The Politics of Decentralization in a Centralized Party System: The Case of Democratic SpainAuthor(s): Alfred P. MonteroSource: Comparative Politics, Vol. 38, No. 1 (Oct., 2005), pp. 63-82Published by: Ph.D. Program in Political Science of the City University of New YorkStable URL: http://www.jstor.org/stable/20072913Accessed: 09/03/2009 16:18

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=phd.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with thescholarly community to preserve their work and the materials they rely upon, and to build a common research platform thatpromotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Ph.D. Program in Political Science of the City University of New York is collaborating with JSTOR to digitize,preserve and extend access to Comparative Politics.

http://www.jstor.org

The Politics of Decentralization in a Centralized Party

System

The Case of Democratic Spain

Alfred P. Montero

William H. Riker argued that the most distinctive aspect of federalism is "the consti

tutionally assured potential for local governments to disrupt."1 Taken too far, inter

governmental conflicts could make policymaking unresponsive and inefficient.

Riker argued that such disharmony, when not effectively eliminated by the (^cen tralization of the administrative structure of the state, can be mediated by the central

ization of the political party system. His argument focused on two dimensions of the

party system: the distribution of partisan loyalties between central and subnational

government and the tendency of locally elected politicians to articulate local inter

ests. According to Riker, the first dimension of disharmony is more likely to be con

trolled where the same party or alliance controls national government and most or all

of the subnational units. This argument assumes that national parties are disciplined around partisan brokers who reside at the national level. These conditions are also

crucial in reducing the tendency of subnational politicians to articulate local interests

to the detriment of their affiliations. Therefore, centralized, disciplined parties with

high degrees of concordance in partisan loyalties across the national-subnational

divide are able to limit intergovernmental conflict.

Riker's theory directly influenced more recent comparative studies of decentral

ization that follow his notion that party system structures are the key independent variables in explaining the nature of intergovernmental relations. Most recently, the

seminal work by Eliza Willis, Christopher Garman, and Stephan Haggard focuses on

the location of party brokers, party leaders who shape the careers of politicians on

the national and subnational levels through control over nominations and placement on electoral lists.2 In party systems in which these leaders are located in subnational

government, the state is more likely to be decentralized in its policy responsibilities,

revenue-raising powers, and expenditures since these resources will flow to the cen

ters of power that shape the political interests of national policymakers.3 There is an

ongoing tension between national and subnational politicians in which national par

tisans, especially those in government, prefer to limit the autonomy of subnational

governments in order to maximize their own capacity to distribute resources across a

country based on the criteria of need and expected political payoffs.4 Closed elec

toral lists empower party leaders who control not only nomination of candidates but

63

Comparative Politics October 2005

also the probability of election through list placement.5 In short, the process of

decentralization is shaped by bargaining among politicians, and "the structure of

political parties provides the medium through which such bargaining takes place." 6

Democratic Spain presents a challenge to arguments about party organization since it mixes a centralized political party system with highly disciplined national

organizations and a decentralized and continually decentralizing state. Spain's transi

tion to democracy during the 1970s was a dual transition in the sense that it was a

regime transition and a state transition, the latter involving the replacement of a cen

tralized bureaucracy with a federal structure. Yet unlike the democratic transition

that was based on consensual pacts among political elites and transparent institu

tions, Spain's regional autonomy system (the state of the autonomies) was born in

the context of intense intergovernmental conflicts and ambiguous institutions requir

ing continued interpretation by legislatures, parties, the judiciary, and even the elec

torate.7 These intergovernmental conflicts continued even during periods when the

same party ruled the majority of regions and maintained an absolute majority in the

governing lower house of parliament (the chamber of deputies or congreso).

Intergovernmental conflicts continued even as the party brokers who managed the

national legislative process retained their capacity to shape the political careers of

their copartisans. Why does Spain's experience diverge so apparently from the pre dictions of the Rikerian framework? To answer this question, arguments about the

power of party organization have to be questioned fundamentally.

Challenging Party Organization Arguments

When applied to a broader set of experiences, party organization arguments fail to

explain some aspects of change in the degree, pace, and structure of decentraliza

tion, especially in cases of increased decentralization with centralized party systems.

First, change in party system structures has been too slow or too shallow to explain

significant shifts in intergovernmental policy responsibilities and resources over time

and across different cases.8 Second, these explanations rely on a one-dimensional

understanding of interests as partisan and not territorial.9 Governors and mayors have interests emanating from the offices they hold and not just the parties they rep resent. These interests are most obvious when regional presidents, such as those in

Catalonia and the Basque Country, claim to represent "nations." Yet territorial inter

ests are also apparent in nonnationalist regions such as Asturias, whose leadership has sought to reverse the area's industrial decline under different party governments

regionally and nationally.10 Party system structures are not irrelevant, but they are

insufficient to explain decentralization.

Party organization arguments focus too much on the mechanisms of shaping decentralization in the legislative arena, and they ignore other foci of political con

64

Alfred P Montero

flict that determine decentralization. These limitations are apparent in countries that

initiated their decentralization processes at the same time as their democratic transi

tions. Simultaneous, dual transitions produced two fundamental tensions that explain the divergence in party system structure and decentralization outcomes: the simul

taneity of democratization and decentralization produced countervailing logics in

favor of both party system concentration and decentralization of the state, and

democratization provided new political spaces for subnational interests to demand

reforms in the state structure after transitions (and their founding constitutions) were

initiated and established. These conditions are common in third wave democracies

where decentralization of the state has continued even as party systems struggle toward greater centralization under the exigencies of implementing reform.11

Spain demonstrates that a political order with a decentralizing state and central

ized and disciplined national parties will still allow subnational interests to influence

public policy either through the articulation of policy differences within the national

parties or through intergovernmental conflicts in which subnational executives repre sent territorial interests vis-?-vis the national government. Where the discipline of

national parties is particularly robust, and the capacity of subnational partisans to

represent their region's interests in parliament is weak, the weight of representing subnational interests will shift disproportionately to intergovernmental conflicts and

away from intraparty conflicts.

Riker's central claim that the political structure of government is more influential

than administrative structure in shaping the degree of decentralization needs to be

reevaluated. The continued decentralization of the administrative structure can have a dynamic at least partially independent from the political structure. This dynamic is

based on intergovernmental conflicts between subnational executives and national

governments, that is, in an arena of political conflict and cooperation not wholly within the purview of parties and legislatures but legitimated and necessitated by the

political logic of a simultaneously decentralizing democracy. Spain illustrates the most prominent ways in which the Rikerian framework fails

to explain the degree, pace, and type of decentralization. The logic governing the

consolidation of national democratic institutions diverged from the logic governing the recomposition of the Spanish state as a federal administration. In the hope of

avoiding the sectarian and ideological conflicts that doomed the Spanish Republic

(1931-36) and led to the Civil War (1936-39), most of the institutional choices dur

ing the regime transition underscored the need for government stability.12 The elec

toral laws devised in December 1976 favored the formation of a small number of

disciplined national parties in full control of closed electoral lists for each of the

fifty electoral districts.13 This system quickly became institutionalized around stable

patterns of party competition including three to four statewide parties (the Union of the Democratic Center, UCD; the Socialists, PSOE; the Popular Party, PP; and the

United Left, IU) and prominent regionalist/nationalist parties, the most important of

65

Comparative Politics October 2005

which were the Basque Nationalist Party (PNV) and the alliance of the two main

Catalanist parties (Convergencia i Uni?). Party identities remained the strongest pre dictors of the electorate's choices during most of the democratic period and, with the

exception of the Basque Country and Catalonia, the two centrist parties following the implosion of the UCD in 1983, PSOE on the left and PP on the right, controlled

the regions and governed the national congreso. The parties themselves remained

highly disciplined and mostly centralized organizations.14 In contrast to the centralization of political society, the regime transition set the

stage for the decentralization of the state as an essential component of the consolida

tion of Spanish democracy. The elites that negotiated the transition and the new

democratic constitution created seventeen "autonomous" regions with designated resources and policy responsibilities. Spain is therefore a useful test case for the

hypothesis that decentralization of the administrative structure of the state proceeds

according to a logic at least partly divorced from the institutions governing parlia

mentary behavior and the party system.

Party Brokers and the Dual Party System

Increasing decentralization in Spain is paradoxical for party organization arguments

precisely because party brokers preside on the national level, the electoral system is

closed-list, and the national legislative process is dominated by the interests of par

liamentary parties. When compared to party systems elsewhere in Europe, the

Spanish parties score at the high end of indicators of internal discipline.15 The

national leaders of the UCD, PSOE, and PP all rely on formal powers reinforced by

party statutes and the regulations governing the legislative process in the congreso that give these individuals singular control over their organizations.16 The UCD is

the exception that proves the rule. The UCD was poorly institutionalized as a party, and its fractiousness led to its demise soon after the 1982 elections.17

The closed-list system provides party leaders with extraordinary control over the

election of candidates to parliament. Proponents of party organization arguments hold

that this institution alone, ceteris paribus, is a factor favoring centralization if national

party leaders control the lists.18 In Spain this assumption holds de jure, but the subna

tional party organizations exert some influence defacto over the composition of elec

toral lists. Formally, both national and regional party leaders design the electoral lists, but the criteria they use in most cases are not subject to a specific outline set down in

party regulations. From extensive interviews of the general and organization secretaries

of the major statewide and nationalist parties in seven regions and interviews of national

party bureaucrats, I surmise that the criteria used to advance politicians' careers differ

primarily on whether the list is for the congreso or the regional assembly. Each regional

66

Alfred P Montero

party office in all of the statewide parties and provincial organizations of the major nationalist ones in Galicia, the Basque Country, and Catalonia nominate candidates for

the regional lists, and they have significant input in the composition of lists in the elec

toral districts in their region for the congreso. Regional party officers may select local

notables for positions following the national notables on lists for the congreso. As Table

1 shows, diverse criteria, including attention to regional concerns, invariably emerge as

part of the process, which indicates that the subnational offices are not mere franchises

of the national party office. But regional concerns shape candidate selection more sys

temically in the composition of regional lists than in the composition of lists for the

congreso, reflected in the more consistent tendency to value experience in subnational

office and service to the party in the composition of regional lists. Because national

party brokers must approve all recruitment choices, subnational party offices select can

didates for the congreso's list that will pass the national electoral committee's scrutiny. The judgment of the national committee, in turn, is governed by the loyalty and service

that the candidate has demonstrated to the party and not necessarily to the region. The

nationalist parties are only partial exceptions, but they too favor national party service

as the key determinant of placement on the national list. While the national electoral

committee has final say over the regional lists, these organizations eschew micro

management of the subnational lists, so most are accepted with few changes.

Table 1 Candidate Recruitment Criteria for Seven Regions

Region National List Regional List

Andalusia Service to national party

Asturias Experience in regional/local office; service to national party

The Basque Service to national party Country (PSOE/PP) or nationalist party

(PNV)

Catalonia Professional non-politician of social

standing; service to national party (PSOE/PP) or nationalist party and local experience (CiU)

Galicia Service to the national party (PSOE/PP); sectoral representatives (labor, business, etc.); experience in local office (PP); party notables

(BNG)

Madrid Service to national party

Valencia Service to national party

Service to local party; experience in local office

Experience serving mining areas, local office

Experience in local office; willingness to serve local party

Experience in local office; service to local party

Service to local party; experience in local office

Service to local party; experience in local office

Service to local party; experience in local office

Source: Based on the most prominent answers in author interviews with general and organization secretaries of PSOE (including PSE and PCE), PP, CiU, BNG, and PNV, September 2002

January 2003 and June 2003. If not specified, entries refer to all partisan respondents.

67

Comparative Politics October 2005

The rules governing the legislative process, codified into the Reglamento del

Congreso de los Diputados of February 1982, strongly support the autonomy and

power of national partisan brokers. Individual deputies have little autonomy from

their grupos parlamentarios (parties) or their leaders.19 The law prohibits individual

politicians from forming their own organizations. Deputies who fail to join one of

the congreso's parties are lumped into a single heterogeneous grouping called the

grupo mixto (mixed group), which must act as one entity. Legislation must be pro

posed by a parliamentary group, and amendments to an entire bill must be presented

by a parliamentary group.20 The parliamentary groups shape the composition of the

official governing body of the parliament, the mesa, that is formed prior to the for

mal instauration of a new government and is responsible for procedure. The parties also elect the president of the chamber, who heads the mesa. Thematic committees

(some permanent and others ad hoc), whose composition is based on the distribution

of seats by party in the congreso, reshape initial legislation and release reports that

are the focus of discussion in plenary sessions before votes are taken. The formation

of committees and even the implementation of parliamentary rules on a daily basis

are determined by the whips of the congreso's parliamentary groups, who negotiate

frequently as a rules committee (junta de portavoces). Consequently, individual

deputies in interviews and surveys report consistently that they have little influence

whenever a single party enjoys an absolute majority, since the party's leaders tend to

determine the agenda.21 In short, the Spanish party system has elements of centralization and decentral

ization that together allow for the paradox of a system of nationally disciplined and

generally centralized partisan organizations that are capable of accommodating their

interests with those of unicameral subnational legislatures and subnational party offices. This dual system, as some scholars have argued, creates a channel for subna

tional interests to pressure national parties to decentralize authorities and resources

to favor their subnational partisans. Boix and Grau i Creus found evidence in pre

vailing spending patterns and policy choice that seem to indicate the influence of

intraparty pressures associated with subnational actors.22 The evidence suggests that

subnational partisans regularly shape the policy choices of national party leaders, who are disposed to be responsive because the electoral interests of the national

organization rely on the need to cultivate subnational constituencies.

If the dual hypothesis is correct, both the national and regionalist parties in con

gress should process subnational influence on policymaking. For example, they should recruit members with substantial subnational political experience and con

tacts.23 Subnational governments and party offices, for their part, have strong incen

tives to promote the political careers of partisans with strong local ties, as they will

provide their local bailiwicks with a say in parliament and the national partisan

apparatus.24 National and subnational party leaders should be interested in perpetu

68

Alfred P Montero

ating the careers of these subnational representatives in congress. The length of a

Spanish deputy's experience in the congreso is a determinant of a deputy's informal

ties and understanding of the parliamentary process.25 Deputies who sit for a single mandate have less access to the backroom deal-making where the details of legisla

tion are composed.

Since the regional electoral committees tend to favor regional interests in their

composition of lists for the regional parliament, national deputies with experience as

regional deputies should have the best combination of service to the party and a pro file of representing the region's interests. Have subnational party offices effectively

promoted these candidates on national lists? Do these national deputies have longer runs in the congreso than deputies with other subnational elected or appointed expe riences? Using data for all seven legislatures and the constituent assembly (LC), Table 2 demonstrates that, while the percentage of deputies with subnational experi ence expanded during the democratic period, municipal experience has been more

representative of the subnational cohort in the congreso than regional parliamentary

experience. Moreover, deputies with municipal experience serve less time on aver

age than deputies with no subnational experience. Deputies with subnational parlia

mentary experience tend very slightly to serve more time, but this tendency is not

appreciable in either the descriptive or the correlative data.

Through bivariate and multivariate ordinary least squares statistical analyses, the

role of regional experience and type of experience were tested as determinants of the

number of terms served in the congreso. None of these variables was significant. Even if the regional party offices have some (albeit not determinate) say in the selec

tion of candidates for the national lists, they do not put up politicians with extensive

subnational experience, and even when these candidates gain seats in the congreso,

Table 2 Subnational Experience and Longevity in the Congreso by Type, LC-VII

% _Average Terms Served**_ Regional Regional Municipal Reg. No

Term Experience Deputies Office* Deputies Mayors Councilors Experience

LC 24.0 9.5 8.9 1.99 1.76 1.88 1.97

I 27.9 i 0.9 11.4

II 36.2 11.2 17.6

III 41.1 14.0 22.9

IV 49.2 20.1 30.4

V 51.2 24.3 33.9

VI 58.1 23.5 37.0

VII 59.6 24.5 41.1

N-1597

aIncludes only mayors and members of municipal councils.

^For all legislatures combined.

69

Comparative Politics October 2005

they do not stay long enough to gain much influence. By contrast, the regional gov ernments are composed by politicians favored by the subnational party office, so

these politicians have both the incentives and the capacity to respond to the particu lar needs of the region. Although a more systemic analysis of the careers of the sub

national cohort in the congreso is necessary to come to firmer conclusions, the evi

dence thus far suggests that the system is dual in the sense that party brokers exist at

two levels. National party brokers articulate the interests of their national affiliations

in the congreso, while regional party offices address the interests of the regions and

shape political careers in their regional assemblies. Yet the latter do not exert their

interests or their representatives in the intraparty dimension in the congreso.

Divided Government and the Nationalist Parties

Some scholars of Spain have argued that decentralization has occurred when nation

al parties are in particular need of subnational support. Cases of divided government, the absence of same-party majorities in the executive and legislative branches or

absolute majorities in systems with coll?gial executives, will therefore be associated

with greater levels of decentralization. In another variant of divided government, the

party or coalition that controls the national government is not the same as the one

that controls a majority of the subnational governments. Colomer usefully defines

the two variants of divided government as "horizontal" (national executive-legisla

tive) and "vertical" (national executive-subnational executive), respectively.26 Riker

refers to the latter as "disharmony" and sees it as both a cause of intergovernmental conflict and a consequence of the inability of national parties to control subnational

opponents.27

In Spain periods of horizontal divided government have coincided with the for

mation of alliances between one of the major parties of the center, the PSOE or the

PP, and the nationalist parties. More than the United Left (Izquierda Unida, IU), the

third statewide party, the nationalist parties of Catalonia (CiU), the Basque Country

(PNV), and the Canary Islands (CC) have been successful in negotiating governing alliances in the congreso. Nationalist parties based in particular regions have fewer

opportunities than statewide parties to acquire discretionary central government resources. Hence they have strong incentives to bargain their votes and support with

governing parties to get a piece of the fiscal pie for the regions they represent.28 These pressures have been most salient when national parties can not form majority

governments or enact extraordinary legislation through supermajorities without the

support of regionalist parties (PSOE in 1993-96 and PP in 1996-2000).29 The

nationalist parties are also ideologically centrist, which favors them as strategic alliance partners when compared with IU.30

70

Alfred P Montero

If correct, the horizontal divided government theory implies that the pattern of

decentralization in Spain is overall one of slow or little decentralization punctuated

by episodes of more intense decentralization during periods of divided government. But the actual pattern of decentralization does not bear out this prediction. As Table

3 demonstrates, the Spanish state became more decentralized, at a faster rate, during

periods when the PSOE held its absolute majorities in the congreso (1982-93) than

in the more recent period of minority governments. To be sure, this decentralization

had much to do with the dynamics of the democratic transition and the creation of

the state of the autonomies, yet minority governments in the 1990s also decentral

ized significantly, even if not at the same high rate.

The argument that the nationalists acted as powerbrokers or veto players during these minority governments must be subjected to a counterfactual test. Would decen

tralization have been less without their influence? A perspective that looks at the

entire process would suggest that the decentralization of the 1980s produced a pow erful trend that could not have slowed down even if the PP, supposedly the most cen

tralizing party (a point contested below), gained an absolute majority in 1996.

Between 1979 and 1999 the central government's share of total public spending in

Spain fell from 90 to 61.3 percent, while that of the autonomous communities rose

from 12 to 26 percent.31 No other decentralized or decentralizing state in Europe

experienced a similar level of fiscal decentralization during this period.32 To be sure, the nationalist parties have been catalysts for some decentralization, as illustrated

below, but their influence does not explain the overall pattern. Has the influence of the nationalist parties appeared primarily in the congreso?

The horizontal divided government argument requires that the national composition of these parties reflect local interests and experiences. Yet the legislative branches of

the nationalist parties do not always directly or unquestionably represent their

regions. On average, members of the nationalist parties tend to have more subnation

al experience prior to taking their seats in the congreso (57 percent for all eight leg islatures) than the PP (52 percent) and the PSOE (44 percent). Yet, if they represent their regions' interests in the congreso, they do not do so very long. Seventy-nine

percent served one or two terms, compared to 73 percent of the PP and 71 percent of

Table 3 Rate of Change in the Territorial Distribution of Public Expenditures, 1981-1999

1981-84 1984-87 1987-90 1990-92 1992-97 1997-99

Central -13% -4% -9% -5% -6% -9%

Regional +300% +20% +40% +13% +16% +23%

Source: Based on author's calculations from figures in Luis Moreno, The Federalization of Spain

(London: Frank Cass, 2001), p. 66.

71

Comparative Politics October 2005

the PSOE, and only nine percent served four or more mandates, compared to 13 per cent of the PP and 19 percent of the PSOE. Members of the nationalist parties are

also a third more likely to ask for a baja, or permission from their party to renounce

their seat prior to the end of the term, than members of the PP and PSOE. Their

motivations as representatives of their regions' interests are also suspect. Surveys of

Catalanist deputies show that they are less likely than deputies from the PP or the

PSOE to define themselves as representatives of their region and they are more like

ly to identify themselves as representatives of their party.33 They are also less likely to meet with members of the Generalit?t parliament, which presumably they repre

sent, than are the PP and PSOE with regional parliaments in general.34 Not surpris

ingly, national-subnational conflicts have emerged regularly within the CiU. For

example, Miquel Roca's stewardship of the CiU in the congreso led him to clash

with Jordi Pujol, the longtime president of the Generalit?t, on numerous occasions.35

Differences among the congressional strategies of the nationalist parties also raise

doubts that the nationalist parties fulfill the same function in the congreso. CiU has

more of a parliamentary strategy in that it has maintained a steady leadership at the

national level, while turnover is higher in the PNV and its legislative production is a

third of the CiU's.36

The vertical divided government argument relies on Riker's point that party bro

kers at the national level must control the political careers of the rank and file.

Otherwise, intergovernmental harmony would not necessarily favor centralization.

Yet in Spain the weakness of the representation of subnational interests in the con

greso and the inability of subnational governments to influence their copartisans in

the lower house through regular party channels give the regions incentives to pursue

extralegislative lobbying. Therefore, a high level of harmony would not necessarily favor centralization in Spain, nor would a high level of disharmony guarantee decen

tralization.

Accordingly, Gordin found that there is no strong correlation between fiscal

decentralization and the degree of partisan disharmony between national and subna

tional levels in Spain.37 One reason for the weak correlation is that governments with

absolute majorities, contrary to the logic of party organization arguments, have not

preferred to centralize resources. Both the PSOE and PP governments have decen

tralized fiscal resources to partisan and nonpartisan subnational governments, and

the data do not show consistently that they have favored their copartisans dispropor

tionately. For example, in 1996-99, when the PP was in power, two Socialist regions

(Andalusia and Castille-Le?n) tended to be among the top three regions (along with

the PP's partner, Catalonia, led by the CiU) to receive transfers and to sign coopera tive agreements with the center. Also, subnational governments, regardless of the

party in power, are not restrained in lobbying for more resources. Subnational inter

ests matter, and they will be represented in intergovernmental relations where subna

tional advocates can find political space to do so.38

72

Alfred P Montero

Party Structure Arguments

Differences in party structure may explain the paradox of an apparently disciplined

party system with continued decentralization. While the PP has a highly hierarchical

structure, the PSOE's federal structure may explain the decentralization during the

Socialists' absolute majorities from 1982 to 1993.39 Some of the key figures on the

governing federal council of the party, who are commonly referred to as the barons

of the organization, are the current or former presidents of regions such as

Extremadura, Castille-La Mancha, and Andalusia. The Catalan and Basque Socialists maintain their own party organizations (PSC and PSE, respectively) that

are affiliated with the national PSOE and follow it in the congreso, but they explicit

ly represent the interests of their regions. Since 2000 candidate selection in the

PSOE has been determined by internal primaries, thus increasing the influence of

subnational interests. The PP, by contrast, had few majorities in the regional parlia ments prior to 1995, so it has less need to have a decentralized internal organization. Also, its internal cleavages are more ideological and less regionalist than the older

PSOE's.40

Yet as an explanation for decentralization party differences seem to matter little.

As reported in Table 3, fiscal decentralization accelerated once the PP came to

power in 1996 and belies the notion that one party is more sensitive to subnational

concerns than the other. The data show that more PP deputies have held subnational

office prior to gaining seats in the congreso, 52 percent compared to 44 percent of

PSOE deputies. In a poll of 212 deputies conducted in 1997, Uriarte showed that

"service to the region" was slightly more important to PP deputies than to Socialist

ones.41 National surveys done by the Centro de Investigaciones Sociol?gicas (CIS) show that deputies from both the PP and PSOE declare that representing their region is important in roughly equal proportions.42

Several factors account for the similarities of the two major parties in terms of

how they mediate the effects of decentralization. First, the parties operate under sim

ilar guidelines in the parliamentary process. Because of the need to maintain parlia

mentary discipline, their leadership is uniformly opposed to factionalism, of which

regional loyalties are one variant. Second, as reported above, candidate selection

methods are similar in both parties. Regional and provincial party offices have the

right to propose candidate lists, but national electoral committees can amend these

lists and hold ultimate authority over their approval. Third, both the PP and the

PSOE have tended to address conflicts involving intergovernmental relations outside

the parliamentary arena. This tendency in itself is indicative of similar internal struc

tures. The PP and the PSOE are centralized and disciplined parliamentary parties that close off the autonomy of regional interests in the congreso. But the federal

structure of the Spanish state creates other arenas in which regional concerns can be

heard.

73

Comparative Politics October 2005

The Weakness of Intergovernmental Institutions

Some observers of the Spanish intergovernmental system have argued that it has

evolved in the last few years into a phase of "cooperative federalism." 43 Since the

intergovernmental system maintains numerous forums for center-periphery coopera tion?sectoral conferences, the senate, and bilateral agreements?the party system need not aggregate the entire range of subnational interests. Various multilateral

organizations emerged during the evolution of the state of the autonomies to iron out

differences in policymaking between the central government and the seventeen

regions. The representatives from the regions, the relevant ministries, and the gov

erning party or coalition sit on twenty-four sectoral conferences (conferencias secto

riales) and assorted interministerial committees that are organized thematically by

policy area. In most cases negotiations on these committees flesh out the technical

details of administrative decentralization, but in cases where there is significant

intergovernmental disagreement, the principals move negotiations to a joint commit

tee for the transfer of administration (comisi?n conjunta a para la transferencia de

competencias), which is staffed by representatives from the responsible ministry, the

regional government(s), and the ministry of public administration.

The activities of these intergovernmental committees and sectoral conferences,

however, demonstrate that they are limited in a number of ways. First, following a

ruling by the constitutional tribunal in 1983 (ruling 76/1983), the sectoral confer ences can not override or change in any meaningful way the authorities already devolved to the regions. This protects the regions' extant policy rights, but it also

limits the extent to which extensions of regional autonomy can be negotiated through the sectoral conferences. Consequently, the relevance of the sectoral conferences has

waned. Ministry of public administration data analyzed by Grau i Creus demon

strates that the frequency of meetings decreased after 1981-86, when the regions

garnered most of their authorities.44 The national government has shown little inter

est in reviving the sectoral conferences. Presently, it does not maintain consistent

representation for permanent agencies on most of them. Second, the sectoral confer

ences and the interministerial committees are also limited in their capacity to allow

the regions a say in intergovernmental relations, illustrated by the example of the

council of fiscal and financial policy of the autonomous communities (consejo de

pol?tica fiscal y financiera de las comunidades aut?nomas), the primary body

responsible for drafting changes to the fiscal and administrative structure of the state

of the autonomies. Council rules give the national government qualified votes, in

practice as many votes as the regions, so the council is nothing more than a vehicle

through which the government gives itself recommendations.45

The weakness of multilateral forms of shared rule between central and regional

governments has underscored the importance of bilateral agreements in resolving

74

Alfred P Montero

disputes. The chief form of bilateral agreements are joint planning agreements known as convenios. Most of the convenios involve the provision of central-level fis

cal resources to finance policy areas previously devolved to the regions; in most

cases, no multilateral negotiation is necessary, as either most are signed by one

region or subsequently other regions bilaterally join the original signatory.46

Corporatist rules promoted a centralization of patterns of labor and business rep resentation of their interests during the first years of democratization.47 The same

can not be said for the regional governments. Given the weakness of intergovern mental institutions and the vacuousness of subnational representation in parliament, no corporatist system of interest intermediation for the regions emerged during or

after the democratic transition. And even though corporatist institutions contributed

to the centralization of economic and fiscal policymaking during the early years of

the new democracy they did so less after years of intergovernmental bargaining over

the fiscal structure of the state.

The Persistence of Bilateral Intergovernmentalism

The autonomy process created a set of institutions that gave the regions incentives

and opportunities to renegotiate their fiscal powers over time. Given the weakness of

subnational representation in the congreso, within the major parties, and in the for

mal institutions of intergovernmental relations, renegotiation was continuous but

largely uninstitutionalized. The resulting pattern of fiscal federalism was one of per

sisting bilateral intergovernmentalism that produced an uneven distribution of resources and powers and a never-ending spiral of cross-regional and intergovern

mental conflict.

Once the state of the autonomies became part of the constitution, negotiations

began in the congreso over the shape and content of the fiscal federal system. Between July 1979 and September 1980 draft legislation went back and forth between

the two chambers of the Cortes. With UCD, PSOE, and PCE for and the AP/PP, the

Catalan nationalists, and the Andalusian Socialist Party against, the Cortes approved

Ley Org?nica 8/1980 on September 22, 1980. The Organic Law of Financing the Autonomous Communities (Ley Org?nica de Financiaci?n de las Comunidades

Aut?nomas) established a provisional period of five years (1980-86) in which the

details of each region's fiscal responsibilities in two succeeding five-year periods

(1987-91 and 1992-96) would be negotiated with the national government. Article 13

of the law defined the role of the regions in collecting taxes for the central govern ment. These tax revenues are defined for each region as the percentage of participa tion in the revenues of the state (porcentaje de participaci?n en los ingressos del esta

do). This figure determines the size of fiscal transfers to each region. Thus, it is usu

ally the focus of intergovernmental negotiation of the fiscal system.48

75

Comparative Politics October 2005

From its inception, the fiscal structure created nodes of cross-regional and inter

governmental conflict by producing an asymmetric model of fiscal federalism. The

seventeen regions were divided into four distinct groups: the special regime regions, the Article 151 regions, the Article 143 regions with multiple provinces, and the

Article 143 regions with single provinces. As Table 4 demonstrates, these differences

determine significantly the resources that each region retains for autonomous policy

making. The multiprovincial and uniprovincial regions spend just over half and one

third, respectively, per inhabitant, than what the Article 151 regions do. Yet even

these regions spend only just over two-thirds of what the special regime regions of

the Basque Country and Navarre spend. These two regions keep most of what they collect in taxes as own resources (approximately 87 percent of their revenues). Other revenues come from fiscal transfers for health care spending (Insalud) and social

security (Inserso). In turn, the Basque Country and Navarre transfer to the national

treasury an amount (the cupo or quota) from their tax revenues that is negotiated every five years.

The Article 151 and 143 regions were subject to a general model of financing. These regions have five types of income. First, and primarily, they collect taxes that

they keep as their own resources, and they receive a percentage of national tax collec

tions through fiscal transfers. Second, they receive subsidies for health care through Insalud and Inserso. Third, they receive fiscal transfers meant to address issues of

underdevelopment and inequality through a national compensatory fund (fondo de

compensaci?n interterritorial) and European structural funding. Fourth, they receive

conditional fiscal transfers after signing bilateral convenios to fund national expendi tures in their territory. Fifth, they can accumulate debt. Fiscal transfers dominate the

Table 4 Fiscal Structure of the Autonomous Communities (in Thousands of Pesetas,

per Inhabitant, Based on the 1994 budget)

Article 151 Regions Average

Canary Is. Catalonia Galicia Valencia 186.7 233.2 240.6 198.9 218.3

Article 143 Regions (Multtprovincial)

Balearic Is. CastiHe- Castille-La Extremadura Leon Mancha

122.2 136.8 166.7 129.2 123.06

Article 143 Regional (Uniprovincial)

Cantabria Madrid Murcia LaRioja 78.5 73.9 103.6 79.9 81.17

Special Regime Regions

Basque Countrv/Navarre 314.8" 314.8

Source: Figures taken from Moldes Teo, "La participaci?n," p. 110,

Andalusia 232.22

Aragon

60.4

Asturias 70.0

76

Alfred P. Montero

financing of the fifteen nonspecial regime regions (approximately 85 percent of all

revenues). The regions spend money that they do not necessarily collect.

Spain's asymmetric fiscal federal structure produced considerable instability in

intergovernmental relations. First, the system created incentives for the regions to

hoard resources and to overstate the costs of the services that they provide; the fiscal

structure is not set and is perpetually subject to renegotiation. Regional governments mobilize politically to engineer terms that are favorable to them individually, even if

the new agreement externalizes costs to their neighbors. Individual regions also

hedge their bets against potential losses in the next round of bargaining by preserv

ing scarce resources or, conversely, spending on visible (though not necessarily use

ful) projects that might justify future funding from the center. Second, the reliance

on fiscal transfers, most of which are unconditional, creates a gap between the costs

of taxation and the benefits of spending. The tax-benefit gap broadens incentives for

the regions to spend in the present and plan future spending without fully under

standing changes in the tax base. Current spending risks overshooting resources over

the medium term. Since fiscal transfers and own resources are often not sufficient to

cover the shortfall, regions must accumulate debt, which contradicts the efforts of

national governments to stabilize public accounts.49 Finally, the inherently unequal distribution of authorities and resources cause the have-not regions to up the ante in

their calls for new authorities. At the same time, the have regions attempt to protect their special rights versus the demands of the common regime regions. Interregional relations, therefore, take on a zero-sum competitive dynamic since all actors know

the national fiscal pie is limited.50 The result is a spiraling of demands for an expan sion of subnational authorities and resources that undermines any kind of rational or

coherent fiscal logic implemented from the center.

The evolution of the Spanish fiscal system reflects this logic of distributive conflict more than the presence of disciplined national parties with absolute majorities in the

congreso. The end of the provisional period in 1986 led to negotiations concerning the

fiscal federal system for the next five years. Each region (except the special regime

regions) negotiated the cost of administering policies within their range of competence. While these costs were based on explicit variables in each area, no central agency was

charged with calculating the real amounts. Each regional government tried to maximize

the cost per policy area in the hope of raising the ceiling of funding. After the first five

year period (1987-91) was negotiated, the second five year period (1992-96) was con

sidered. Both the central government and the regions agreed that total funding per poli

cy area could not decline; therefore, the regional governments had an even stronger incentive to inflate the presentation of administrative costs to raise the funding ceiling still higher. These perverse incentives softened budget constraints considerably and put a premium on the ability of each region to exert leverage on the central government and

the council of fiscal and financial policy.

77

Comparative Politics October 2005

At the same time, the regions were increasingly pressed to find their own fiscal

resources. The imposition of the VAT in 1985 as a prelude to accession to the EU

reduced the domain of taxes already in the hands of the regional governments,

including luxury taxes and the main nonincome taxes, the impuesto sobre transmi

siones patrimoniales and actos jur?dicos documentados. Efforts to decrease the tax

burden and encourage investment led to declines in these taxes after the 1985

reform. The Generalit?t of Catalonia responded by issuing a white paper calling for

the devolution of large percentages of the income tax (impuesto de la renta sobre las

personas fisicas) and an expansion of own resources for the region.51 Such pressure

expanded to include virtually all of the other regions, leading directly to a set of rec

ommendations distributed by the council of fiscal and financial policy in June 1992

to decentralize 15 percent of the income tax to the nonspecial regime regions.52 The

reform of the Ley Org?nica de Financiaci?n de las Comunidados Aut?nomas pro

posed by the council of fiscal and financial policy conditioned fiscal transfers to the

regions on 15 percent of their income tax collections. Once it was implemented in

1994, the nonspecial regime regions gained substantial control over taxation.

The Catalans, once again, proved to be the spearhead of reform, but in a way that

led to the generalized decentralization of the income tax. Having secured extraordi

nary influence in the national government following the governing pacts of 1993, Jordi Pujol demanded a doubling of the region's collection of income tax to 30 per cent. The Socialists acceded to 15 percent, but they had to buy the support of the less

well-off regions of Andalusia, Galicia, and Extremadura by guaranteeing these areas

additional monies from a national compensatory fund.53 Once the PP came to power in 1996 and signed agreements with three nationalist parties?PNV, CiU, and CC?

to maintain a governing majority, it implemented a series of fiscal reforms that

greatly expanded subnational resources. From 1997 to 2001 the national government

agreed to cede the property tax, inheritance and donation taxes, gaming taxes, and

another 15 percent of the income tax.

The strategic-interactive dynamics of Spain's asymmetric fiscal federal system set

up a decentralizing dynamic that superseded the controls sought by governing national parties. Fiscal decentralization proceeded in this open-ended fashion during

majoritarian and minoritarian governments. Conflicts and side payments on the

dimension of national-regional government, not intralegislative or intraparty politics, determined outcomes.

Conclusions

The central paradox of Spanish decentralization is the design of national representa tive institutions of democracy to enhance centralized administration but the continu

78

Alfred P Montero

ing decentralization of the larger apparatus of the Spanish state. The party system failed to reconcile these tensions. The national and regional parties do not aggregate subnational interests either within the congreso or systematically within the parties themselves. Given the mostly nonfunctional role of intergovernmental committees,

most of the major issues facing Spain's federal system have been addressed in the

poorly institutionalized arena of intergovernmental distributive conflict. Iterated bar

gaining, asymmetry of authority and resources among the regions, and a weak center

that was incapable of imposing solutions without resorting to periodic mediation

with regional presidents shaped intergovernmental distribution. Future research must

explore these variables in distinct policy areas to assess intergovernmental dynamics in Spain.

Rikerian party organization arguments provide insufficient explanations of the

degree or pattern of decentralization. The major parties of Spain are disciplined and

centralized organizations for the most part, and they have tremendous control over

the legislative process. But they have not been immune from intergovernmental pres sures to decentralize policy authorities and resources emanating from regional gov ernments headed by subnational copartisans and opposition groups. This finding is

consistent with work on party structures and state reform in other countries. In the

experience of statewide parties in territorially decentralized (and decentralizing) countries in western Europe, organizations that previously favored centralization can

become advocates of decentralization later (for example, the British Labour Party and the Flemish Liberal and Socialist parties), even when they are in government and even if, as with the statewide parties in Spain, they are relatively centralized.

The preferences of central party brokers and their capacity to impose them on their

subnational copartisans vary according to electoral concerns and policy choices.54 The nature of intergovernmental distributive conflict plays a key role in this strategic game. Clearly, the argument that decentralization can be fully understood with refer ence to the degree of centralization in governing parties is not sustainable in compar ative perspective.

Most important in the failure of party organization arguments, the game of car

rots and sticks that supposedly links national and subnational copartisans is not

exclusive to the legislative arena. The behavior of copartisans in the legislature and

in the larger multiarena of intergovernmental relations can differ substantially. The most disciplined parties operating in the most structured legislatures can assure that

their deputies will vote according to the interests of the party brokers, but these insti

tutional sources of discipline are of little use when attempting to convince subnation

al incumbent copartisans to follow the party line. Such requests can threaten the

electoral and policy interests of regional incumbents and require side payments or

wholesale rethinking of reform and policy options. Scholars of representation argue that agents have imperative mandates to do what the principal would do. In the case

of politicians motivated to represent their regions, decentralization of the state and a

79

Comparative Politics October 2005

centralized party system produce a dual imperative to represent both the party and

the region. Politicians in parliament in these cases are poorly placed to fulfill the

dual imperative. Subnational executives, however, can do both, and they have the

political space and the means to reconcile the two imperatives of representation when they enter into conflict. This factor more than any other explains how a cen

tralized and disciplined party system may have to deal with the dynamic of a decen

tralizing state that it does not sufficiently control.

NOTES

The author would like to thank Lourdes L?pez Nieto, Pedro Puy Fraga, and the three anonymous review

ers for Comparative Politics for their helpful comments and suggestions. 1. William H. Riker, The Development of American Federalism (Boston: Kluwer Academic

Publishers, 1987), p. 74.

2. Eliza Willis, Christopher Garman, and Stephan Haggard, "The Politics of Decentralization in Latin

America," Latin American Research Review, 34 (1999), 7-56; Christopher Garman, Stephan Haggard, and

Eliza Willis, "Fiscal Decentralization: A Political Theory with Latin American Cases," World Politics, 53

(January 2001), 205-36. These authors do not distinguish between politicians and party bureaucrats. For

our purposes, the relevant division of interests lies between national and subnational politicians and

bureaucrats, not between politicians and bureaucrats. Party brokers may be likened to what Kitschelt calls

the dominant coalition within the party organization. See Herbert Kitschelt, The Transformation of

European Social Democracy (New York: Cambridge University Press, 1994). 3. See Stephan Haggard and Steven B. Webb, "Political Incentives and Intergovernmental Fiscal

Relations: Argentina, Brazil and Mexico Compared," in Alfred P. Montero and David J. Samuels, eds., Decentralization and Democracy in Latin America (Notre Dame: University of Notre Dame, 2004).

4. Garman, Haggard, and Willis, "Fiscal Decentralization," p. 209.

5. Haggard and Webb, pp. 241^13.

6. Willis, Garman, and Haggard, "The Politics of Decentralization," p. 48.

7. See Jos? Ram?n Montero and Mariano Torcal, "Autonom?as y Comunidades Aut?nomas en

Espa?a: Preferencias, dimensiones y orientaciones pol?ticas," Revista de Estudios Pol?ticos (Nueva

?poca), 70 (October-December 1990), 34.

8. See Alfred P. Montero, "After Decentralization: Patterns of Intergovernmental Conflict in

Argentina, Brazil, Spain, and Mexico," Publius: The Journal of Federalism, 31 (Fall 2001), 43-64.

9. See Tulia Falleti, "Of Presidents, Governors, and Mayors: The Politics of Decentralization in Latin

America," paper presented at the meeting of the Latin American Studies Association, Dallas, March

27-29, 2003.

10. See Alfred P. Montero, Shifting States in Global Markets: Subnational Industrial Policy in

Contemporary Brazil and Spain (University Park: Penn State University Press, 2002), ch. 4.

11. See Hector Schamis, Re-Forming the State: The Politics of Privatization in Latin America and

Europe (Ann Arbor: University of Michigan Press, 2002). 12. See Josep M. Colomer, "The Spanish 'State of the Autonomies': Non-Institutional Federalism,"

West European Politics, 21 (1998), 40-52.

13. See Richard G?nther, "Electoral Laws, Party Systems, and Elites: The Case of Spain," American

Political Science Review, 83 (September 1989), 835-58.

14. Ibid.

15. See Kitschelt, pp. 223-24.

80

Alfred P. Montero

16. Juan J. Linz, "Spanish Democracy and the Estado de las Autonom?as," in Robert A. Goldwin, Art

Kaufman, and William A. Schambra, eds., Forging Unity Out of Diversity: The Approaches of Eight Nations (Washington, D.C.: American Enterprise Institute for Public Policy Research, 1989), pp. 297-98.

17. Richard G?nther and Jonathan Hopkin, "A Crisis of Institutionalization: The Collapse of the UCD

in Spain," in Richard G?nther, Jos? Ram?n Montero, and Juan J. Linz, eds., Political Parties: Old

Concepts and New Challenges (New York: Oxford University Press, 2002). 18. Garman, Haggard, and Willis, "Fiscal Decentralization," p. 212.

19. Lourdes L?pez Nieto, "Las Cortes Generales," in Manuel Alc?ntara and Antonia Mart?nez, eds.,

Pol?tica y gobierno en Espa?a, 2nd ed. (Valencia: Tirant lo Blanch, 2001), p. 228.

20. Amendments to particular articles may be made by individual deputies, but the party's spokesper son (portavoz) must sign off before the change is presented formally. See Pablo O?ate, "La organizaci?n del Congreso de los Diputados," in Antonia Mart?nez, ed., El Congreso de los Diputados en Espa?a:

Funciones y rendimiento (Madrid: Tecnos, 2000). 21. Irene Delgado, "Elites pol?ticas y vida parlamentaria: Actividades y motivaciones de los diputados

espa?oles," in Mart?nez, ed.

22. Caries Boix, Political Parties, Growth and Equality: Conservative and Social Democratic

Economic Strategies in the World Economy (New York: Cambridge University Press, 1998), pp. 142^45;

Mireia Grau i Creus, "The Effects of Institutions and Political Parties upon Federalism: The Channelling and Integration of the Comunidades Aut?nomas within the Center-level Policy Processes in Spain

(1983-1996)" (Ph.D. diss., European University Institute, Florence, 2000). 23. See L?pez Nieto, "Notas sobre los pol?ticos: Opiniones de alcaldes y diputados espa?oles sobre su

quehacer," Working Paper No. 179 (Barcelona: Institut de Ci?ncies Politiques i Socials, 2000), p. 9.

24. See Michael Keating, The New Regionalism in Western Europe: Territorial Restructuring and

Political Change (Cheltenham: Edward Elgar, 1998). 25. See Lourdes L?pez Nieto, Mercedes Alda, Esther del Campo, Jos? Ram?n Laorden, El?seo L?pez,

Teresa Lorenzo, and Antonia Monteagudo, "Un primer balance sobre la actividad de los parlamentos auton?micos" (mimeo, 2003), p. 19.

26. Josep M. Colomer, "Las instituciones del federalismo," Revista Espa?ola de Ciencia Pol?tica, 1

(2000), 41.

27. Riker, pp. 93-94.

28. Juan J. Linz, "Democracia, multinacionalismo y federlismo," Revista Espa?ola de Ciencia

Pol?tica, 1 (2000), 25.

29. William B. Heller, "Regional Parties and National Politics in Europe: Spain's Estado de las

Autonom?as, 1993 to 2000," Comparative Political Studies, 35 (August 2002), 657-85.

30. Jordi Capo Giol, "Sistema electoral y gobernabilidad espa?ola," Revista Espa?ola de Ciencia

Pol?tica, \ (2000), 61.

31. See Pedro Puy Fraga, "Financiaci?n auton?mica y reforma del Senado," in Enrique Moldes Teo

and Pedro Puy Fraga, eds., La financiaci?n de las Comunidades Aut?nomas (Madrid: Minerva Ediciones,

1996), p. 60.

32. See Francesc Morata, "El Estado de las Autonom?as: Veinte a?os de rodaje," in Alc?ntara and

Mart?nez, eds., p. 131.

33. Antonia Mart?nez and M?nica M?ndez, "La representaci?n pol?tica en el Congreso espa?ol," in

Mart?nez, ed., pp. 236-37.

34. Ibid, p. 261.

35. Linz, "Democracia," p. 25.

36. See L?pez Nieto, "Las Cortes," p. 238.

37. Jorge P. Gordin, "Unraveling the Politics of Intergovernmental Transfers: Comparative Evidence

from Argentina and Spain," paper presented at the meeting of the Midwest Political Science Association,

Chicago, April 3-6, 2003.

81

Comparative Politics October 2005

38. William Downs, Coalition Government, Subnational Style: Multiparty Politics in Europe's

Regional Parliaments (Columbus: Ohio State University Press, 1998), p. 218.

39. See Mireia Grau i Creus, "Spain: Incomplete Federalism," in Ute Wachendorfer-Schmidt, ed., Federalism and Political Performance (New York: Routledge, 2000), pp. 72-73.

40. Luz Moran, "Renewal and Permanency of the Spanish Members of Parliament (1977-1993): Reflections on the Institutionalization of the Spanish Parliament," Working Paper No. 81 (Madrid: Center

for Advanced Study in the Social Science, Juan March Foundation, 1996). 41. Edurne Uriarte, "La pol?tica como vocaci?n y como profesi?n: An?lisis de las motivaciones y de la

carrera pol?tica de los diputados espa?oles," Revista Espa?ola de Ciencia Pol?tica, 3 (October 2000), 108.

42. Mart?nez and M?ndez, p. 236.

43. For example, see Tanja A. B?rzel, States and Regions in the European Union: Institutional

Adaptation in Germany and Spain (Cambridge: Cambridge University Press, 2002); and Robert Agranoff, "Las relaciones intergubernamentales y el Estado de las Autonom?as," Pol?tica y Sociedad, 13 (1993), 87-105.

44. Grau i Creus, "The Effects of Institutions," p. 117.

45. See Puy Fraga, p. 67.

46. Luis Moreno, "Decentralization in Spain," Regional Studies, 36 (2002), note 12. The special fiscal

regime in Navarre and the Basque Country usually obviates the need for these regions to sign convenios.

47. For more on this argument, see Omar Encarnaci?n, "Federalism and the Paradox of Corporatism," in Joanne Bay Brzinski, Thomas D. Lancaster, and Christian Tushhoff, eds., Compounded Representation in West European Federations (Portland: Frank Cass, 1999), pp. 102-3.

48. See Enrique Moldes Teo, "La participaci?n en los ingresos del Estado," in Moldes Teo and Puy

Fraga, eds.

49. See Mar?a Casais Mira, "Los recursos tributarios de las Comunidades Aut?nomas: Especial refer

encia al caso gallego," in Moldes Teo and Puy Fraga, eds., p. 103.

50. See Puy Fraga, pp. 64-65.

51. Generalit?t de Catalunya, Llibre blanc del govern de la Generalit?t de Catalunya sobre elf?nan?a ment autonomie (Barcelona: Departement d'Economia i Finances, 1985).

52. See Emilio Alvarado P?rez, "Veinte a?os de proceso auton?mico: Del federalismo cooperativo al

federalismo dual," in Juan Luis Paniagua Soto and Juan Carlos Monedero, eds., En torno a la democracia

en Espa?a: Temas abiertos del sistema pol?tico espa?ol (Madrid: Editorial Tecnos, 1999), p. 370.

53. Morata,p. 138.

54. For example, see Seth Goldstein, "Party Leaders, Power and Change," Party Politics, 8 (2002),

327?48; Downs, Coalition Government.

82