Ivo Vasconcelos Pedrosa University of Pernambuco, Brazil.

Transcription

1 Interactions Between Fiscal Federalism and the Environment: the case of the Sustainable Fiscal Transfers in the State of Pernambuco, Brazil ( ) Felipe Luiz Lima de Paulo Professor at the Federal Rural University of Pernambuco, Brazil PhD student, University of Minho, Portugal Ivo Vasconcelos Pedrosa University of Pernambuco, Brazil. Pedro Jorge Sobral Camões University of Minho, Portugal Acknowledgements: this work was conducted during a scholarship financed by CAPES Brazilian Federal Agency. Process number: / The authors are also grateful for the financial support by the Centre for Research in Political Science (CICP) from the University of Minho in Braga, Portugal. To be presented at the 24th World Congress of Political Science of the International Political Science Association, Poznan, July 23-28, 2016

2 Abstract Environmental issues in Brazil became a central issue in public policy debates. Air pollution, water resources, soil pollution and siltation of the water body are recurring problems in Brazilian environmental scenario. "ICMS-E" was created in order to induce local authorities to create actions towards the improvement of environmental conditions (Ring, 2008). Portugal also adopted a similar public policy instrument (Santos, et al 2012). This study aims at understanding and describing the functioning of this public policy instrument called "ICMS Socioambiental" in the state of Pernambuco, Brazil ( ). The theoretical background is framed in the classical federalism (Oates 1999, 2001; Buchanan 1950; Olson 1965, 1969; Tiebout 1956) and public policy literatures (Lowi 1972; Ostrom 1990). The methodology is anchored on qualitative and quantitative data analysis (spatial autocorrelation data), with data collected by the Secretary of Finance of State of Pernambuco. This study points to three conclusions: first, at the beginning of its implementation, there were several conflicts among the municipalities negatively affected by ICM-E; second, there was a large increase of the social, economic and environmental indicators that meet criteria for the transfer of the "ICMS Socioambiental" to municipalities; and third, there are a high concentration of the transfer of the "ICMS Socioambiental" in the "conservation units" criteria for a few municipalities. Finally, we also suggest some insights for future research and some notes to improve this particular policy instrument. Keywords: Fiscal federalism, Sustainable Fiscal Transfer, Brazil. 1

3 Introduction Theoretically, the sustainable fiscal transfers (SFT) mechanism have its origin in the existence of spillover s problems among jurisdictional units in the provision of social and environmental public goods. The arguments are based on the idea that decentralization of environmental policies is more efficient than its centralization, given that local authorities have lower information costs to meet the preferences of the local citizens with regard their accepted level of environmental quality. However, this idea is contrary to the movement of centralization, which tends to emphasize economies of scale arguments. In this context, the SFT is created, a relatively novel instrument in the policy world landscape. It has already been adopted in two countries: Brazil and Portugal. In the Brazilian case, the SFT involves a state-municipality relationship implemented by the fiscal transfer of ICMS ( Imposto sobre Circulação de Mercadorias e Serviços ), a type of value added tax. In a nutshell, the SFT intends to compensate and encourage local governments to adopt policies towards the improvement in some social, economic and environmental indicators. The compensation refers not only to the costs incurred by local government in the prevention of environmental damage, or their recovery, such as the collection and proper disposal of solid waste, but also to the reduction in economic activities within their territory originated by the existence of conservation units. In both cases, SFT applies the "principle of the protector-receiver". The pioneering state to adopt the SFT mechanism in Brazil was Paraná. Other states followed the adoption of the policy instrument, including Pernambuco, the object of our empirical study. In the case of Pernambuco, the importance of stimulating environmental policies is given in two ways: first, a recent increase in local governmental actions towards environmental issues can be observed, but not as fast as the changes in environmental conditions (Pedrosa & Kakuta, 2008); second, as the analysis progresses from the Metropolitan Region of Recife, the capital of the state, to the countryside, a performance decrease on environmental indexes can be observed, particularly the criteria regarding conservation units and the treatment and disposal of solid waste (Paulo, 2013). In this paper we want to understand how the sustainable fiscal transfer improves the environmental conditions of local government s territories? The empirical analysis focuses on the Environmental and Social ICMS (ICMS Socioambiental the name given to the SFT mechanism in this state) experience in the state of Pernambuco, Brazil in the years of 2011, 2012, 2013, 2014 and We analyse a specific indicator: conservation units, because this was the only environmental variable that has not changed in the ICMS pass percentage collected by the state to the municipalities. The paper is structured into eight sections. The next section revises fiscal federalism literature, the second deals the specific issues of environmental federalism, and the third presents the theoretical framework of sustainable fiscal transfers. The fourth section presents the methodology the fifth details the design of ICMS Socioambiental in the state of Pernambuco, Brazil, the sixth, evaluates its functioning between 2011 and 2015 through an exploratory analysis of spatial data. We end with some conclusions and some policy recommendations and future research. 1. Fiscal Federalism The interest on fiscal federalism, as a sub field of public finance, went to two periods: firstly, strongly prescriptive and anchored on welfare economics, with theoretical and empirical work about US; the second, with the inclusion of political economy concerns. Additionally, the main concern of these studies has been the enduring institutional design question of whether centralizing or decentralizing public sector activities. However, according to Oates (1999), the basic question of fiscal federalism is not necessarily this issue, but the emphasis on two broad aspects: what are the appropriate levels of government and the corresponding division of responsibilities and what are the most appropriate financial instruments to be adopted. 2

4 The literature of fiscal federalism addresses the role of the different levels of government in positive terms as well as in normative terms (Oates 1999). The positive perspective describes how the vertical relationship among the different levels of government works. The normative perspective focuses on how these relationships should be, including concerns of welfare economics and public policy. The classics of the field highlight two very important issues for the development of our theoretical framework: one that deals aspects of equity of fiscal federalism (Buchanan, 1950) and another that explores the theory of local expenditures (Tiebout, 1956). This approach deals specifically with horizontal relations among local governments, while the standard fiscal federalism deals with vertical relationships. James Buchanan (1950) clarifies the policy implications of interstate fiscal adjustment to mitigate inter-regional disparities. In a federal system, the subnational units have a certain autonomy with regard to the political and administrative structure as well as the management of their financial resources. In this perspective, the author favour a process of decentralization. The key point of the analysis is the discussion of the fiscal imbalance among states, related with the emergence of the welfare state, since the citizens demanded for higher levels of public services. Therefore, Buchanan suggests a fiscal transfer mechanism to promote an "equal treatment of equals" and the "unequal treatment to unequal people." In the words of the author, "an individual should have the assurance that wherever he should desire to reside in the nation, the over-all fiscal treatment which he receives will be approximately the same (Buchanan, 1950, p. 589). Charles Tiebout (1956) presents the famous model based on the mobility of families in local jurisdictional units, arguing that they end up having their preferences revealed by the choice of the tax package adopted by the local government, the famous metaphor that they "vote with their feet". In this analysis, decentralization would capture the gains of competition among jurisdictional units. An opposite view, presented by Gordon (1983), points that crossjurisdictional competition in a non-cooperative environment provides no benefits to local governments position. To sum up, the literature does not provide a general consensus on what kind of public policy should be centralized or decentralized, especially when we are dealing with a novel issue in the context of the fiscal federalism field or, in other words, the environmental federalism. In the next section we address this environmental federalism in more detail, based primarily on the study of Oates (2001). Thus, based on the classic study of Tiebout (1956), our basic assumption to the study of sustainable fiscal transfer is that decentralization provides gains to local governments with regard to environmental policies, gains that, according to Oates (1999), are observed in industrialized countries as well in developing countries. Thus, decentralization combined with the use of fiscal instruments would bring welfare gains, because the population would have their revealed preferences locally served by such levels of government. When economies of scale are present, centralization turns out to be preferred. 2. Environmental Federalism The study of environmental federalism involves, in a federal system, vertical relationships among levels of government addressing environmental policies, focusing on the explanation of what should be the most appropriate role for each jurisdiction. But we cannot also ignore the role of horizontal coordination to solve environmental problems, a solution that is implied by the Coase's (1960) analysis, in which it is theoretically possible to obtain efficient allocation of productive resources through negotiation among parties. However, when we consider these coordinative solutions at the local government level, there are some types of barriers which are thoroughly analysed by Feiock (2008). Giving the focus on the problems that effectively involve environmental federalism, there are points in the analysis by Oates (2001). The first is the problem of competition among local governments, which may in some way affect environmental quality. This perspective is based on 3

5 the idea that there is a trade-off between economic development and environmental quality, because municipalities, in order to attract firms to their territories, may be tempted to relax the requirements of environmental regulations. This position is one of the main arguments for those who argue for greater centralization of environmental policy but that is not what supports Oates (2001). The decentralization of environmental policies especially to the local level has, however, several benefits. First, a strong argument for decentralization is that the local governments have a lowest cost to obtain the information on the preferences of citizens about the expected level of environmental quality. Consider the huge cost a central government would have in order to know the preferences of a given population of a local area. Another point is that the population makes decisions also considering the aspect of capitalization, because future decisions are generally well thought, since they have, in theory, impact on present (would be sound to estimate the net present value of environmental public goods that will be consumed by the population in the future). The participation of citizen s in local decision-making with respect to a given environmental public policy is essential, therefore the local government is the most appropriate to level of government to allow that participation. Thus, adapting the concept of environmental federalism to fiscal federalism, particularly the principle of fiscal equivalence detailed in the seminal paper by Olson (1969, p. 483), it is important to note "that there is a need for a separate governmental institution for every collective good with a unique boundary, so that there can be a match between those who receive the benefits of a collective good and those who pay for it". Two empirical studies are relevant for our study, supporting the idea that decentralization does not compromise the quality of environmental requirements at local level. Fredriksson & Millimet s study (2002), dealing with the US, conclude that what really matters is not the raceto-bottom for decelerations or not of environmental quality at the local level, but the spatial dependence of rules that exist among neighbouring jurisdictions. McConnell & Schwab (1990), also based also on the US reality, deal with the impact of environmental regulations on industrial location decisions. The conclusion is that the differences in environmental regulations between regions exerted no significant influence on the choice of the location of an industry. There are certain types of natural resources exploitation that go beyond the limits of local jurisdiction (spillover effect), causing corresponding economic anomalies, such as the externalities (positive as well as negative) and the appearance of the free-rider behaviour. Let us imagine, for example, that a local government has in its territory some biodiversity preserved (conservation units), which freely ensures clean air and clean water to other municipalities. Based on the theory of collective action (Olson, 1965), this is a typical example that ends up encouraging the free-rider behaviour by benefited municipalities, just because an individual action in a large group ends up unnoticed, as is the case in Brazil, where there are municipalities. In this way, this situation ends up generating an external savings money for the municipalities that do not support the costs of preservation in an environmental zone in a single jurisdiction, or in a small group of municipalities. However, in such cases, the traditionally accepted solution is environmental public policy made at the central level, in order to establish a minimum standard of environmental quality for all federal entities. Another example that can be given is when a negative externality is generated, such as an irregular garbage dump in a municipality, which ends up impacting with pollution into other jurisdictions. The literature also details specific ways to correct the anomalies cited above. The most wellknown is the establishment of a solution based on Coase (1960), but there are barriers to internalize an externality and minimize free-rider behaviour involved in collective actions, the socalled transaction costs. Feiock (2008) works with this approach to solve production problems of collective public goods among jurisdictional units. He studies extensively what are the main barriers of the institutional collective action and how to overcome them. In this study we are more interested in the vertical relationship among the states and local government to solve environmental collective problems using fiscal transfers instruments. 4

6 In sum, from this discussion emerges an incentive-based solution to fix these economic anomalies, a public policy instrument of fiscal transfers between state and local government. This instrument aims to encourage and compensate municipalities to produce environmental public goods, generating a positive externality for the other jurisdictions, so as to make the competition among municipalities non relevant to the relaxation of environmental standards at the local level. This public policy instrument ends up doing an important contribution in the production of environmental public goods. Another aspect is that the municipality turns out also to have a selective positive incentive (Olson 1965), making relevant the participation of each local jurisdiction in the large group to produce an environmental public good, as the case of the conservations units and the public consortia law adopted in Brazil since Sustainable Fiscal Transfers Environmental federalism deals with the vertical relationships among the levels of government for allocation of responsibilities with respect to environmental issues. We discussed the trade-off between centralization and decentralization in allocating these responsibilities. We also reviewed the economic anomalies resulting from the spillover effect in large groups that give rise to the free-rider behaviour, causing the externalities to emerge. The solutions presented in the literature are either based on horizontal solutions, such as the presented by (Coase, 1960), or based on a vertical solutions, such as the incentives induced by fiscal transfers, addressed in this paper. The vertical solution of sustainable fiscal transfers have already been adopted in two countries, Brazil and Portugal (Ring, 2008; Santos, 2012) and are known in the literature as "ecological fiscal transfer" (EFT). EFT s can be seen as a type of intergovernmental fiscal transfer that intends to encourage and compensate local governments that improve environmental, social and economic results. Reconciling theory and practice, in the Brazilian case these types of intergovernmental fiscal transfers are known under various names ( ICMS Ecológico, ICMS Verde and ICMS Socioambiental ) and are part of the share that is allocated to municipalities as Imposto sobre Circulação de Mercadorias e Serviços (ICMS) a kind of value added tax (VAT) levied by state governments. In the Portuguese case, this type of instrument was introduced by the local finance law (Santos et al., 2012). The ecological fiscal transfer is, in both countries, a compensation mechanism to mitigate the spillover effect that a local jurisdictional unit may have in the production of an environmental, social and economic public goods. It is more appropriate to use the expression SFT since these dimensions need to be linked to the requirements of what we may want to be sustainable development. Thus, we explain the SFT in its two main characteristics: first, how it can be understood as an incentive for local governments to internalize the spillover effect; second, how it can be seen as a financial compensation instrument for producing environmental public goods beyond its borders. These two characteristics are coherent with what is advocated, respectively, by the theory of collective action (Olson, 1965) and on the theory of fiscal equivalence (Olson, 1969). The incentives provided by the sustainable fiscal transfer are the remedy to the basic problem of collective action, in other words, it is used to correct the free-rider problem which is often seen in large groups. In federal countries the number of local governments can be up to thousands, making individual choices or actions not relevant to obtain a collective gain (Olson, 1965). From this principle, the solution proposed by the theory of collective action to this kind of problem is to produce a positive benefit as a mean to turn individual action important. Amidst the shortage of financial resources, politicians and bureaucrats making rational decisions to seek higher revenues to expand their budgets. In that regards, the SFT emerges as an incentive to municipalities through the fiscal transfer by ICMS Socioambiental. However, this is a ceteris paribus condition, without any asymmetric information among the political actors. However, it is also important to note that there are several factors that can mitigate the compliance with environmental criteria involving common pool resources that may be required by the state 5

7 government, such as the transitions costs and the congruence between the rules and local conditions (time, available territory, available technology, financial resources and others) (Ostrom, 1990), anthropogenic pressures and exploitation of economic activities in the municipality, as these may interfere with the environmental quality of the city, in order to form as environmental rules are established between neighbouring jurisdictions (Fredriksson & Millimet, 2002). The compensation provided by sustainable fiscal transfers to municipalities as a means to internalize the spillover effect is mainly based on the theory of fiscal equivalence (Olson, 1969). The principle of fiscal equivalence is based on three important assumptions. Firstly, it is assumed that when the collective public good goes beyond local government boundaries ends up generating a sort of external savings for other municipalities, which ends up generating the spillover effect. Secondly, the given collective public good reaches only part of the citizens, meaning in this case that the provision of the collective public good in a local government will be prejudicial for more people than help them, even if the Pareto optimality requires that the collective public good should be provided (Olson, 1969). Thirdly, the boundaries of the collective public good are the same as the government territory providing it. Based on these three assumptions of the fiscal equivalence, the sustainable fiscal transfer ensures that the marginal costs incurred in providing the environmental, social and economic public good is compensated by the marginal benefit that has source in the sustainable fiscal transfer (in the relationship between state and local government). Additionally, we should point out to some other important characteristics of the SFT. However, this type of public policy instrument has some redistributive features, i.e., a great advantage in the use of such transfers is that the cost for the implementation of these instruments is very small because the state government will not have any waiver of revenue from taxes. However, referring to the study of Lowi (1972), we believe it is important the participation of the various actors involved in this type of public policy instrument, just to minimize any kind of conflict and thus not compromise the design and scope of the objectives laid down by SFT. 4. Methodology This study aims at understanding the mechanism of sustainable fiscal transfers towards the improvement of environmental conditions in local governments and also describing how this mechanism was implanted in the state of Pernambuco, Brazil, through a short evaluation of this instrument of public policy between the years of 2011 and The methodological approach is both qualitative, analysis of the legislation dealing with the "ICMS Socioambiental", and quantitative, an exploratory analysis of spatial data in the distribution of financial resources allocated to municipalities through fiscal transfer using the criterion of "conservation units". The empirical analysis of this study focuses on a specific indicator of the ICMS Socioambiental : conservation units. This analytical approach is precisely due to two aspects: i) this article aims to evaluate the contributions of SFT for the environment of the state of Pernambuco; ii) this indicator has been used since the beginning of the creation of this public policy instrument at the state and with the same percentage of transfer of ICMS to municipalities. 4.1 Data and Measurement The database is a public source of secondary data and can be consulted in the official government newspaper of the state (called Diário Oficial do Estado de Pernambuco - DOE-PE). We use a biodiversity conservation index, called Indicador de Conservação da Biodiversidade do Município ICBM, which is managed by the state environmental agency, called Agência Estadual de Meio Ambiente CPRH. The ICBM is calculated by the following formula: 6

8 ICBM = ( CBM ) X 100 CBE In which: - ICBM: Index of Biodiversity Conservation of Municipality (Índice de Conservação da Biodiversidade do Município ICBM). - CBM: Sum of all biodiversity conservation coefficients calculated for the municipality (Conservação da Biodiversidade do Município CBM). - CBE: Sum of all biodiversity conservation coefficients calculated for all municipalities in the state (Conservação da Biodiversidade do Estado CBE). Thus, for the measurement of biodiversity conservation coefficient is used the following formula: CB = ( AUC X FC) X AQUC AM In which: - CB: coefficient of the conservation unit. - AUC: area of the conservation unit - AM: municipality area - FC: conservation factor - AQUC: quality of the conservation unit In order to obtain additional information on the quality of the conservation unit, we also interviewed the ombudsman of the state of Pernambuco by . With regard to local public finances data, we made use of a public database, called FINBRA - Finance of Brazil, particularly the expense function of the environmental management, of the national treasury (Secretaria do Tesouro Nacional STN). 4.2 Data analysis Quantitative analysis Two quantitative analyses were performed: i) analysis of the spatial distribution with equal intervals; ii) analysis by global spatial autocorrelation, using the statistic of Moran index, a correlation coefficient varying from -1 to +1. The spatial dependency is also analysed in the Moran scatter diagram analysis. The Moran scatter plot is a graph that the variable of interest is on the x-axis and the spatial lag on the y-axis (Anselin, 2005). The diagram is divided into four quadrants, as follows: The interpretation of the diagram is as follows: Q1 (high-high) shows the locations with high values for the variable analysed, surrounded by places with values above average for this variable; Q2 (low-high) are the locations with low value for a given variable, surrounded by places with 7

9 high values; Q3 (low-low) are local with low values surrounded by places with low values too; Q4 (high-low) are places with high values and surrounded by places with low values. Locations in the quadrants 1 and 3 denote points with positive spatial association, while the quadrants 2 and 4 present negative spatial autocorrelation. (Anselin, 2005) The areas were classified into five levels of significance (Anselin, 2005): no significance; significance level of 0.05 (95% confidence); 0.01 (99% confidence); 0,001 (99.9% confidence); and (99.99% confidence) Qualitative analysis The qualitative analysis was based on critical analysis of Brazilian legislation on the "ICMS Socioambiental through the lens of the sustainable fiscal transfers theoretical framework, particularly the principle of the incentive and the compensation. 5. ICMS Socioambiental in the State of Pernambuco: the design of the public policy instrument 5.1 The prospect of the legislation Two countries adopted the mechanism for sustainable fiscal transfer, Brazil and Portugal (Ring, 2008; Santos et al., 2012). The first Brazilian state to adopt this intergovernmental fiscal transfer mechanism was Paraná in the early 90s (Paulo, 2013) and currently there are 16. In the case of the state of Pernambuco, this policy instrument was enacted by the state law number , of December 21, 2000, based on article 158 of the 1988 Federal Constitution of Brazil, which authorized the use of social, environmental and economic criteria for the distribution of the tax on transactions relating to the movement of goods and of interstate transport services, intercity and communication (Imposto sobre Circulação de Mercadorias e Serviços ICMS). This SFT mechanism adopted in state of Pernambuco emerged under the name ICMS Socioambiental, while in other states have adopted the designation of ICMS Verde or ICMS Ecológico. This difference is relevant and was due to the implantation of social, economic and environmental criteria in Pernambuco, while other states have used environmental criteria only. Nevertheless, a common feature of these instruments was the proposal to internalize the spillover effect resulting from the production of environmental public goods, stimulating and compensating each municipality. According to the article 158 of the Federal Constitution of Brazil, the municipalities (184 in this case) are entitled to a share of 25% of the total amount raised by ICMS. Considering the way municipalities participate in the intergovernmental transfer revenues adopted in Brazil, all municipalities have the right to participate in the distribution of revenue of ICMS in each state in 25% of the total collected. The mechanism of transfer that allows states to use the SFT is just part of those 25% that are intended to municipalities. Of the 25%, ¾ (three-quarters) are passed municipalities using the criterion of fiscal value added, and only one-fourth (¼) are freely distributed to states so that they can define what criteria is to be used to establish of these intergovernmental fiscal transfers. Figure 1 shows the percentage that each state can adopt to distribute as SFT, which can be based on social, economic and/or environmental criteria. In the end, only 25% of the 25% of ICMS revenue can be seen as SFT. [Insert Figure 1] Thus, ¼ of the 25% that are allocated to municipalities may be linked to the fulfillment of criteria related on environmental, social and economic indexes required by the states. In the state of Pernambuco, such criteria undergone various changes, precisely because of the redistributive effect generated by the division of the revenue tax collected by the state related to the ICMS to municipalities. 8

10 This redistributive effect generated various conflicts, which were evident from the beginning, because the municipalities negatively affected lobbied the association of municipalities in the state of Pernambuco, Brazil (Associação Municipalista do Estado de Pernambuco - AMUPE) in order to pressure the executive and legislative to change the law (Jatobá, 2005). Given its redistributive effects, in this political process the SFT instrument became known as the Robbin Hood law. At the beginning of the implementation, some social, environmental and economic criteria that were adopted initially were changed, including the weight of each criterion. To get an idea, for the year of 2002, the state law established that of the 25% of the total ICMS revenue collected that should be passed on to the municipalities, ¼ of this amount would be distributed on the following criteria: relative share of each municipality (15%); conservation units (1%); composting plants units or landfill (4%); health (2%); education (2%); criterion for encouraging the economic development of the municipalities by encouraging the increase of their own tax revenue (1%). Figure 2 shows how these criteria were at the beginning of use of this instrument of public policy. [Insert Figure 2] An important point that we must explain is the mentioned criteria that relates to the relative share of each municipality. In fact, this criterion refers to the relative share of the sum of the differences between the percentage index prevailing for each municipality in the previous year and the percentage of their relative share in the value added of the state. According to Paulo (2013), this mechanism was created in the 80s to reduce the impact on the index, and hence, on transfers of large variations in the fiscal value added owing to the precariousness of the "administrative" statistics up the fiscal value added calculated based on the tax collecting activity. Regarding the other criteria, it is observed that refer to environmental criteria (conservation units and composting plants or controlled landfill), social (education and health) and economic (stimulus to increase the revenue generated by municipalities). However, as explained before, these criteria have undergone changes over the years 1. The fact is that this range of criteria and the subsequent spraying of the percentage share of the same of ICMS transfer to municipalities may have constrained the full scope of objectives initially proposed by the ICMS Socioambiental as a model of SFT (Jatobá, 2005). But this claim still lacks empirical support and should be seen just as a possible cause. Thus, to the exercise of the year 2017, according to SEFAZ-PE (Finance Secretary of the state of Pernambuco) (1990), the ICMS Socioambiental turned to include fourteen criteria in environmental, social and economic areas. In the environmental area are observed four criteria: conservation units, treatment and disposal of solid waste, preserved water source and rivers, and waste recycling plants. In the social area, it is observed eight criteria: child mortality (inversely), teams in the family health program, the enrolment of children s in early childhood public education - kindergartens, proficiency in the 3rd year of elementary public schools, development index of education in Pernambuco state, enrolment in primary public education, number of prisons and penitentiaries and number of intentional lethal violent crimes (inversely). In the economic area, are observed two criteria: encouragement the municipalities to increase their own tax revenue, and GDP "per capita" (inversely). Figure 3 shows how this will be distributed, in percentage, from 2017 on. [Insert Figure 3] We must emphasize some important points. The first is that not all of these criteria are put into practice today. Those being used in the state of Pernambuco: conservation unit, treatment and disposal of solid waste, resident population, municipal GDP "per capita" (inversely), number of teams in the family health program, index of violent crimes, number of inmates (three-year 1 Interested reader in knowing all the changes made to the legislation by the year 2009 we suggest carefully read the master s thesis of Paulo (2009), especially section 3, which is a comprehensive review of the ICMS Socioambiental law in the state of Pernambuco, Brazil. 9

11 average), own tax revenue "per capita" and the basic public education development index of municipalities. Another important factor is that for our analysis, as described in the section on methodology, we will analyse only the empirical data of transfers related to the environmental criteria, and of these just one being applied since the beginning of this instrument of public policy (and with the same percentage for the transfer): conservation units. In this sense, we must strengthen an important observation about the criteria of conservation units that we have said in the section on methodology, which is the fact that although they are in the municipality, they may be responsibility of federal government, states and the municipalities. Thus, if a protected area is not a municipal responsibility, it is not applied the criterion of incentive for them because the municipality does not have any interference in the decision to create or not a conservation unit. However, we understand that applies the principle of compensation, in which case is applied the opportunity cost. Another important aspect to be highlighted in relation to protected areas is that the qualitative assessment that makes up the "Index of Biodiversity Conservation of Municipality" (Indicador de Conservação da Biodiversidade do Município ICBM) it is not been applied in practice, although is part of the ICBM calculation (as described in the section on methodology). This information was given by the agency of the protected areas, that is a state environment agency (Agência de Meio Ambiente e Recursos Hídricos CPRH) in the state of Pernambuco, Brazil. Finally, it is important to note that in the case of fiscal transfers that arising from the ICMS Socioambiental, the municipality has no obligation to hold spending on specific functions for receiving financial resources that is originated from this fiscal transfer, i.e., there is no linking of the transfer with any spent on specific expense. 5.2 The compensation of ICMS Socioambiental The principle of the compensation for the preservation of a "conservation unit" in the ICMS Socioambiental perspective in the state of Pernambuco, is very important for two reasons. First, the perspective of the compensation provides environmental public goods beyond its borders, in the case when the conservation unit created is the municipality's responsibility, being compensated for the marginal costs. Second, the perspective in which the municipality has little power to influence the creation of a conservation unit, which is the case of those is the federal or state responsibility (Grieg-Gran, 2001), in which case is most appropriate the local government be compensated in the perspective of opportunity cost, i.e., offset by the loss of economic activity that ceased to be exploited for the preservation of that conservation unit. 5.3 The incentive of ICMS Socioambiental Our evaluation of the "ICMS Socioambiental" is restricted to only one criterion: conservation units. In this sense, an important point to be made is that protected areas can be of the responsibility of the federal, state or municipal government. Thus, in the case of creation of a conservation unit that is not of the responsibility of municipalities, cannot apply the principle of the incentive mentioned in Section 3, because the creations of them do not have any interference from local government. However, applies the principle of compensation, in which case we believe that the perspective to evaluation of this public policy instrument is the opportunity cost, or in other words, it is not the compensation on the bank of those environmental benefits that going beyond its borders. Also in relation to protected areas, it is also important to mention the research by (Fernandes et al, 2011) about the Ecological ICMS in Minas Gerais, Brazil. In that case study, the conservation units in those municipalities had as explanatory variables: the value received from the "ICMS Socioambiental" for the subcriterion of conservation units; the number of inhabitants that the city has (in reverse, as observed that the smaller was the antropogenic pressure higher was the amount of preserved area); and the ICMS received by the criterion of fiscal value added (in reverse, as the lower was the economic activity in the municipality, the greater the natural area to be preserved). These results may suggest future studies with explanatory designs, in order to 10

12 understand the extent to which the "ICMS Socioambiental" works as an incentive for municipalities in creating and maintaining a conservation unit. 6. Evaluation of the "ICMS Socioambiental ( ) 6.1. Analysis of the spatial distribution with equal intervals In this section we present how the ICBM are distributed in the years of 2015, 2014, 2013, 2012 and The index reflects the conservation units located in the municipalities of the state of Pernambuco, as described in the methodology section. Thus, the municipality with the highest performance receives the highest transfer of ICMS collected in the state, according to the rules established by the ICMS Socioambiental. In this sense, we will present the spatial distribution for the ICBM in year [Insert Figure 4] Only two (Ilha de Itamaracá and Itapissuma) of the 184 municipalities of the state, have high positions for the year 2015 in the ICBM index, and , respectively. Two other municipalities are in the range 4.99: 7.49 (Buíque and Tupanatinga) with ICBM index position situated in and , respectively. Seven other municipalities are situated in the range of 2.5: 4.99 and the rest (174 municipalities) are in the range 0: 2.5. There is, therefore, a concentration of municipalities with high positions in the index of ICBM, culminating that these locations have received higher financial transfers resulting from the ICMS Socioambiental in the criterion of conservation unit in the year To get an idea, in the year 2015, the total of the ICMS" transferred to municipalities in state of Pernambuco amounted to ,51 Brazil Real (BRL) 2. Of this amount, ,12 BRL correspond to the portion of the ICMS that is part of the environmental criteria (conservation unit and treatment and disposal of solid waste). With more detail, it is possible to note that the total transferred corresponding to the percentage related to the conservation units (1%) and treatment and disposal of solid waste (2%), which in total corresponds to 3% for environmental issues. The calculation below presents this distribution: Environmental ICMS (%) = ( ,12 ) x 100 = 3% ,51 In the municipalities of Ilha de Itamaracá and Itapissuma in the year 2015, the fiscal transfers of ICMS Socioambiental values in the criteria "conservation unit" were ,17 (BRL) and ,21 (BRL), respectively. This represents almost 12.34% and 12.48% of the total of the ICMS Socioambiental transferred to municipalities for the criterion "conservation units", which together amount to almost 25% of total ICMS Socioambiental transfer to municipalities in this criteria. However, when we compare the amounts received by the municipalities with the highest ICBM indexes in the ICMS Socioambiental for the criterion "conservation units" with the expenses paid in the year 2015, we realized that this revenue that comes from the tax ICMS transfer corresponds to about 86.77% of expenditure in the environmental management function (3,589, BRL) for the municipality of Ilha de Itamaracá, and 16.92% of expenditure in the environmental management function (18,601,835,67 BRL) for the municipality of Itapissuma. The fact is that this concentration of the ICBM in some localities did not change so much between the years 2011 and 2015, as we can see in the histograms shown in sequence. Other fact is that the behaviour of the revenue transfers of ICMS Socioambiental for the conservation units and 2 Which corresponds to ,98 US Dollar (USD) with exchange on 06/14/2016 corresponding to 1 USD to 3,4903 BRL. 11

13 the expenditure of environmental management (that includes all the expenses of the environment management in the municipality) did not change significantly. [Insert Figure 5] Thus, we can observed that there are higher index in a few municipalities, and we think that this behaviour may indicate some factors: municipalities do not have land to be preserved (anthropogenic pressures, for example); municipalities have no interference on the creation of the conservation units that may be responsibility of the states or federal jurisdiction; or that local governments do not see some sort of compensation or incentive for the creation and maintenance of new conservation unit. An important point to be made is that a better distribution of ICBM index between the municipalities into the state of Pernambuco does not imply necessarily that local government no longer has a conservation unit that already existed, but that did not increase new ones at a faster rate than other municipalities. Thus, as we can see on the map of the year 2011, the same two municipalities that in the year 2015 were with the highest ICBM index are also present in the wider range of class when we analyse the spatial distribution. The municipalities of Ilha de Itamaracá and Itapissuma have the largest indexes, corresponding to and , respectively. Further, we can observe two municipalities that fall under the class that is between 5.91: 8.86, which are the municipalities of Tupanatinga and Buíque with ICBM index in and , respectively. Another six municipalities are in class between 2.95: 5.91, while the rest, 175, are in Class 0: 2.95, indicating that the concentration of ICBM index, and the financial resource transfer of ICMS Socioambiental, consequently, for this sub-criterion, is still concentrated a few municipalities in the state of Pernambuco, Brazil. [Insert Figure 6] Thus, as we can see during the study period that we have detailed for the ICBM index (years 2011 to 2015), which is one of the criteria used for the transfer of financial resources from ICMS collected by the state to municipalities, that it not was changed significantly. It is important to deepen our study to identify the extent to which the ICMS Socioambiental has worked as a mechanism of incentives or as financial compensation for the municipalities of the state of Pernambuco during this period. 6.2 Correlation analysis of spatial data In this section, we will present the values of spatial association of the ICBM index in the state of Pernambuco, for the years 2015 and The choice to show only the data for those years, and not for the years in this interval of time is because we observe that the values did not change significantly over this period. In this sense, we present below the Moran index for year [Insert Figure 7] As we can see in the chart, the Moran index for the year 2011 was , indicating that there is positive spatial autocorrelation. However, it is observed that there is a greater concentration in those municipalities that have low values and are surrounded by low values. To check the level of statistical significance of the data presented, we calculate the pseudo significance level with 999 permutations, that was 0,001. We follow this analysis with the years 2012, 2013 and 2014 and the results have not changed significantly. In this sense, we believe it would be sufficient to exhibit the result just for the year 2015, as we can see in sequence. [Insert Figure 8] 12

14 Finally, as we can see in the chart above, the Moran index for the year 2015 corresponds to , indicating that there is positive spatial autocorrelation. However, as we can observe in the years 2011 to 2014, it is noted that there is a greater concentration in those municipalities that have low values and that are surrounded by low values too. To check the level of statistical significance of the data presented, we calculate the pseudo significance level with 999 permutations, that was 0,001. Conclusion The sustainable fiscal transfer emerges as a new alternative to solve the spillover problems among local jurisdictional units in the production of environmental public goods. In Brazil, it is already in place a public policy instrument that is based on this theoretical perspective, while using different names from state to state ("ICMS Ecológico", "ICMS Socioambiental" and "ICMS Verde"). In the specific case of the state of Pernambuco, Brazil, we realize that: 1) at the beginning of its implementation, this public policy instrument raised several conflicts among the municipalities negatively affected; 2) there was a large increase of the social, economic and environmental criteria for the transfer of the "ICMS Socioambiental" to municipalities; and 3) there are a high concentration of the transfer of the "ICMS Socioambiental" in the "conservation units" criteria for a few municipalities. A strong recommendation to improve the effectiveness of instrument is to effectively include the qualitative criteria in the evaluation of the conservation units, since the criterion at work so far is the total area of the territory. A specific limitation is that the ICBM index used in this study includes all types of conservation units in the territory of the municipalities (whether federal, states or municipalities). Accordingly, the results can be limited with regard to the non-disaggregation of that measure. Another limitation of this study is related to the expenditure on environmental management functions that we use as a comparative example with the revenue from the "ICMS Socioambiental" related with the "conservation units" criteria. It should be noted that these expenses are aggregated, i.e., constitute all expenditures with environmental management in the municipality, and not only those related with the conservation units. These exploratory results (a high concentration of the ICBM index in some municipalities), suggest future work that should deal with explaining what factor contribute to the creation of a conservation unit in a municipality. Thus, we suggest a quantitative analysis by panel data using a model based on Fernandes et al. (2011): Y it = β 0 + β 1 CICMSE it + β 2 POPULATIO it + β 3 VA it + β 4 DEPFPM it + μ it In which: - Ƴ it = municipal conservation unit (i) the time (t), through the ICBM - VICMSE it = revenue of the ICMS Socioambiental for the criterion "conservation units" in the municipality (i) the period of time (t) - POPULATION it = local population (i) in the period of time (t) (inversely). - VA it = value added in the city (i) in the period of time (t) (inversely) - DEPFPM it = municipality dependence on the transfer of the municipality participation fund. We suggest use the following index: 13

15 In which: DEPFPM = FPM RO - FPM: municipality participation fund - RO: municipality budget revenue. 14

16 Bibliography Anselin, L. (2005). Exploring Spatial Data with GeoDa: A Workbook. Geography, Buchanan, J. M. (1950). Federalism and Fiscal Equity. American Economic Review, 40(4), Coase, R. H. (1960). The Problem of Social Cost. The Journal of Law and Economics. Feiock, R. C. (2008). Institutional collective action and local government collaboration. Big ideas in collaborative public management. Retrieved from Ideas_ch10.pdf Fernandes, L., Coelho, A., Fernandes, E., & Lima, J. E. (2011). Compensação e Incentivo à Proteção Ambiental: o caso do ICMS Ecológico em Minas Gerais. Revista de Economia E Sociologia Rural, 49(03), Fredriksson, P. G., & Millimet, D. L. (2002). Strategic Interaction and the Determination of Environmental Policy across U.S. States. Journal of Urban Economics, 51(1), Grieg-Gran, M. (2001). Fiscal Incentives for Biodiversity Conservation: The ICMS Ecologico in Brazil. SSRN Electronic Journal, (December). Gordon, R.H. (1983) An optimal taxation approach to fiscal federalism, Quarterly Journal of Economics 98, Jatobá, J. (2005). El impuesto sobre circulación de mercaderías y servicios de transporte (ICMS) como instrumento económico para la gestión ambiental: el caso de Brasil. In Política fiscal y medio ambiente: Bases para una agenda común (pp ). Oates, W. E. (1999). An Essay on Fiscal Federalism. Journal of Economic Literature, 37(3), Oates, W. E. (2001). A Reconsideration of Environmental Federalism, (November), -. Retrieved from Ostrom, E. (1990) Governing the commons: the evolution of institutions for collective action. Cambridge, MA: Cambridge University Press Olson, M. (1965) A lógica da acção colectiva: bens públicos e teoria dos grupos. Oeiras: Celta Editora. Olson, M. (1969). The principle of fiscal equivalence. American Economic Review, 59, 479. Paulo, F. L. L. de. (2009). O Imposto sobre Circulação de Mercadorias e Serviços - ICMS Socioambiental como Incentivo à Melhoria das Políticas Públicas Ambientais em Nível Municipal: o Caso de Municípios da Região Metropolitana do Recife - PE. Universidade de Pernambuco. Faculdade de Ciências da Administração - Mestrado em Gestão do Desenvolvimento Local Sustentável. Paulo, F. L. L. de. (2013). Imposto Sobre Circulação De Mercadorias E Serviços Icms Socioambiental : Avaliação Da Política Do Estado De Pernambuco Nos Últimos Cinco 15

17 Anos Na Perspectiva Da Gestão Ambiental. Fórum Ambiental Da Alta Paulista, Pedrosa, I. V. &, & Kakuta, A. É. (2008). Os Governos Municipais e a Crise Ambiental: o Caso de Pernambuco. In IV Encontro Nacional da Anppas Brasília - DF Brasil (pp. 1 14). Ring, I. (2008). Integrating local ecological services into intergovernmental fiscal transfers: The case of the ecological ICMS in Brazil. Land Use Policy. Santos Rui, R., Ring, I., Antunes, P., & Clemente, P. (2012). Fiscal transfers for biodiversity conservation: The Portuguese Local Finances Law. Land Use Policy. Tiebout, C. M. (1956). A pure theory of local expenditures. The Journal of Political Economy, 64(5),

18 Figures Figure 1 Intergovernmental Transfer of ICMS, Fiscal Value Added and State Law Figure 2 Initial Distribution of the ICMS Socioambiental 15% 15% 2% 2% 4% 1% Relative share of each municipality Composting plants units or landfill Education Conservation units Health Tax revenue generated by municipalities 17

19 Figure 3 Distribution of ICMS Socioambiental in 2017 Number of prisons and penitentiaries Number of intentional lethal violent crimes (inversely) GDP "per capita" (inversely) Tax revenue generated by municipalities Number of enrollments in primary public education Development Index of state education Proficiency in the 3rd grade of elementary school Number of children in public childhood education - kindergartens Teams in the Family Health Program Infant mortality (inversely) Waste recycling Protection of water sources and rivers Treatment and disposal of solid waste Conservation units 0% 1% 2% 3% 4% 5% 6% Figure 4 18

20 Figure 5 19

21 Figure 6 Figure 7 Figure 8 20