Fiscal decentralization and inequality: the case of Spain

Abstract What are the distributional effects of power allocations between different levels of government in a country? This article examines whether having regions with more fiscal authority is significantly associated with higher national economic inequality in a country. The literature has long established the positive link between decentralized governance structures and varying levels of redistribution within a country. Redistribution is channel through which governments tackle inequality, and if redistribution is asymmetric across sub-national units, this is expected to increase inequality. Yet, there is a gap in the literature that systematically links these three components: decentralized governance, redistribution and economic inequality. By focusing on the fiscal aspect of decentralization, this article provides the theory and the causal mechanism for understanding why the decentralization levels of governance in a country should matter for income inequality. Using data measured along different components of regional authority, such as policy scope, fiscal autonomy and fiscal control, it explains why fiscal decentralization is expected to increase inequality, while predicting that the co-sharing of fiscal power between the regions and central government will to lead to lower inequality. The article tests this hypothesis with a case study on Spain.

conducted in 2013, supplemented by extant literature. The theory is tested with a statistical analysis using RAI scores in 17 European Organisation for Economic Co-operation and Development (OECD) countries between 1960 and 2006, using a fixed-effects model. This research idea is inspired by Spain, which is an illustrative case for the argument because (1) it has one of the highest levels of interpersonal income inequality in Europe and (2) it is a highly decentralized country, yet the fiscal system carries both centralized and decentralized characteristics. Spain provides an interesting example from which lessons can be learned for countries that use different types of fiscal equalization instruments and for regional studies in general.

LITERATURE, THEORY AND METHODOLOGY
In the political economy literature, federalism and decentralization have been associated with less and more uneven redistribution, with limited empirical studies (Rodden, 2006;Treisman, 2007;Wildavsky, 1984). Rodden (2004) states that federalism is a negative factor for the growth of redistribution and the welfare state, but also argues that empirical support for this argument is limited to research on 15 OECD countries.
Several studies refer to the positive link between decentralization levels and higher inequality through the channel of the welfare state (Boix, 2003;Glaeser, 2006;Reuveny & Li, 2003). The concept of 'fiscal federalism' provides a channel to link the territorial governance structure with inequality. Oates (1972) claims that if fiscal systems are very decentralized, it enables the existence of various communities that support different levels of redistribution in a country, which would result in varying redistribution levels and patterns.
This fragmentation attributed to federalism derives from varying decision-making points and through decentralized fiscal structures, which can decide on their own taxation and redistributive policies. These decentralized fiscal structures can also be found in constitutionally unitary states, which are then prone to diverging within country redistribution patterns. Regional governments can carry out their own policies in their jurisdiction, which indicates 'self-rule'; and/or they can co-decide on country-wide policies by sharing the authority with other regional governments and the central government, indicating 'shared rule' (Elazar, 1987). Furthermore, countries that have centrally governed taxation, but still asymmetrically distribute their equalization grants to their regions are prone to varying levels of redistribution as well, as illustrated with the case of Spain.
Based on this literature, for fiscal decentralization/regional authority to be a significant predictor of higher income inequality, the following are assumed. First, (fiscal) decentralization can obstruct the willingness of the central government to implement inequality-reducing policies. If there were high regional authority, regions would have varying redistributive capacities and policies, which would result in varying services and transfers available. Most importantly, (fiscal) decentralization and more regional authority would hinder the capability of the central government to redistribute in a more standardized way through taxes and transfers. Therefore, due to more regional authority, the discrepancies of redistribution should lead to higher overall national economic inequality in a country. Taxes and redistribution are one of the main channels that the welfare states use to target income discrepancies. Higher levels of revenue sharing granted to regions through fiscal autonomy should also have an increasing impact on inequality due to the lessening of the standardization of redistribution. On the contrary, if the sub-national units and the central government interact and decide together on redistributive issues towards more standardized goals, this is expected to be associated with lower income inequality. 1 In short, more fiscal decentralization should be associated with more inequality, and more fiscal shared rule with more equality.
This statistical study used aggregated national scores from RAI -looking at the general sub-national authority levels -as well as general scores of the sub-national units in countries for fiscal autonomy and fiscal control 2 . Fiscal autonomy measures 'the extent to which a regional government can independently tax its population'. Fiscal control measures 'the extent to which regional representatives co-determine the distribution of national tax revenues' (Hooghe et al., 2010, Appendix A). The present study controls for variables that have been shown to impact inequality, such as unemployment, social expenditures, wage coordination and social transfers.
This statistical study shows that more fiscal regional authority (fiscal autonomy) is significantly associated with more overall national income inequality in Europe (Gini inequality), while more co-sharing of fiscal power between regions and the central government (fiscal control) is significantly associated with lower inequality.
This indicates that countries where the regional government sets the rate of more of the major taxes (personal income, corporate, value added or sales tax) are expected to have more inequality. On the other hand, as expected, higher shared rule on taxation might be 'bringing the regions together' and contributing to lower overall income inequality levels because there are channels for the sub-national units to discuss redistributive preferences at the national level.

BACKGROUND OF THE CASE OF SPAIN THE CASE OF 'EVOLVING' FEDERALISM
'Spain evolved to a federal form without the proper coordination mechanism needed for a federal system'. 3 Spain has been called a 'quasi-federation' (Agranoff, 2010) or a 'de facto federal state in all but name' (Moreno & Colino, 2010). With the end of the dictatorship and the new constitution of 1978, Spain transitioned from a unitary to a more federal arrangement with the gradual creation of 17 autonomous communities (ACs). There is a three-tiered governance system in Spain; with the (central) government, the regions (ACs) and the local governments (including provinces and municipalities) (Agranoff, 2010). The regions are not granted direct participation at the central level of government and there is no chamber based on the boundaries of the territories (Encarnacion, 2008). Also, Spain is a highly decentralized country in service delivery, with most of the power of delivery resting in the ACs.
Spain has two types of financing for its ACs: the 'common' system and the 'autonomous' system for the Basque Country and Navarra, pointing out to the asymmetry of the system. The central government retains the main control in the financing of the ACs in the 'common' system. In the common system, the taxes are collected by the central government. At the end of the fiscal year, the central government distributes the capital transfers to the ACs based on population, insularity, age of population and dispersion of municipalities. Transfers from the state amount to 95% of total revenues received by ACs (Beramendi, 2012). In the 'autonomous system', the Basque Country and Navarra can levy, manage, settle and collect their own taxes, and they transfer a fixed quota to the central government. 4 . Furthermore, there is a different economic and tax system for the Canary Islands 'due to historical and geographical reasons' and in line with the European Union's provisions for ultra-peripheral regions. 5 Spain exemplifies the complications of a delicate balance between the self-rule and shared-rule components of Elazar's (1987) federalism concept: 6 The so-called federal governance structure of Spain is of 'evolving' nature prone to changes concerning the power structure between the central government and the ACs. Powers have been devolved to the ACs gradually and some of them were on a 'fast track' (such as Catalonia, Basque Country and Galicia) -on a demand basis -whereas other ACs took more time to get powers in taxes and public services. The 'fast track' devolution was also linked to demands for devolved rule/independence in ACs, such as Catalonia, the first 'pre-regional autonomous body' to be founded in Spain. Beramendi (2012) claims that the granting of powers to the regional governments resulted in a fast pace of the decentralization of expenditures for investments in infrastructure, education and health. He argues: The fiscal flows of public expenditure in general, and social security in particular, generate significant redistributive effects across regions. These flows shape Spain's geography of income inequality and nurture distributive tensions between those parties and regions benefiting from the status quo and those parties and regions that feel unduly exploited by an excessively centralized fiscal structure.(p. 185)

DISCUSSION OF THE IMPACT OF SPAIN'S DECENTRALIZATION LEVELS
Looking at quantitative decentralization indices in the RAI (Hooghe et al., 2010), the aggregated scores of fiscal authority of the ACs have gradually increased from 2.0 for the period 1983-96 to 3.0 for 1997-2006, and its different regions also have varying fiscal authority scores. 7 For example, Navarra and Basque Country's fiscal autonomy scores are both 4.0, whereas Catalonia and the other ACs have a score of 3.0. On the other hand, the fiscal control (shared rule) scores of all the regions is 1.0, indicating low levels of shared rule, which according to the author's theory and statistical analysis is conducive to higher inequality. However, there are still differences among the regions as well, unlike many other European countries. Some regions can vary the tax rate; for example, Catalonia has increased the income tax rate compared to other regions. Spain's fiscal decentralization concerning its intergovernmental grants has been classified as 'collaborative decentralization' by Gordin (2009) as a 'game being played between the central government and the regions but also as an issue subject to inter-state conflict'.
The ACs are financed through the Organic Law on the Financing of the Autonomous Communities (LOFCA), which has been reformed in 2009 at the Fiscal and Financial Policy Council, an intergovernmental feature of the system whereby different levels interact on redistributive issues. The main aims of the system are stated as the reinforcement of Welfare State benefits; the increase in the equality and adequacy in the joint financing of autonomic authorities; the increase in autonomy and joint responsibility; and the enhanced dynamics and stability of the system and its ability to meet the needs of the public.
Thus, in Spain certain mechanisms exist to ensure that all the regions have adequate funds for services. Among others, the total adequacy fund guarantees the financing of the agencies of the ACs and cities with a statute of autonomy. According to this fund, all regions should receive the same resources, 'in terms of corrected population or unit of need'. 8 One example for how the territorial structure would impact income distribution can be found by looking at the yearly fiscal equalization grants. There are certain studies that show that the results of equalizing attempts are mixed. Herrero-Alcalde et al. (2012) show that Spain's Interterritorial Compensation Fund distributes more resources to the regions with lowest economic activity, but not all the regions that would need this fund are eligible. 9 Before the 2009 reforms, the territorial distribution system was regarded as arbitrary and also over-equalizing, which meant that ACs that had larger fiscal capacity had fewer resources at the end of the year (Bosch, 2011). After 2009, the Competitiveness and the Cooperation Funds have aimed to bridge the gaps in per capita resources among regions, in order to help poorer regions and/or those that suffered from the 'overequalization' (Bosch, 2011).
As an interviewed scholar of administrative law 10 argues: Spain is decentralized in terms of expenses of its regions but centralized in terms of its revenues. In other words, the ACs can spend their own money and provide basic services like education, housing and health, however their revenues for these services largely depend on how much the central government provides to them, which is not always the desired amount.
Regional StudieS, Regional Science It can be argued that the power of the central government in transfers and the relative low level of regional revenue raising authority have prevented Spain's overall inequality levels from rising substantially but the level remains above European average (Gini = 0.335), as Figure 1 shows. Spain's inequality level is higher than Germany (Gini = 0.289), a federal country with a similar fiscal setup but one with lower asymmetries. Therefore, one can presume that if the sub-national units (ACs in this case) and the central government interacted and decided together on fiscal issues towards similar goals, and fiscal asymmetries did not exist, this could have paved the way for lower inequality. The fact that the Senate in Spain is not considered as an institution that represents the interests of the regions, as the Bundesrat (upper house) does in Germany, also detracts from representation of ACs at the central level. As an interviewed expert stated: Spain and Germany are in the same box for fiscal equalization but Germany seeks to reach it culturally as well.
[…] There is a different 'commitment' for redistribution [in Germany]. And there are some unexplained cultural dispositions that distinguish these countries [which have a bearing on redistribution] 11 .

CONCLUSIONS AND IMPLICATIONS FOR ACADEMICS, POLICY AND PRACTICE
The hypothesis of this article is that more fiscal regional authority is significantly related to more income inequality. The case study of Spain sheds light on how the evolving nature of federalism in the country does not really have symmetric 'shared rule' characteristics. Fiscal asymmetries resulting from the territorial setup can be a contributing factor for relatively higher inequality levels in Spain. It shows that the allocation of competences between different levels of the government, especially in fiscal powers, is a key factor for income distribution and national/regional income inequality. The implications of this theory and the case study of Spain can be used in (1) designing more efficient and equalizing intergovernmental fiscal transfers in similar territorial governance structures, (2) identifying the relevant state and non-state actors at the central and regional levels for redistributive policies, and (3) designing more symmetrical interactive redistributive policies whereby governments can reach their equalizing goals by engaging with sub-national governments. This article goes beyond the dichotomy of federal and unitary states and how to reconcile fiscal autonomy and equality and makes a contribution by highlighting the importance of equalizing policies. The case of Germany is a counter example where the redistributive policies are regarded to adhere to the policy of overall 'equality' in a federal country. While Germany is an extreme Fiscal autonomy scores of specific ACs in Spain: • Basque Country: 1979: 4.0 • Catalonia: 1979-96: 2.0, 1997: 3.0 • Galicia: 1981-96: 2.0, 1997: 3.0 • Andalusia: 1981-96: 2.0, 1997: 3.0 • Navarre: 1982