, 71-103). In turn, in order to attract nationalist parties to the constitutional consensus, asymmetry had to be implicidy or explicitly present in the constitutional text. Hence, the distinctions analyzed abovewere introduced to both reflect the existing differences and facilitate their future reinforcement beyond the constitutional arena.
However, there was a price to be paid for such a strategy. The EA lacks institutions of shared rule (as explained above), and conflicts about asymmetry have been ongoing since 1978. Given the initial conditions of openness, asymmetry, and lack of institutionalized cooperation, bilateral exchanges and pressures became the engine of the system. This dynamic has generated two factors that have increased the tension between harmonization and asymmetry and kept the system far from the perfect institutional stability described above. The first is an expansion in the number of regionalist and cryptonationalist parties in AC where they had previously had virtually no political presence. Table 4.5 presents the evolution of the voting share of those parties in the regional elections for all the AC except for Galicia, the Basque Country, Navarra, and Catalonia.
These parties foster the tension between harmonization and asymmetry because they pressure incumbents in their ACs to demand that their local region's autonomous powers match those of the original nationalist ACs.
The second factor increasing the tension between harmonization and asymmetry is the fact that the process of decentralization has come closer and closer to its constitutional limits. Paradoxically, as the design of EA becomes more complete, the trade-offs become harder. As the system approaches its final form and the case for higher levels of harmonization is supported both by statewide and new regionalist parties, it loses
Spain 149
j
Table 4.5 Share of Regionalist and Nationalist Voting in Regional Elections of the AC plus
the Canary Islands
	1983	1987	1991	1995	1999
Aragon	19.8	26.0	26.4	24.8	24.38
Baleares	13.92	9.19	3.98	11.02	10.97
Asturias	0.7	1.3	3.1	3.1	9.76
Canarias	23.1	29.0	38.8	42.4	42.1
Cantabria	8.6	13.6	41.3	33.1	16.6
Castüla-La Mancha	0.0	0.3	0.7	0.8	0.8
CastfflaYleon	2.8	2.4	2.5	5.1	6.7
Extremadura	10.4	7.0	4.6	5.3	3.9
Madrid	0.4	0.1	0.9	0.9	-
Murcia	2.7	3.4	3.0	1.3	2.7
Valencia		9.14	10.37	7.01	8.9
La Rioja	7.5	6.5	5.5	7.6	7.02
Sources: Ofiate et al. 1999 and Ministry of Internal Affairs. Note Figures are percentages.
flexibility and thus pushes nationalist parties to move toward a different strategy. As illustrated by the empirical evidence provided in the third section of this essay, these actors have turned from arguments about shares within the system to questions about the overall structure of the system. System stability is still elusive.
This said, the EA has, by historical and comparative standards, worked rather well as an institutional device to manage the coexistence of multiple identities. It has also helped Spain complete the transition to democracy.
On a more theoretical level, the Spanish experience raises some interesting questions about the interaction between federalism and nationalism (Linz 1997; Watts 1998). On the one hand, the experience of the EA shows that federalism is the most successful institutional device in integrating several nationalisms within the same polity. Its advantages in relation to its alternatives, especially in relation to a centralized/unitary system, are illustrated by the long-term historical analysis of the Spanish experience. On the other hand, the evolution of EA also shows how nationalism introduces an additional, entirely independent dimension into the relationships between several levels of government: asymmetry (and its opposite, harmonization). While decentralization concerns the sharing of powers between the center and the regions, asymmetry is a function of nationalism, and applies to the horizontal differences between regions in terms of their power and capabilities. In this sense, our chapter shows that when federalism and nationalism are combined, the political system will never be at equilibrium.