McGill University POLI 631 Seminar paper Montreal, April 26th, 2010

Devolution all'Italiana: The Rationale of Fiscal Decentralization in Italy.

Jonas Teusch M.A. Political Science Comparative Politics

1. Introduction It is widely acknowledged that the basic structure of government is transformed around the world as authority and resources migrate from central to subnational governments (Rodden 2004: 481). Of greatest theoretical interest is hereby fiscal federalism, as theory suggests that fiscal decentralization would be associated with a number of normatively desirable outcomes, in particular increased national income and the provision of higher quality public goods. Not surprisingly, many empirical studies try to verify if - and under what conditions - these promises of fiscal federalism hold true in practice (e.g. Rodden 2006). Yet, there is also a growing body of literature that is more interested in the causes than the effects. This seems important too, as a better understanding of the complex rationale that underlies fiscal decentralization could help explain why reforms have occurred in some states while institutional structures remained unchanged in others. In recent years, both quantitative and qualitative empirical studies have contributed to this research agenda, identifying crucial institutional, structural, political, and international explanatory variables. However, instead of complementing each other, these two approaches do mostly ignore the other. While quantitative studies theorize and test a vast array of potentially powerful explanatory factors but remain rather vague about their specific effects, qualitative scholars focus on few fairly idiosyncratic aspects of fiscal federalism, but often fail to specify a complete model. This paper will test theoretically well-grounded hypotheses, drawing from the qualitative evidence provided by the Italian case. Italy seems to be a particularly promising subject of analysis, because it fulfils many of the conditions that the theoretical and empirical literature on the subject matter identify as being associated with fiscal decentralization. This paper will therefore be able to verify if the rationale of fiscal decentralization does indeed manifest itself as expected. It will be shown that, while structural factors do have effects similar to the ones suggested by quantitative studies, the

complex political rationale, which proves to be crucial in the Italian case, is not adequately accounted for by Large-N studies. This paper proceeds as follows. Section two provides a literature review, discussing both theoretical and methodological shortcomings of existing research. Section three conceptualizes fiscal federalism. Section four develops theoretically informed hypotheses about the institutional, structural, political, and international rationale underlying fiscal decentralization. Section five lies out a methodological approach that allows for a sophisticated assessment of these hypotheses. Section six provides a factual account of Italian decentralization, assessing to what extent the institutional reforms have made it approximate fiscal federalism as conceptualized in section three. Section seven will test the theoretically informed hypotheses from section four. Section eight concludes.

2. Literature Review Ever since the publication of Tiebouts seminal article positing A Pure Theory of Local Expenditure (1956), political scientists and economists have been arguing about the normative implications of fiscal federalism (e.g.: Oates 1999; Rodden 2006; Sorens 2010; Weingast 1995). However, comparatively few studies analyze empirically for what reasons fiscal decentralization occurs. First of all, it has to be noted that there is no unique and well-accepted theory to be tested regarding the identification of causes for FD [fiscal decentralization] to vary across countries and over time (Letelier 2005: 156). Instead, the literature provides various independent and often contradictory hypotheses, involving institutional, structural, political, and international explanatory variables.1 Probably because of data limitations, political factors are less emphasized in quantitative studies (for a recent exceptions see: Jametti and Joanis 2010). Qualitative work, by contrast, puts great emphasis on the political rationale (e.g. Mazzoleni 2009). All the different hypotheses that may be amenable to the Italian case will be discussed in detail in

For a concise overview of the quantitative literature, see Letelier (2005: 162).

section four; instead, this part focuses on the major flaws that are inherent to both the quantitative and the qualitative empirical approach to fiscal decentralization.2 The major problem with the quantitative empirical literature is conceptual. The theoretical literature on fiscal federalism stresses that, for fiscal federalism to unfold its hypothesized utility maximizing effects, subunits need to enjoy both fiscal and programmatic autonomy (Riker 1964; Rodden 2004; Weingast 1995).3 Yet, quantitative studies pay little - if any - attention to the latter dimension. Fiscal decentralization is almost without exception simply operationalized as either the share of general government expenditures made by sub-national governments or as the share of sub-national government revenues in the general governments revenues (e.g. Arzaghi and Henderson 2005; Oates 1972; Panizza 1999). Stegarescus data is slightly improved (2005; 2009), as he identifies autonomous sub-central tax revenues. Yet, as Sorens points out, this is still no good proxy of sub-central policy autonomy (2010). The qualitative literature usually takes a more comprehensive approach to decentralization, discussing institutional reforms with regards to both programmatic and fiscal autonomy which only when taken together constitute fiscal decentralization as it is understood in this paper (Fabbrini and Brunazzo 2003; Roux 2008). But while avoiding conceptual flaws, this part of the empirical literature suffers from two other major shortcomings. First, most qualitative research that deals with the Italian case does not develop a theoretically well grounded framework of analysis, but is limited to a rather descriptive account of the various stages of decentralization (see: Amoretti 2002; Keating and Wilson 2010; Roux 2008). It might well be true that the Italian case is mainly driven by idiosyncratic factors; yet, especially since many authors stress the importance of international factors such as Globalization and Europeanization (e.g. Garret and Rodden 2003), insights from other advanced industrialized

The former approach usually takes a cross-country perspective (Italy is almost always included). Qualitative work mainly takes the form of case studies. 3 See section three below.

democracies that are subject to similar constraints, would certainly merit further attention (Rodden 2004). Second, while quantitative work systematically tests a vast array of potentially relevant independent variables, qualitative work tends to take a fairly narrow focus. For example, Mazzoleni analyzes decentralization through the lenses of party competition (2009), Huysseunes work is centered on the extreme right (2006), and Fabbrini and Brunazzo discuss Europeanization and domestic mobilization (2003). However, as sophisticated quantitative work demonstrates, independent variables may interact with one another (Stegarescu 2009). Therefore, by focusing on few explanatory variables only, these authors may miss the big picture and come to flawed conclusions (Rodden 2004). This study will contribute to both the qualitative and the quantitative literature on the subject matter. First, before empirically assessing the Italian case, this study will theorize hypotheses about the institutional, structural, political, and international factors that are potentially relevant for fiscal decentralization in an advanced industrialized democracy such as Italy. This will allow for a more comprehensive analysis of Italian decentralization than the ones provided by the existing case study research. In addition, by qualitatively testing hypotheses usually only examined in quantitative research designs, it will be possible to shed some light on how exactly these factors unfold their effects.

3. Conceptualizing Fiscal Federalism It lies outside the scope of this paper to evaluate if the promises of fiscal federalism do indeed manifest themselves in practice.4 Yet, if this paper wants to truly contribute to the growing literature on fiscal decentralization, it is crucial to capture the concept in a way that is in line with fiscal federalism theory. Only by doing so, research on both the causes and the effects of fiscal decentralization can complement one another. Therefore, this section tries to identify the important elements of fiscal federalism as they

For an excellent analysis, see Rodden: Hamiltons Paradox: The Promise and Peril of Fiscal Federalism (2006).

are suggested by the theory of fiscal federalism. Drawing from Riker, Weingast, and Rodden, Jason Sorens identifies four elements of the ideal type of fiscal federalism (2010):
1) Sub-central governments (SCGs) enjoy programmatic autonomy, i.e., exclusive authority to decide a subset of economic policy (creating, repealing, and adjusting programs and regulations). 2) SCGs face a hard budget constraint, funding their own spending largely out of autonomous revenues, i.e., revenues raised through taxes over which the SCGs enjoy the authority to set rates, base, or both, and may not have access to unlimited credit. 3) There is a common market, so that SCGs may not enact barriers to the free flow of goods, capital, and labor across their borders. 4) The system is institutionalized, so that the central government may not alter it at will.

Obviously, only truly federal political systems can approximate all four elements of this ideal type, because in merely decentralized unitary polities the central government may still unilaterally decide to take away programmatic or fiscal autonomy from subunits (Osaghae 1990). Therefore, the fourth condition can only be met by states that dispose of a federal constitution. Yet, it should be clear that unitary systems may come closer to the ideal of fiscal federalism than some federal political systems. In the Federal Republic of Germany, for example, the laender, though enjoying highly institutionalized competencies (criterion no. 4), have relatively little exclusive competencies. On the contrary, in the unitary Italian political system, the powers granted to the regioni by the Italian constitution are limited, but in recent years they gained some programmatic and fiscal autonomy.5

Table 1: Typology of Programmatic and Fiscal Autonomy (Sorens 2010)

See section six.

Fiscal decentralization, then, as it is understood in this paper, refers to institutional reforms that move a political system closer to the ideal of fiscal federalism. Consequentially, the concept of fiscal decentralization is amenable to the analysis of both federal and unitary political systems. Institutional reforms can encompass two conceptually distinct dimensions, namely fiscal and programmatic autonomy. Since these two dimensions complement one another, institutional reforms of one dimension will only lead to greater levels of fiscal federalism if the other dimension disposes of some minimal degree of autonomy as well. For example, a central government that empowers subunits with fiscal autonomy, but does not grant any programmatic to them, does not approximate fiscal federalism.6

4. Theorizing Fiscal Decentralization The Institutional Rationale The major theoretical approaches comparative political scientists apply to explain institutional change are the three new institutionalisms, namely Rational Choice Institutionalism (RI), Sociological Institutionalism (SI), and Historical Institutionalism (HI). The three schools differ in explaining how institutions originate and change, but all three agree that new institutions are created or adopted in a world already replete with institutions (Hall and Taylor 1996: 949). Consequentially, it seems to be obvious that one needs to take into account the existing institutional structure. Probably because constitutional settlements vary greatly across countries, the institutional rationale is heavily underrepresented in Large-N studies.7

Denmark is a particularly striking example of how misleading measures that focus solely on the fiscal autonomy dimension can be. The Scandinavian country scores third on Roddens ranking of fiscally decentralized countries, though the central government tightly regulates virtually every aspect of local government finance (Rodden 2004: 483). 7 Not surprisingly, there is only one major hypothesis advanced by quantitative scholars. It stresses that federal political systems are associated with increases in fiscal autonomy, because if regional politicians have de iure or de facto veto power in the process of federal policy making, this should allow subnational officials to bargain for larger shares of the public sectors resources (Garrett and Rodden 2003: 96). Obviously, this hypothesis could only be tested in a cross-country design and is therefore not relevant here.

RI is an intentionalist approach, driven, not by impersonal historical forces, but by a strategic calculus (Hall and Taylor 1996: 945). Institutions originate because they perform certain functions that are of value to those who have the power to create them. In other words, existing institutions change and new emerge because alternate institutional forms better suit the preferences of the actors in power. In short, RI scholars apply a logic of instrumentality (Campbell 2004: 11). Applied to the Italian case, this rationalist logic may become relevant when demographic or economic developments change the opportunity structure for the actors in power. By contrast, sociological institutionalists argue that institutions change not because new institutional forms are more efficient but rather because the latter are widely valued within a broader cultural environment (Hall and Taylor 1996: 949). This constructivist approach follows a logic of appropriateness (March and Olsen 2004). Here, incentives for change might arise as the spread of regional authority becomes more and more common among Western democracies (Hooghe et al. 2008). This will be discussed in detail below with regards to the impact of European Integration. HI scholars take both strategic calculus and constructivist considerations into account. Here, change is explained with reference to critical junctures in the past which path dependently influence processes of institutional creation and reform. In other words: Rejecting a functionalist view of institutions, historical institutionalists see institutions as enduring legacies of political struggles (Thelen 1999: 388). Actions are explained with reference to both a logic of instrumentality and a logic of appropriateness (Campbell 2004: 11). HI scholars usually derive their hypotheses inductively and often working with mid-range theories, a fact that makes generalizations across cases difficult if not impossible. Applied to the Italian case, such an approach might stress Italys history of city states (Putnam 1993; Ziblatt 2006).

The Structural Rationale There is considerable theoretical and empirical evidence that levels of economic development have a positive effect on fiscal decentralization (Letelier 2005). The explanations for this suggested by the relevant literature do differ. Yet, the main rationalist argument goes as follows. At higher levels of economic development, there will be an increasing demand for high quality public services (Wheare 1964). Here, Economists focus on the increased demand for certain expensive public goods such as highways (e.g. Pryor 1967). A more political argument is that granting authority to subunits will allow policy makers to take decisions closer to the citizens, thereby potentially improving the quality of public services.8 In addition, increased competition between different jurisdictions (Tiebout 1956) could lead to innovative public services, as new policies could be tested in some jurisdictions first (Rose-Ackerman 1980). Further, since economic development is highly correlated with higher levels of education, decentralization would allow central governments to respond to the increasing demands for greater government accountability, put forward by more enlightened citizens (Seabright 1995). The quantitative empirical evidence for the positive impact of higher levels of economic development on fiscal decentralization hypothesis is solid (e.g. Oates 1992; Panizza 1999; Stegarescu 2009). Yet, quantitative studies cannot control if the rationale manifests itself as theorized, since they use a rather crude measure, namely GDP per capita. By contrast, a qualitative research design is able to assess which, if any, of the above mentioned rationales played a role in the political process of fiscal decentralization. Another potentially relevant structural factor is government expenditure, which is expected to be positively correlated with fiscal decentralization as well. Partly because, for the reasons outline above, government expenditure should be correlated with higher levels of economic development. Yet, government expenditure is more than merely a proxy for economic development. Rodden explains this with a rationalist argument, stating that representatives will seek to externalize the cost of government
8

However, Rodden and Rose-Ackerman argue that this effect is unlikely to happen in practice, mainly because the very smallness of local jurisdictions might increase the likelihood of corruption (1997).

expenditure (2003: 700). Thus, at higher levels of government expenditure, the central government might try to shift the burden (and the blame) of funding government activities to their subunits. This effect should mainly apply to the dimension of fiscal autonomy. More precisely, the central government might still prefer to have control over programmatic autonomy, because public spending can be used, for example, to buy the loyalty of voters in would-be breakaway regions (Garrett and Rodden 2003: 95). Decentralization scholars agree that domestic preference heterogeneity is an important determinant of regionalization. Three distinct kinds of cleavages rank prominently in the literature, namely ethnic (e.g. Hooghe et al. 2008; Panizza 1999), linguistic (Stegarescu 2009) and economic (Lessmann 2009) cleavages. Ethnic cleavages seem to be particularly important in tribal or other forms of traditional societies. Yet, in advanced industrialized democracies such as Italy, the language criterion is more appealing than the one related to ethnies. As Ernest Gellner put it, the minimal requirement [] of a modern community is literacy (1994 [1964]: 55). Language determines therefore the minimal political unit in the modern world (ibid: 56). Consequentially, linguistic cleavages should be positively correlated with increases in both fiscal and programmatic autonomy. Regional economic disparities seem to be highly relevant too, as comparatively rich regions might not be willing to finance less prosperous regions. Thus, the higher the cleavages, the greater should be the demand of wealthy subunits for decentralization. Poorer regions, on the contrary, should oppose this. Therefore, political factors are expected to interact with structural factors. Large-N studies have problems dealing with all three kinds of cleavages as they are hard to measure. They usually work with rather crude measures such as the ethnic fractionalization index (Fearon 2003), the Linguistic Diversity Index (Lewis 2009), or the GINI coefficient. A qualitative research design is better suited to analyze whether these cleavages are territorially concentrated and politically organized (see below).

The Political Rationale One political factor that merits attention is party ideology as there is considerable empirical evidence for the thesis that Left and Democratic parties in government increase social welfare spending (Huber and Stephens 2001; Swank 2002). Keeping in mind that welfare programs are usually administered at the central level, these parties can be expected to have little incentive to decentralize. Therefore, the presence of Left and Christian Democratic parties in government should be associated with decreases in degrees of fiscal autonomy. By contrast, the effect of party ideology on programmatic autonomy is hard to predict. This rather rationalist logic can be complemented by a more constructivist argument. Since Left parties identify themselves with the idea of solidarity and equality, decentralization may be regarded as inappropriate. Yet, the existing literature pays more attention to a rationalist logic of parties trying to increase electoral success (Jametti and Joanis 2010; Mazzoleni 2009). Here, ideology is largely ignored. Parties are driven by strategic considerations. More precisely, parties who have a strong territorial basis will be interested in decentralization.

The International Rationale Since Gourevitchs seminal article The Second Level Reversed: The International Sources of Domestic Politics (1978), international level factors have attained increased attention among Comparativists. Applied to this papers topic, the potential impact of Globalization on fiscal decentralization is a major subject of debate. One strand of the literature believes that Globalization is associated with increases in decentralization. Alesina and Spolaore, for example, posit that a breakup of nations is more costly if it implies more trade barriers and smaller markets (1997: 1041). Globalization and the associated diffusion of free trade regimes would reduce these costs and therefore make secessionist threats more credible. In order to avoid secession, central governments in heterogeneous countries might therefore

10

decide to concede both fiscal and programmatic autonomy to subunits. Stegarescus study of the effects of economic and political integration on fiscal decentralization provides empirical evidence for this interaction effect (2009). However, other scholars argue that Globalization leads to centralization of fiscal autonomy, while simultaneously enhancing the political autonomy [of subunits] (Garrett and Rodden 2003: 92). This argument is based on the assumption that larger fiscal units are better at risk sharing. Drawing from the vast International Political Economy literature on the implications of economic Globalization, they emphasize that Globalization increase[s] volatility and hence aggregate[s] economic risk (ibid). Following this logic, pooling economic resources by increasing the centers fiscal autonomy, would provide an insurance both citizens and regional officials might be interested in. Another related international phenomenon that has gained the attention of decentralization scholars is Europeanization. As Oates points out, it is striking to witness in the European Community the moves toward devolution in many member countries, while, at the same time, the Community develops a set of supranational institutions for governance and economic management (1999: 1145). Accordingly, both Hooghe et altri (2008) and Stegarescu (2009) observe a decentralizing effect of European integration. Yet, the results of the former scholars analysis need to be treated with caution as they are based on descriptive statistics and do not include relevant controls for preference heterogeneity. Stegarescus results, by contrast, are based on advanced econometric analysis. The interpretation of a possible impact of Europeanization is therefore not straightforward. On the one hand, the EU does indeed foster heterogeneity within its member states, exemplified by the EUs motto United in Diversity. This manifests itself in the various structural funds for which only regions are eligible. Yet, Europeanization does share many characteristics with Globalization, too. More precisely, the combination of Maastricht and the euro render Europe the closest example we have of a completely open international market (Garret and Rodden 2003: 106). Therefore, Europeanization

11

could well be associated with increased demands for risk sharing. Hence, central government officials in homogenous EU member states may decide to centralize their political systems. Yet, strong preference heterogeneity within a country, as it is arguably the case for Italy, could lead to the opposite outcome.

5. Methodology Italy is a particular promising subject of analysis as it fulfils at least seven causal factors that are supposed to trigger fiscal decentralization. First, it has a history of regionalism (Putnam 1993; Sabetti 2000; Ziblatt 2006). Second, levels of economic development are among the highest in Europe. Third, a public deficit of more than 100% of its GDP has been accumulated over recent decades, heavily constraining government expenditure, Fourth, Italy has the third highest linguistic diversity score of all OECD member states (Lewis 2009).9 Fifth, huge economic disparities exist between North and South. Sixth, a strong regionalist if not secessionist party (Lega Nord) enjoys popular support. Last but not least, Italy is a founding member of the European Union. Not surprisingly, it has decentralized considerably over the last decades. This makes Italy an excellent subject for our case study, as it allows to analyze if the rationale suggested by the literature can indeed explain decentralization or if idiosyncratic Italian factors are more important. The time period of this study comprises Italys Republican History (1948 to present), although some references to earlier historical events will be made if these should seem to be critical junctures. In general, since fiscal decentralization did majorly occur in the last two decades, the analysis will mainly focus on these years. Fiscal Decentralization will be assessed according to the criteria developed in section three, distinguishing between programmatic and fiscal autonomy.

The Linguistic Diversity Index (LDI) expresses the probability that any two people of a given country selected at random would have different mother tongues and is measured on scale ranging from 0 (everybody speaks the same language) to 1 (everybody speaks another language). Italy has a value of 0.586. Among OECD countries only Canada (0.599) and Belgium (0.747) are linguistically more diverse.

12

When interpreting the rationales underlying the institutional reform processes, particular attention will be paid to identifying the underlying logic of change, as theorized in section four. Analyzing which logic of change has the greatest explanatory power, at least in the Italian case, would be another important contribution to the literature because few authors confront the different approaches systematically.10 Yet, it may well be the case all the different approaches reveal important causal mechanisms, a finding that would be in line with Hall and Taylors conclusions that each has something to learn from the others (1996: 955).

6. Categorizing the Italian Political System The Postwar Settlement Although federal ideas have a long history in Italy (Putnam 1993; Sabetti 2000; Ziblatt 2006), the 1948 Constitution created a fairly centralized unitary state, dividing the country into 15 regioni a statuto ordinario (ordinary status regions). Although there were provisions for granting some minor powers to them, these were not implemented until the 1970s. However, the solution was highly asymmetrical because, in addition, regioni a statuto speciale (special status regions) were created. Sicily, Sardinia, Val dAosta, Trentino Alto Adige - and eventually also Friuli-Venezia Giulia (in 1963) were granted some powers in relation to legislation, administration and finance. Most importantly, they were allowed to keep large parts of their levied taxes. In short, with regards to the ordinary regions, the constitution of 1948 created a polity that approximated the ideal type of fiscal unitarism (see table 1). Yet, considerable powers in terms of both programmatic and fiscal autonomy were granted to the special regions. In the early 1970s, the institutions of the fifteen ordinary status regions were finally set up, each with a directly elected regional council and an executive responsible to it. In 1977, they gained some more programmatic autonomy in urban planning, regional development, urban and rural policing,

10

For a notable exception see Kelemen (2003).

13

health and hospital assistance, education and culture, communications, environment, and craft industry. In addition, regioni were granted the authority to exercise some direct administrative control over local government (Hooghe et al 2008: 196). Yet, since these regions had still no financial autonomy at all, Italy ignoring the special regions - still approximated the ideal type of fiscal unitarism fairly well.

Reforms in the 1990s Regionalization was considerably deepened after the collapse of the First Republic in the early 1990s. In the late 1990s, the so called Bassanini laws transferred more competencies to the regions. More precisely, all administrative responsibilities confined to the regions territories (the so called residual administrative powers) were now assigned to them, only excluding competencies dealing with major infrastructure projects. In short, the Bassanini reforms increased the administrative capacity for regions in the design and allocation of public services (Keating and Wilson 2010; Roux 2008: 328-329). This can be seen as an important step towards greater programmatic autonomy. In addition, first steps towards regional fiscal authority were taken, as the Bindi reforms attributed health care management to the regions. Associated with these reforms was the establishment of a new system of regional finance set up in 2000. It assigned certain tax revenues to the regions. Here, of particular importance was the fact that the revenues generated by a new business tax (IRAP), remained with the regions. Yet, this was far from approximating the ideal type of fiscal federalism: not only did regions continue to rely on transfers payments and proceeds from state income taxes for the majority of their public budget, they didnt even have the competence to set the tax rate, not to speak of creating new tax bases (Keating and Wilson 2010; Mazzoleni 2009: 141). Further, reforms in 1999 and 2001 introduced direct elections of regional presidents in ordinary and special status regions (except for Val dAosta and the Bolzano province). This is important in terms of programmatic autonomy as it arguably strengthens regional identity and allows for a stronger

14

bargaining position of regional leaders, as they enjoy direct legitimacy from their respective population (Roux 2008: 328-329). To sum up, both programmatic and fiscal autonomy of regions were increased in the 1990s. Since the fiscal authority of subunits was still quite limited, the Italian State did not really approximate the ideal type of fiscal devolution. However, it became less unitarist than before.

The 2001 Constitutional Reform The most important reform for the purpose of our study is the constitutional reform of 2001, consolidating the principle of residual powers and extending it to legislative competencies concurrent with the central government in international and EU relations, foreign trade, job protection and industrial safety, education, scientific research, health, food, sport, civil protection, town planning, ports and airports, cultural and environmental resources, transport and energy (Hooghe et al. 2008: 196). From then on, the central government does no longer have the power to suspend regional legislation; disputes between regions and the central government are now settled by the constitutional court. The competencies of ordinary status regions have become more and more similar to those of the special status regions. Although, the term federalism is not mentioned in the legislation, some observers conclude that the reform created a quasi-federal state (ibid).

Recent Developments In 2005, another constitutional reform which would have shifted significant authority on health and education to the regions is rejected in a popular referendum. In 2009, Parliament agrees on the principles for implementing fiscal federalism, approving a rather vague framework legislation. Currently, Italys minister for federalism, Roberto Calderoli (Lega Nord), is working on a reform that would move Italy closer towards the ideal of fiscal federalism (Calderoli 2010).

15

Crucial Aspects of fiscal federalism that are still not fulfilled are related to criteria 1 and 3 presented in the conceptual part. More precisely, especially southern regions are continuously bailed out by the central government because they cannot fund the public health care system with their own limited revenues (Keating and Wilson 2010). In addition, most competencies for the regions are established by ordinary legislation. The central government could therefore take away these powers from the regions by simple majority. The lack of institutionalization may be exemplified by the fact that Italy does not have a territorial second chamber. Although two legislative chambers were created in 1948, namely the House of Deputies and the Senate, the latter is no territorial chamber at all. By contrast, this so called bicameralismo perfetto gives both chambers almost identical law making powers. In addition, they are also elected in a similar manner, thus senators did and do not represent the interests of specific regions (Guarnieri 2007).

7. Explaining Institutional Changes in the Italian Republic Institutional Analysis Rational Choice Institutionalism helps explain why all attempts to make the Senate a true regional chamber have failed so far. Obviously, senators are not interested in seeing their chamber abolished or its composition significantly changed, as they might not get re-elected (Amoretti 2002). Yet, direct elections have led to the development of new regional elites which might change the rules of the institutional game in the long run (Fabbrini and Brunazzo 2003: 115). However, the regional elites are not well represented at the central level as the only channel of influence consists of the State-Regions Standing Committee (Amoretti 2002: 135). This is supposed to coordinate central and regional policies but the role of the regions is strictly subordinated to the central governments one. More precisely, the agenda is fixed and determined by the central government (Keating and Wilson 2010). Thus, from a rationalist perspective, the fact that regions lack representation at the center is a factor potentially

16

limiting further decentralization, as the existing institutional structures provide few opportunities for regional actors to make their voices heard in Rome. Yet, the new regional elites, legitimated by direct elections, might make decentralization seem an appropriate choice for the general public and central government officials. The reform regarding the role of Italys Constitutional Court, mentioned in the preceding section, is important as well as it changes the power relations between regions and center. The 2001 constitutional reform makes the Court the arbitrator in disputes between different tiers. Yet, there are reasons to assume that the Court is more likely to favour the central level. More precisely, the regions have no say in the make-up of the Court (Amoretti 2002: 136).

Structural Analysis As Fabbrini and Brunazzo note, the central [Italian] state was forced to acknowledge that the exclusively centralized management of its economic and social responsibilities was inefficient and uneconomical (2010: 104). This is perfectly in line with the rationalist hypothesis theorized in section four, stating that at higher levels of economic development states face increasing utility maximising incentives to decentralize their political systems. Although regions enjoyed only minor programmatic autonomy before the reforms of the 1990s, the fact that ordinary regions enjoyed some spending authority but did not have any kind of fiscal responsibility before the Bindi reforms, had disastrous consequences for the levels of public debt. Roux makes clear that this split between revenues generated by the central government and spending partly controlled by subunits, institutionalized political irresponsibility in the budgetary process (2008). In addition, Italys government expenditure was already very high because of the so called grow the pie strategy (see political analysis below).

17

Amoretti points to the multifaceted nature of Italian cleavages, rooted in a history of territorial fragmentation and overlapping layers of cultural, social, economic, and political differentiation (2002: 129). A historical perspective would stress that, between the fall of the Roman Empire in the fifth century and Italian Unification in 1861, Italy was hardly more than a geographical expression (ibid). This resulted in a diverse array of historical experiences (ibid) which could not easily be overcome by imposing a centralist polity from above (Ziblatt 2006). Economically, the north Italian city states were far ahead at the time of unification and this holds true until today as the best-known single divide separates the urbanized, economically developed North from the poorer, more rusticated, and less fully modernized South (Amoretti 2002: 129-30). Hine outlines that the economic disparities between North and South make fiscal federalism less likely because it could only be achieved at the cost of establishing a massive system of intergovernmental transfers (1996: 125). Yet, such a system might turn the promises of fiscal federalism into perils (Rodden 2006; Sartori 2010). During the negotiations about the Italian constitution of 1948 regionalist demands emerged in the bi-lingual border regions of northern Italy (Keating and Wilson 2010). These regions became autonomous in order to take into account linguistic and cultural differences, such as the linguistic minorities in Trentino-Alto Adige/Sdtirol, Val dAosta, and Friuli-Venezia Giulia, or geographic isolation in the case of the two greater islands, Sicily and Sardinia (Lepschy et al. 1996). Thus, the linguistic cleavage dimension did play an important role for the establishment of the special status regions, precisely as hypothesized. Yet, with regards to the demands for decentralization in the 1990s and 2000s, linguistic variety seems to be a less salient issue.

18

Political Analysis In 1948 federalism did not rank prominently on the agenda of the members of the Contituent Assembly because the crucial cleavage was between pro-Soviet and pro-Western forces. Thus, in the postwar years federalism was not crucial in an ideological sense. Obviously, the many cleavages outlined above did bring about heterogeneous demands. However, these were not addressed in an exclusive fashion but by the favourite grow the pie strategy (Amoretti 2002: 133). Namely, instead of privileging one region over the other when distributing revenues, the overall amount of spending was increased. Not surprisingly, Italys public debt became the highest in Western Europe. Yet, there were ideological positions with regards to decentralization that can be attributed to party lines. Originally, the Christian Democratic DC was in favour of decentralization since this was in line with the Catholic doctrine of subsidiarity (Mazzoleni 2009: 136). By contrast, the Italian communist party (PCI) strictly opposed it as Marxism calls for economic development through centralization (ibid). Over time these positions changed, however. Since the PCI was continuously excluded from the governing coalition, though enjoying strong electoral support, decentralization was seen as an alternative strategy that would allow the PCI to take over some of the so called red regions, particularly in central Italy. For this very reason the DC then strongly opposed decentralization, a fact that can explain why the provisions regarding regional institutions were only implemented with a delay of 22 years (Amoretti 2002: 133). Maybe the most important partisan actor is the Lega Nord. Founded in 1991 as a gathering of several Northern regional parties, it was successful in activating the territorial cleavage mentioned above (Roux 2008: 330). The Northern Leagues interest in federalism (or, in its most accentuated from, secessionism) can be explained by applying a logic instrumentality and one of appropriateness. The former can be related to the economic cleavages between North and South outlined above. In line with our hypothesis from section four, the comparatively wealthy regions want to keep as much of their tax

19

money as possible. This is nicely captured by the Legas slogan Roma la Drona (Rome the thief). However, increasing regional authority would also help the Northern League to construct a northern identity (Huysseunne 2006). Thus, fiscal decentralization might be useful for achieving their ultimate goal of secession. But at the same time, federal reforms threaten the Legas agenda; because if most demands are already satisfied by a federal settlement, it might be harder to mobilize the public for the secessionist policies. This explains why the Lega strongly opposed the 2001 reform and campaigned for a no vote in the referendum, blaming the Left of promoting False Federalism. Apparently, the Lega feared that incremental steps towards a moderate form of fiscal federalism would harm their cause (Keating and Wilson 2010). The role of the centre-left is interesting as it apparently contradicts the hypothesis of our theoretical part. Instead of opposing reforms as hypothesized, it has initiated all important reforms in the 1990s. Sometimes they did this with cross-partisan consensus, but the most important constitutional reform of 2001 was made against the explicit opposition of both the Lega and Berlusconis Forza Italia. Roux explains the Lefts support of federalism as a strategy set in place to limit the electoral support of the centre-right in northern regions (2008: 332). Apparently, a rationalist vote maximizing strategy explains this policy better than an ideological argument which would make centralist policies seem more appropriate for parties interested in providing equal standards of living. However, applying a logic of appropriateness shows how problematic this strategy may turn out to be. As Amoretti notes, the traditional concern of the left is economic and social equality, a goal that is easier to pursue under a unitary as opposed to a federal government (2002: 137). The failed reform of 2005, which was initiated by the centre-right coalition, including the Lega, was a highly contradictory reform package. Because parts of the governing coalitions, especially the Christian Democratic Centrist party (UDC), were strictly against decentralization, the supposedly federalizing reform even contained centralizing measures. For example, a clause was added that would

20

have allowed the central government to over-rule regional laws if this was in the national interest (Keating and Wilson 2010). Most importantly, programmatic decentralization was still not complemented by the necessary fiscal autonomy to finance the spending. In other words, the perils of fiscal federalism were, again, not addressed (Rodden 2006). This is hardly surprising as for the reform to find a majority in the parliament it needed to find the support of southern parliamentarians as well. Naturally, the tax-absorbing South opposed significant decentralization of financial authority. The reaction of the centre-left coalition to this reform package is in so far interesting as it allowed them to get back to the ideological position this paper originally hypothesized. Namely, the reform was critiqued as a divisive threat to equality, citizenship and standards (Keating and Wilson 2010). The results of the failed referendum merit some attention because the yes vote actually won everywhere in the North except in the red Emilia Romagna. The economic regional disparities seem to have played a role here (Keating and Wilson 2010). To conclude, granting more fiscal and programmatic autonomy to Italian regions is a partisan issue. While the Left was and possibly still is generally willing to grant more programmatic autonomy to subunits, it strictly opposes true fiscal federalism for ideological reasons. The only party that is clearly in favour of full-fledged fiscal federalism, satisfying the no bail-out criterion and limiting intergovernmental grants as much as possible, is the Lega Nord. Yet, since it has to share the power with Berlusconis Popolo della Liberta, a party that certainly does not want to put its southern vote base at risk, only further programmatic decentralization seems feasible. The observation that with regards to the partisan rationale, consensus on the programmatic dimension is not too hard to reach, while reforms dealing with the fiscal autonomy dimension are considerably more conflictual, seems to represent one of the key perils of fiscal federalism in practice (Sartori 2010). Both a rationalist and a constructivist logic need to be considered in order to fully understand party agendas. On the one hand, parties seem to follow rationalist vote-share maximizing strategies.

21

This explains, for example, the decision of the centre-left coalition to push forward its own federalizing reform in the late 1990s, although it conflicted with the coalitions solidaristic identity. On the other hand, the rationalist logic of vote-maximization cannot really be separated from the constructivist politics of identity, because the identity of a party significantly constrains its strategic options.

22

International Analysis This paper cannot test the hypothesis that globalization is associated with increased demands for risk sharing and therefore centralization because fiscal decentralization in Italy is not yet developed far enough to create these demands. Yet, the influence of Europe can be assessed quite well. Fabbrini and Brunazzo emphasize two main ways the EU has influenced fiscal decentralization in Italy. First, beginning with the 1980s, EU cohesion policies directly encouraged the more robust activation of territorial units (2003: 105). This point needs some explanation. Italy traditionally pursued a top-down regional policy, mainly through the Cassa per il Mezzogiorno (CASMEZ). More precisely, until the 1990s, the central government had tried to develop the South centrally from Rome, for example by financing infrastructural projects that were supposed to bring about trickle-down effects. These development project were notoriously unsuccessful, and most infrastructural projects eventually failed, leaving cattedrali nel deserto (cathedrals in the desert). Yet, applying a rationalist perspective, this programmatic authority in the hands of central government officials provided an important power base for the centre which could be used for vote maximizing strategies. EU cohesion policy, by contrast, is basically a bottom-up policy. Projects are expected to be developed locally and if they fulfil the EU criteria, they will receive funding directly from the European Union. Since this policy proved to be much more effetive than the original Italian approach, the CASMEZ was eventually dissolved and replaced by new tool of negotiated planning, in line with EU cohesion policy principles (Fabbrini and Brunazzo 2003: 109). However, this significantly increased programmatic autonomy of the subunits and weakened the center. Yet, one does not need to see cohesion policies through rationalist lenses, a more constructivist perspective might also do. Following such a logic, the emphasis EU cohesion policy places on regional actors, encouraging them to develop independent policies tailored to the specific needs of the respective subunit, might increase the legitimacy of regionalism (Fabbrini and Brunazzo 2003: 113; Roux 2008: 329).

23

The second EU effect is more indirect but nonetheless highly relevant. The provisions of the Treaty of Maastricht, signed in 1991, uncovered the inefficiencies of the central Italian government. This was mainly due to the new standards the EU established in order to achieve a Monetary Union. The stability and growth pact and the convergence criteria revealed that Italy was close to running bankrupt. In order to fulfil these criteria and be able to join the Euro, the central government had to cut down expenditure. In other words, the notorious increasing the pie policies were no longer feasible (Amoretti 2002: 135). This again weakened the centers power vis-a-vis the regions because the former was no longer able to buy the political support of the latter. At the same time, this made the regional disparities between the net-taxpaying North and the tax-absorbing South more visible (Roux 2008). Lastly, it is important to note that the EU potentially interacts with the political and structural rationale outlined above. Politically, this applies at least to the Lega. Only because the EU guarantees a free market within the European Union the Lega can reasonably campaign for an independent northern state of Padania. Structurally, the fact that Italy is a fairly heterogeneous country seems to be a necessary condition without which the opportunity structure provided by the EU would not trigger domestic changes (Fabbrini and Brunazzo 2003: 118).

24

8. Conclusion Not surprisingly, this paper has painted a more nuanced picture of the rationale underlying fiscal decentralization than those commonly revealed by quantitative studies. Particularly, political factors that are underrepresented in quantitative studies prove to be essential for explaining institutional reforms in Italy. More precisely, the analysis of the political rationale has shown that the causes for decentralization in terms of programmatic autonomy diverge from the ones underlying decentralization of fiscal autonomy. While programmatic decentralization can find the consensus of parties from all political spectrums, significant decentralization in terms of fiscal autonomy is more conflictual. From a rationalist perspective, it decreases the power of the center to buy electoral support. Following a constructivist logic of appropriateness, it potentially undermines the solidaristic identity of Left and Christian democratic parties. The finding that an agreement about increasing programmatic autonomy of subunits is reached more easily than one aiming at decentralizing fiscal authority has some problematic normative implications. As Rodden has outlined in considerable detail (2006), for fiscal decentralization to bring about the desired utility maximizing effects, decentralization needs to take place on both dimensions. Yet, in the Italian case, fiscal reforms are quite limited, as the main reforms were centred on the dimension of programmatic autonomy. Therefore, the perils of fiscal federalism seem to have

manifested themselves in the Italian case yet again, as the fiscal aspect is mainly rhetorical in nature (Calderoli 2010; Sartori 2010). Generally, it has been shown that structural factors are important, but they are only associated with the hypothesized outcomes if the opportunity structures they provide are exploited by political actors. While members of the Italian Constituent Assembly were responsive to the linguistic and cultural cleavages which found expression in the assignment of specific programmatic and fiscal authority to the so called special status regions, the economic cleavages between North and South seem to have

25

attracted political attention in recent years only. Yet, since the latter disparities trigger conflicting demands - fiscal decentralization from the net-taxpaying North as opposed to continued centralized redistribution from the tax-absorbing South - the only feasible agreement was related to programmatic autonomy. Changes in fiscal autonomy remained rather marginal and it seems not very likely that the current minister of federalism, Roberto Calderoli, will be able to put a new bill through that would bring true fiscal federalism. An international level factor that could be shown to interact considerably with domestic factors is European Integration. In the Italian case, domestic actors have been willing and able to exploit the new opportunity structures that evolved due to the process of European Integration (Fabbrini and Brunazzo 2003). This interaction effect is in line with the one found by recent quantitative work (Stegarescu 2009) and confirms the insights of the Europeanization literature (Risse et al. 2001). Last but not least, this papers analysis suggests that both a logic of appropriateness and a logic of instrumentality can help explain decentralizing reforms, at least in the Italian case. This is in line with Hall and Taylor (1996) who expected that the two contrasting logics of change complement one another.

26

Bibliography Alesina, Alberto, and Enrico Spolaore (1997): On the Number and Size of Nations, The Quarterly Journal of Economics 112 (4): 1027-56. Amoretti, Ugo (2002): Italy Decentralizes: A New Look at Federalism, Journal of Democracy 13 (2): 126-140. Arzaghi, Mohammad, and Vernon Henderson (2005). Why Countries are Fiscally Decentralizing, Journal of Public Economics 89: 1157-1189. Calderoli, Roberto (2010): Trasferire le funzioni non basta : il momento del federalismo fiscale, Corriere della Sera, April 24th,2010. Campbell, John (2004): Institutional change and globalization, Princeton: Princeton University Press. Fabbrini, Sergio, and Marco Brunazzo (2003): Federalizing Italy: The Convergent Effects of Europeanization and Domestic Mobilization, Regional & Federal Studies, 13(1): 100-120. Fearon, James (2003): Ethnic and Cultural Diversity by Country, Journal of Economic Growth 8 (2): 195222. Garrett, Geoffrey, and Jonathan Rodden: Globalization and Fiscal Decentralization, in Miles Kahler and David Lake (eds.): Governance in a Global Economy: Political Authority in Transition. Princeton University Press: 87-109. Gellner, Ernest (1994 [1964]): Nationalism and Modernization, in John Hutchinson and Anthony Smith (eds.): Nationalism, Oxford: Oxford University Press: 55-63. Gourevitch, Peter (1978): The Second Image Reversed: The International Sources of Domestic Politics, International Organization 32 (4): 881-911. Guarnieri, Carlo (2007): Il sistema politico italiano, Bologna : Il Mulino. Hine, David: Federalism, Regionalism, and the Unitar State, in Carl Levy (ed.): Italian Regionalism: History, Identity and Politics, Oxford: Berg. Hooghe, Liesbet, Gary Marks, and Arjan H. Schakel (2008): Regional Authority in 42 Democracies, 19502006, Regional and Federal Studies 18 (2-3): 111-302. Huber, Evelyne, and John Stephens (2001): Development and Crisis of the Welfare State: Parties and Policies in Global Markets. Chicago: University of Chicago Press. Huysseune, Michel (2006): Modernity and Secession: The Social Sciences and the Poltical Discourse of the Lega Nord in Italy, New York: Berghahn Books.

27

Jametti, Mario, and Marcelin Joanis (2010): Determinants of fiscal decentralization: political economy aspects, Working Papers 2010/7, Institut d'Economia de Barcelona (IEB). Keating, Michael, and Alex Wilson (2010): Federalism and Decentralization in Italy, PSA Conference, Edinburgh, March-April 2010. Kelemen, Daniel (2003): The Structure and Dynamics of EU Federalism, Comparative Political Studies 36(1-2): 184-208. Lepschy, Anna Laura, Giulio Lepschy, and Miriam Voghera (1996): Linguistic Variety in Italy, in Carl Levy (ed.): Italian Regionalism: History, Identity and Politics, Oxford: Berg. Lessmann, Christian (2009): Fiscal Decentralization and Regional Disparity: Evidence from Cross-Section and Panel Data, Environment and Planning 41(10), 2455-2473 Letelier, Leonardo (2005): Explaining Fiscal Decentralization, Public Finance Review 33: 155-183. Lewis, M. Paul (ed., 2009): Ethnologue: Languages of the World, Dallas: SIL International, Online version: http://www.ethnologue.com [last access: April 19th, 2010]. March, James, and Johan P. Olsen (2006): The Logic of Appropriateness, in Michael Moran, Martin Rein, and Robert E. Goodin (eds.): The Oxford Handbook of Public Policy, Oxford: Oxford University Press, 2006. Mazzolini, Martino (2009): The Italian regionalization: a story of partisan logics, Modern Italy 14 (2): 135-150. Meadwell, Hudson(2009): The Political Dynamics of Secession and Institutional Accommodation, Regional & Federal Studies 19 (2): 221-235. Oates, Wallace (1972): Fiscal federalism, New York: Harcourt Brace. Oates, Wallace (1999): An Essay on Fiscal Federalism Journal of Economic Literature 37 (3): 1120-49. Osaghae, Eghosa (1990): A Reassessment of Federalism as a Degree of Decentralization, Publius 20. Panizza, Ugo (1999): On the determinants of fiscal centralization: theory, and evidence, Journal of Public Economics, 74(1): 97-139. Pryor, Frederic (1967): Elements of a positive theory of public expenditures Finanzarchiv 26 (3): 40530. Putnam, Robert (1993): Making Democracy Work: Civic Traditions in Modern Italy, Princeton: Princeton University Press. Riker, William H. (1964): Federalism: Origin, Operation, Significance, Boston: Little, Brown.

28

Risse, Thomas, Maria Green Cowles and James Caporaso (2001): Europeanization and Domestic Change: Introduction, in Maria Green Cowles, James Caporaso and Thomas Risse (eds.): Transforming Europe: Europeanization and Domestic Change, Ithaka (Cornell University Press): 1-20. Rodden, Jonathan (2003): Reviving Leviathan: Fiscal Federalism and the Growth of Government, International Organization 57 (4): 695-729. Rodden, Jonathan (2004): Comparative Federalism and Decentralization: On Meaning and Measurement Comparative Politics 36 (4): 481-500. Rodden, Jonathan (2006): Hamiltons Paradox: The Promise and Peril of Fiscal Federalism. Cambridge: Cambridge University Press. Rodden, Jonathan, and Susan Rose-Ackerman (1997): Does Federalism Preserve. Markets?, Virginia Law Review 83, 7: 1521-72. Rose-Ackerman, Susan (1980): Risk taking and reelection: Does federalism promote innovation? Journal of Legal Studies 9 (3): 593-616. Roux, Christophe (2008): Italys path to federalism. Origins and paradoxes, Journal of Modern Italian Studies 13(3): 325339. Sabetti, Filippo (2000): The Search for Good Government: Understanding the Paradox of Italian Democracy, Montreal: McGill-Queens University Press. Sartori, Giovanni (2010): Il federalismo e il mistero del silenzio tombale, Corriere della Sera, April 21st, 2010. Seabright, Paul (1995): Accountability and decentralisation in government: An incomplete contracts model, European Economic Review 40 (1): 61-89. Sorens, Jason (2010): The Institutions of Fiscal Federalism, Publius (forthcoming). Stegarescu, Dan (2005): Public Sector Decentralisation: Measurement Concepts and Recent International Trends, Fiscal Studies 26: 301-333. Stegarescu, Dan (2009): The effects of economic and political integration on fiscal decentralization: evidence from OECD countries, Canadian Journal of Economics 42(2): 694-718. Swank, Duane (2002): Global Capital, Political Institutions and Policy Change in Developed Welfare States. Cambridge: Cambridge University Press. Thelen, Kathleen (1999): Historical Institutionalism in Comparative Politics, Annual Review of Political Science 1999(2): 369404. Tiebout, Charles (1956): A Pure Theory of Local Expenditures, Journal of Political Economy 64: 416-24.

29

Weingast, Barry R. (1995): The Economic Role of Political Institutions: Market-Preserving Federalism and Economic Development, Journal of Law, Economics, and Organization 11 (1): 1-31. Wheare, Kenneth (1964): Federal government, Oxford: Oxford University Press. Ziblatt, Daniel (2006): Structuring the State: The Formation of Italy and Germany and the Puzzle of Federalism, Princeton: Princeton University Press.

30