FISCAL WAR AND STATE DEVELOPMENT POLICIES IN BRAZIL Sergio Prado INTRODUCTION Item 1. THE FEDERATE DISPUTE AND THE CENTRAL GOVERNMENT’S CO The VAT taxation system and the role of CONFAZ The weakening of the central government’s control power Item 2. ALLOCATIVE DECISION AND REACH OF THE STATE INCENTIVE 2.1. The long-term trends Resumption of private investment and the regional developme 2.3. Limits of the fiscal war: the impact of incentives on Item 3. FISCAL WAR, STATE FINANCES AND TAX REGIME 3.1. A typology of intervention instruments Selective benefits 3.2. Alternative mechanisms for the use of tax competence By way of conclusions Item 4 – RESOLUTION ALTERNATIVES FOR THE FISCAL WAR First approach: guidelines for elimination of state developm Second approach: the limited autonomy of the SNGs in the re Final remarks FISCAL WAR AND STATE DEVELOPMENT POLICIES IN BRAZIL Sergio Prado 1 ABSTRACT This paper2 addresses the phenomenon of the “fiscal war” in Brazil, attempting to identify its historical and institutional origins, the types of instruments used and the impacts of its practice on the finances of state governments. Two basic views are supported here. First, that the “fiscal war” does not necessarily result in losses for the states that practice it, but does result in loss for the whole of the country, since it does not create new investments and maximizes the fiscal cost of subsidizing private investment projects. Secondly, that the unusual recent expansion of the fiscal war is closely linked to the weakening of the central government’s regulation capacity and to the current administration’s decision, in the recent period, to abandon and reject any initiative in the scope of discretional development policies. INTRODUCTION The use of exemptions and subsidies of various types, but mainly related to taxes, has been a common and enduring practice of Brazilian state governments at least since the sixties. After a period of exacerbation between the late sixties and early seventies, when the so-called “fiscal war” became news for the first time, the trend lost evidence until the late eighties. In the beginning of the nineties, it returned, incipiently at first, and later breaking out, as from 1993/4, into a major national controversy. There was never another time, since the fifties, when there weren’t at least a dozen states (and, more recently, also large municipalities) with regional/local development programs in operation, employing mainly tax incentives. The “fiscal war” can therefore be viewed as a peculiar state of exacerbation of the competitive nature of these policies. For this reason, and perhaps other even more relevant reasons that will be outlined below, the approach adopted here is to analyze the fiscal war in the context of regional development policies. 1 Professor of the Economics Institute of UNICAMP. 2 This paper summarizes the results of a broad research on the theme, carried out by the Directorate of Public Sector Economics of FUNDAP – Foundation for Administrative Development – a body linked to the State of São Paulo, under the coordination of the author, and published in PRADO, S./CAVALCANTI, C.E. (2000). This approach unfolds into a focus on at least three important analytical dimensions of the phenomenon. First, the most evident and perhaps the main one, is the dispute in the federate context. The argument developed in item 1 is that the Fiscal War, as practiced today, is dependent upon the combination of a certain legal arrangement on the taxation of the VAT (value-added tax) and the gradual weakening of the regulation capacity (in a broad sense) of federate relations on the part of the central government. The developments that took place in this context made room for a new and relevant event, the strong resumption of internal and external private investment in the nineties, that triggered the expansion and intensification of the ” fiscal war”. The second dimension involves the fiscal war as a peculiar form of regional industrial development policy and involves, therefore, another sphere of interaction, that between governments and the private process of spatial allocation of investment. This is the context in which we should discuss the characteristics and limits of the fiscal war as a regional policy instrument, as will be done in item 2. The basic argument developed in item 2 is that, under the present conditions of the federate arrangement, the fiscal war as a mechanism to guide the allocation of large blocks of investment is a perverse process, controlled entirely by private companies and leading to increased fiscal costs for the implementation of each project. The third dimension is related to the fiscal consequences of the subsidies used in the fiscal war, in other words, the major issue of possible negative impacts of incentives on state finances. In item 3, we propose a typology of the instruments used in the fiscal war, analyze its characteristics and support the argument that, even though the fiscal war results necessarily in a negative impact for the country as a whole, it is not at clear that the states that practice it lose individually in the medium and long run, if the subsidized projects are successful. In fact, depending on the evolution of the projects involved, the state government can end up in a better relative position than if the subsidies had not existed, even if the fiscal cost for the country as a whole is high. Item 1. THE FEDERATE DISPUTE AND THE CENTRAL GOVERNMENT’S CONTROL. The few attempts to conceptualize the “fiscal war” tend to emphasize, on the one hand, its conflictive and disordered nature: the attitude of the leaders expresses a non-cooperative behavior, whose practical outcomes converge towards local alliances, that is, they are aimed at ensuring the well-being of the citizens living in the territory in question. VARSANO (1996, p. 2) conceptualizes the fiscal war as “a situation of conflict in the federation where the federate entity that gains – when, in fact, there is a gain – imposes, in general, a loss to one or more other federate entities, since it is rarely a positive sum game”. However, the emphasis on the conflictive nature in the scope of the federation does not render an appropriate characterization of the phenomenon. It means conceptualizing the Fiscal War exclusively on the basis of the behavior of sub-national governments (SNGs), which can restrict the concept to the somewhat naive view of cooperative versus non-cooperative behavior, normally disseminated by the media. The core issue is that to fight for self-interests in a non-cooperative manner is inherent to federate agents: the federation is, in this sense, intrinsically conflictive, made up of agents that are largely (and very much so) competitive among themselves, which requires the presence of a body to regulate federate relations – the central government together with the Congress. They must both be, at least, the sponsors of a whole legal body aimed at minimizing, at least, the damages that individual SNG initiatives generate for the whole of society. It is proposed here that the focus of the analysis be shifted to the behavior of the SNGs and its implications at a wider level, where the phenomenon can be analyzed as a case of dysfunction of the political-institutional process of regulation of conflicts among regional economic and political interests. Seen from this perspective, the fiscal war is a case of a general class of phenomena that emerge when political initiatives provided by SNGs take on negative connotations and generate perverse economic effects as a result of the insufficient or conjuncturally inoperative nature of the political-institutional framework that regulates federate relations, when such framework proves to be incapable of ensuring a minimum balance among the various regional/local interests to prevent perverse economic and social effects. On the other hand, in the specific case addressed in this paper, that of development policies, the size and severity of conflicts are directly linked to the reach and power of the instruments available to the SNGs. In the Brazilian case, as we will show below, the way the tax structure created in the 1967 reform evolved, ultimately gave the state governments a powerful autonomous instrument of developmental activism. If we add to this the gradual weakening of the central government’s control capacity, we have the necessary ingredients for the outbreak of the fiscal war. The VAT taxation system and the role of CONFAZ The most obvious component of control over federate dynamics is conferred by the existing legal and constitutional provisions in the scope of the tax and budgetary system of the Federation. Although this set of provisions is much broader, we will focus on two of its aspects: the current structure for value added taxation and specific provisions that regulate the state’s competence over the ICMS. The main factor for the viability and sustaining of the fiscal war among Brazilian States is to be found in the taxation mechanisms established in interstate commerce. The Brazilian model, virtually unchanged in its bases since the sixties, has two characteristics that are essential for the problem analyzed here. First, the competence over the main value added tax – the ICMS – belongs to state governments, contrary to what can be considered the world pattern: keeping control of this type of tax in the scope of the federal government or, at least, through shared competence, ensuring some degree of regulatory influence to the central power. Up to 1988, there were still some limits to the exercise of state competence. There was a ceiling and a requirement for uniform rates, under the control of the Federal Senate. The 1988 constitutional reform increased the autonomy of the States to establish, through state laws, the ICMS rate on internal operations. The new Constitution also removed the Federal Government’s power to grant state tax exemptions and deductions, without approval by the states. Finally, the expansion of the tax base – with the extinction of the so-called “single taxes” and extension to services – resulted in the States becoming key parties in the formulation of the national tax policy, since the ICMS became the broadest tax on Brazilian domestic production and consumption. Second, the 1967 tax reform chose to adopt the origin principle in the formatting of the Brazilian VAT. This is not the place to describe the motivations for such option3. The relevant point is that it turned the previous ICM tax, since its origin, into a typical tax on production, an aspect that is completed by non-exemption for capital goods. For our purposes, the main result is that each state’s tax collection is defined by the share of production carried out in its territory, and is therefore independent from the actual payment of the tax by its inhabitants4. The direct connection between the principle adopted and the fiscal war emerges from the fact that this principle is the essential condition that allows the local government to negotiate with each company, and even the obligation to pay the tax. Even in the extreme situation where all the production is exported to other units of the federation (several cases come close to this), the government that harbors the production, since it is the legal destination of the collection, might eventually grant tax incentives, awards and exemptions. Thus, due to the macroeconomic stimulus generated, it is interesting for a state to attract enterprises from other regions, giving up its tax collection, through tax refunds, even if the company’s consumer market is located elsewhere. 5 3 On the advantages of the use of the origin principle and the difficulties in the other options, see VARSANO, R. (1979), 1980A) and (1980b) and LONGO, C. (1979), (1980a), (1980b) and (1981). 4 As long as a state achieves a surplus in its internal commercial exchanges, it appropriates a revenue share that is larger than the one generated internally by its taxpayers. In this case, the inhabitants of other states are financing the Treasury of the state that has the commercial surplus. The opposite would be true if the principle adopted were that of destination, in which case the tax, from the point of view of each federative unit, would be closer to a typical consumer tax. In the Brazilian case, the significant imbalance in the participation of the states in the generation of value added would result in the “pure” principle of origin leading to strong concentration of collection in the most developed region. It was therefore decided to tax interstate trade, through the former ICM, a combination of the origin and the destination principles, based on different rates for internal and interstate operations (the latter, albeit lower than the former, is positive). This affects the division of revenue among the consuming states and the producing states, allowing trade-deficient states to appropriate a larger share of the revenue. As highlighted by LONGO, C. (1980a) “The Tax Reform Commission expected the system of differentiated rates to assure appropriate participation in the total revenue of the States by the net importers in interstate commerce (that, generally, correspond to the less developed States of the Country)”. 5 It is important to note in this case that the approximation to the principle of destination, resulting from adoption of reduced but positive rates in interstate commerce, has different dimensions depending on the regions considered. The model was constructed in order to, on the basis of the origin principle, favor the less developed states of the north and northeast regions. Thus, the rate for trade between those regions and the south/southeast, which is much lower, ensures this appropriation. However, regarding internal trade in the south and southeast regions,, higher rates are adopted, resulting in an actual larger appropriation of the revenue by the producing state. The high concentration of markets in these regions, therefore, leads to a strong dominance of a system of interstate revenue allocation by origin. In short, the possibility of using the state competence on tax as an instrument of regional development policies was introduced from the beginning in the model adopted for the VAT. This leads to the following aspect to be highlighted here: the provisions created precisely to inhibit or restrain this possibility. In the last three decades, certain outbreaks of fiscal war type initiatives, making use of the ICMS, occurred at different times. In all of them, the possibility of intervention by the federal government was limited by the attribution of competence over the tax. Any restriction in the normative plan would necessarily have to go through Congress. Under the Military Regime, Constitutional Amendment n.1, of 17/10/69, in its article 23, establishes that the granting of state tax incentives must be subordinated to accords among the states. In the 1966-1969 period, various regional accords6 proliferated in the country, with the purpose of establishing harmonized systems of tax incentives among member states, and had the effect of forcing other states to react with similar measures. From 1970, the Federal Government began to coordinate meetings of all state finance secretaries to try to prevent the competitive processes of granting of subsidies. The restrictive legislation was later restated by Complementary Law n. 24/75, which establishes the principle that all types of exemption must result from formal decision legalized in accords entered into by the states (article 1), in a meeting attended by the representatives of the majority of the states ((article 2 paragraph 1) and upon unanimous decision of the represented states (article 2 paragraph 2), with the absentees subject to the provisions of the accord7. This law creates the CONFAZ – Council of Financial Policies – composed by those secretaries, whose main purpose was to regulate consensually the special treatments in ICM taxation. Therefore, it is clear that the legislation, while transferring to the states the power to decide upon ICM incentives, and removing this prerogative from the federal government, also proposes a rigid mechanism for its approval, which requires homogeneity in its coverage and unanimity among the states. 6 Accords of Salvador (22/11/66), Fortaleza (22/02/67), Rio de Janeiro (27-0267) and Natal I and II. 7 The 1988 Constitution, although it reduced considerably the powers of the central government over areas and issues that belong to sub-national governments – for example, removing its competence to grant state tax exemptions, provided in article 19 paragraph 2 of the 1969 Constitution – clearly maintains the same former orientation as regards state incentives. Moreover, article 152 establishes that “it is forbidden to the states, the Federal District and the Municipalities to establish a tax difference among goods and services of any nature according to their origin or destination”. A first conclusion, therefore, would be that the provisions established in Complementary Law n. 24 are more than enough to restrict most of the procedures that were, for a long time, adopted in the fiscal war: all those that support direct award of privileges in ICMS collection. The reality, however, is quite different. Complementary Law no. 24 directly stipulates the need for regulation through ordinary law, which was never done. Consequently, the existing legal rules on the subject were gradually generated in a fragmented way by means of accords entered into in the scope of CONFAZ. As far as one can observe, there was no consolidation of these norms, throughout the period after 1975. Neither is there any record, until recently, of any legal dispute over the practice of granting benefits, supported by these norms. There seems to be quite a lot of room for different interpretations with regards to what exactly a fiscal benefit is, as well as on what type of policy would constitute a benefit clearly vetoed by the norms. Until the mid eighties, apparently CONFAZ was able to exercise some control on the autonomous incentive policies of state governments, more due to the strong and watchful presence of the Minister of Finance than for the merits of its own institutional performance8. As from the New Republic, this role was gradually weakened, and state governments gradually extended the use of benefits without considering the existing legal restrictions, leading to a situation of total discredit of the CONFAZ as a body capable of inhibiting these practices. The weakening of the central government’s control power The tax and institutional configuration outlined above are enough to express the limitations of any restrictive action on the autonomy of state governments as regards tax manipulation with a view to industrialization. Other aspects were added along this period. We will now refer to the capacity held by central government – more specifically the Federal Government – in certain circumstances and historical periods, for the control it exercises on fiscal resources and allocative decision-making processes, “directly regulating” the regional allocation of public investment and, directly or indirectly, strongly influencing the allocation of private investment. This power, which was strongly concentrated in the hands of the federal government in the beginning of the Military Regime, was gradually lost by the Federal Government, along the eighties, as a result of the crisis of the authoritarian regime and the collapse of the funding pattern of the public sector in the previous decade. This pattern, which enabled the control and planning of the main investment decisions by the central government (such as the II National Development Plan), lead to conflicting interests among regional blocks becoming, to a certain extent, subordinated to the guidelines issued by the central power, through the use of financial and fiscal resources9. Another aspect, no less important, of this direct loss of “regulatory” capacity, lies in the advance of the political-fiscal decentralization process, which culminated in the increased federate autonomy established by the 1988 Constitution10. 8 Depositions gathered from technical staff working at CONFAZ at the time describe the essential role of the federal government in disciplining unruly states and minimally maintaining the dynamics of negotiation of the most important incentives. The near stagnation of the investment controlled by the federal government, mainly of the state productive sector, had effects that have not been appropriately mapped out yet, mainly in its regional distribution. The federal government’s policy of liberal cuts, in turn, has rejected a more active role through development policies at the national level. This leads to a virtual omission in the role that would exclusively belong to it, that of coordination of policies with this purpose. This set of factors, associated with the expansion of unemployment in the industrial sector, seems to be leading to a return to the initiatives of regional policy systems, reflected in the current group of uncoordinated policies that interfere in the allocation of investment without explicit and organized criteria at the national level, called “fiscal war”. 9 The data referring to public investment illustrates some aspects discussed above. It is observed that the aggregate rate of investment of the public sector (governments plus Federal companies) at constant prices dropped from 6.9% in 1980 to 3.8% of the GDP in 19909. The result of the state-owned companies confirms this trend by showing that the average of the investments of the state productive systems went from US$ 12.411 million in the 1985/89 period to US$ 6.848 million in the 1990/94 period, a 44.82% reduction (Table I. 1). In fact, there is a movement, already defined in the late eighties, of substitution of the federal government for the SNGs as more dynamic agents in the allocation of public investment (see PRADO, S. (1995)). 10 In the situation that persisted at least until the late seventies, the strong control imposed by the federal government on the allocation of tax resources, including those in principle under the control of the states – participation funds – made the SNGs dependent on a tiresome process of attracting the so called negotiated transfers (ALFONSO, J.R.R.(1994) to enable investments. This situation was an additional disciplining factor that, no doubt, contributed to the federal government inhibiting any more aggressive initiatives of autonomous policies of the type analyzed here. Item 2. ALLOCATIVE DECISION AND REACH OF THE STATE INCENTIVES The second basic dimension of the fiscal war phenomenon refers to the fact that it is a peculiar form of regional development policy. The focus shifts to the relationship between the SNG and the private process of investment allocation decision. The media often attributes to state initiatives and programs a determining role in a company’s decision on where to base its productive activity. This implicit assumption of strong interference power and leverage of the incentives is often taken to the extreme (which is clear in the rhetorical justifications of governments) of claiming that these programs create new investments that, in their absence, would not exist. However, it is obvious that the rationale of such policies is to try to intervene in a process of private locational decision process that is, essentially, subordinated to determinants of a much broader nature, related to costs, as the basic driver, infrastructure, distance from markets, availability of qualified manpower, trade union action, etc. From this perspective, it can be suggested that even the emergence of fiscal war initiatives is, to some extent, associated with the long-term trends of the locational decision process, as will be shown ahead. Assuming the total absence of discretional initiatives on the part of governments, the allocative assessment of private agents, driven by strategic, logistic and operational criteria, would result in a certain aggregate level of investment under a given regional distribution that, in a stricter sense, we could take as “optimum”11. Any alteration in this allocative profile would therefore involve, in principle, an additional “cost” that would have to be covered in order for the allocative decision to be changed. The basic argument of this item is that there are factors of a more general nature that determine a basic profile of spatial allocation, which establishes the context in which the fiscal war operates. In this context, the core element is the fiscal cost required to move the company away from the preferred location it would adopt in the absence of the incentives. Our hypothesis is that, under the current conditions in Brazil, the process maximizes the fiscal cost. 11 The sense of “optimum” here would usually be of the allocation of resources that would have achieved the best economic results for the resources available. In order to avoid the debate on the suitability of the concept, which would be inappropriate in this context, we chose to use it in the more restricted sense of the spatial allocation that, from the particular perspective of each company, would maximize its goals concerning profitability, reduction of risks and future expansion paths. Two essential questions emerge in the historical and analytical spheres: 1) What are the former and recent trends of private allocation processes that define the framework for discretional action by governments, and how these trends may have acted as additional inducers of the fiscal war? Here it is crucial to separate long-term investment trends – the problem of industrial concentration/deconcentration in the country in the last four decades – from the more recent issue of resumption of investment, mainly foreign investment concentrated in the automotive sector, throughout the Latin American continent. 2) What is the reach of the current fiscal war practices as regards expanding the aggregate volume of investments in the domestic economy and achieving efficient allocation of the fiscal resources involved? 2.1. The long-term trends The historical evolution of industrial allocation trends in the country could be divided, in the long run, into three very distinct phases and a possible new and more recent stage. The first, thoroughly analyzed12, corresponds to the concentration of the industry in the State of São Paulo. In the second phase, which took place within the 1970-1985 period, there was a clear predominance of strong deconcentration, whose epicenter is the metropolitan region of São Paulo. The attenuation and near elimination of this trend characterizes the third phase, between 1985 and the early nineties, when there was evident stagnation of the relative participation of regions in the industrial capacity. Finally, the perception is emerging, still as a vague assumption generating considerable dispute, that in the recent period a peculiar movement of reconcentration around São Paulo would be taking place. Let us consider the more recent stages of the process, between 1970 and the early nineties. 12 CANO, W. (1981). The phase that began in 1970 is typically one of deconcentration13. Attention should be drawn to some aspects of this stage that are important to understand the fiscal war phenomenon. First, the long period of deconcentration resulted in a movement of spatial allocation that privileges an absolute majority of Brazilian states, basically to the disadvantage of the states of São Paulo and Rio de Janeiro. Even so, strictly, the strongest impact is on their metropolitan regions, much more relevant in the case of Rio de Janeiro: in São Paulo, the growth of the large inland cities was not far from that of other dynamic centers of the polygonal line. This factor, even though it might not constitute an explanation, no doubt helps compose a scenario that does not encourage strong fiscal war initiatives. These would be predictable isolated initiatives on the part of those few losing states, which, in a context of strong regulatory capacity of the federal government, would be difficult to legitimize. Secondly, the existing interpretations stress the basic fact that this is not about a process of broad centrifugation centered in São Paulo, but rather the existence of significant factors of removal opposed by other restrictive and agglutinating factors, whose combination results in a concentration that extends to a bigger geographic area, a “polygon” formed by large municipalities bordering the more developed region (DINIZ, C.C.(1991). One can picture that around the metropolitan region of São Paulo, curves or zones of equivalence would be defined, from the point of view of private allocation decision, so that various alternative allocations would present an almost identical cost-benefit configuration regarding access to markets, infrastructure, etc. Close to the epicenter, these curves would indicate increasing allocative advantages as they move away. From a certain point, however, the diseconomies would predominate, defining a strong discontinuity that corresponds to the notion of “polygon”. We will later return to this image to discuss the reach of fiscal war initiatives. Moving on to the next stage, the available data shows that, between 1985 and 1990, the alterations in the regional distribution of industrial activity were insignificant. As for the economic nucleus of the country, the basic relative positions were maintained. Recently, various experts have suggested the possible reversion of the deconcentration process. The possibility of such dynamic inflection in the locational context is an incipient point, still lacking further debate. The largely unexpected success achieved so far by the MERCOSUR takes on an increasingly prominent role as a factor that induces a pull towards the south. On the other hand, the recent process (nineties) of opening of the economy also brings other opposing factors, such as the attractiveness of locating exporting plants in the north/northeast, given the closer access to markets, particularly NAFTA. At any rate, such development, if effective, would rekindle old conflicts among the macro-regions of the country14. The most recent interpretations of this phenomenon associate it with factors of a strong structural nature, derived from the major changes in course in the Brazilian economy. From this stance, given the nature of the determinants, there would actually be the possibility of reversing the process, with the economy resuming a movement of reconcentration around the developed region. Both Diniz (1995) and Cano (1995), for example, point to a reconcentration trend of the so-called service industry in the more developed area of the country. This recent movement is related to technological changes and productive reorganization, which tend to alter local requirements, especially of the more expertise-intensive activities. Moreover, the Mercosur also plays a decisive role in this process15. Recent data, raised by the NATIONAL INDUSTRY CONFEDERATION, record a clear growth in the participation of the Southeast region in the GDP. Between 1985, approximately when the deconcentration halted, and 1995, the region grew from 58.18% to 62.6%. As shown in the table below, São Paulo accelerated its growth resumption in the second lustrum, while Rio de Janeiro, which had isolatedly persisted on a stagnating path in the first period, reversed this trend dramatically. 13 The State of São Paulo suffered an overall drop of participation in the industrial product from 56.4% in 1970 to 49.1% in 1990. In turn, the metropolitan area of São Paulo witnessed a reduction of its participation in employment and industrial production from 34% and 44%, respectively, to 25.2% and 26.3%, between 1970 and 1990. In parallel, the economy of the state of Rio de Janeiro comes up as the second major “loser” in the period, having its participation in the value of industrial transformation reduced from 15.7% in 1970 to 9.5% in 1985. 14 A clear indication of this is the recent case of the Provisional Measure of the automotive sector, ironically a typical case of interregional fiscal war operated through the federal government. 15 Cano (1995) draws attention to the following factors that depressed the peripheral domestic growth and reversed the deconcentration: i) the dismantling of the National State, and, therefore, of the regional development policies; ii) the trade opening; iii) the maturity of the investments of the II National Development Plan; iv) the drop of federal and state investments, derived from the financial crisis they were going through; v) the reduction in the growth rate of exportations; and vi) the concentration in São Paulo of the computer, microelectronics, telecommunications and automation companies, which Cano identifies as the reconcentrated form of the current productive restructuring of the country. TABLE I BRAZIL – PARTICIPATION OF REGIONS IN THE GDP 1985 1990 1995 BRAZIL 100 100 100 NORTH 4.35 3.48 3.24 NORTHEAST 3.55 13.18 12.58 CENTER-WEST 6.24 5.92 5.86 SOUTHEAST 58.18 60.79 62.60 MINAS GERAIS 9.67 12.49 13.12 ESPÍRITO SANTO 1.67 1.71 2.10 RIO DE JANEIRO 12.78 10.89 13.17 SÃO PAULO 34.07 35.70 37.45 SOUTH 17.69 17.34 15.72 RIO GRANDE DO SUL 7.93 7.00 6.61 SANTA. CATARINA 3.50 3.32 3.37 PARANÁ 6.25 6.31 5.95 Source: CNI – National Industry Confederation Nevertheless, although the idea of reconcentration is questionable, the point referring to the stagnation of the deconcentration process is consensual16. This means, essentially, that an important driving factor of peripheral regional economies, even the most advanced ones, was halted. This is, again, a very general factor, but that cannot be rejected in the attempt to outline the conditions that lead to the worsening of the fiscal war. As the picture of stagnation and crisis advanced in the eighties, the first manifestations of individual initiatives of the federate entities began to appear, aimed at maintaining their relative positions in the distribution of the industrial activity. 16 It would be possible to propose another set of explanations, derived from the more conjunctural aspects that marked the second half of the eighties. This period witnessed the most dramatic stage of instability in the long process of recent crisis in the Brazilian economy. It also witnessed, as already mentioned, the most drastic and rapid reduction of public investment in all the recent history of the Brazilian economy. After 1986, the economy took a swinging path, where the horizon for forming expectations was drastically shortened. Private investment, after some recovery in 1986, also took a clear declining path. In this context, it is viable to assume that, without damage to the structural factors mentioned, a long period set in when the defensive positions on the part of most of the private sectors privileged strategies to expand capacity with a view to optimizing existing plants and/or modernizing and increasing productivity, in a “non-optimizing” option under certain aspects, but minimizing the risks of further immobilization of capital in a context of high uncertainty. This would result, in certain inductive factors of deconcentration losing importance in the private allocation decision. Resumption of private investment and the regional development policies The previous sections featured the general conditions that opened the possibility for emergence of autonomous local policies to create regional development programs. There are records of similar processes taking place in previous decades. However, it can be claimed that at no moment the phenomenon reached the magnitude that it has today. In the period between 1975 and the late eighties, the information available suggests a clear loss of drive of these development policy initiatives17. There are still state-level industrial development programs. However, there is no record of any conflict among SNGs that suggests the competitive use of benefits and incentives, such as had occurred earlier and would occur in the recent period. Should this be interpreted as an indicator of the capacity of the federal government and the Congress to establish limits and restrictions over conflictive and disordered practices by the federate agents? A more thorough analysis of some aspects of the Brazilian economy in this period does not provide support to this hypothesis. It is characterized by a strong reduction of aggregate investment, caused not only by the autonomous reduction of a large share of internal private investment, but also by the negative effect derived from the sharp reduction of state investment (that inflects in 1981/2 and reaches its lowest point in 1993) and, finally, by the reduction to extremely low levels of direct foreign investment along the eighties, with a decreasing trend as the macroeconomic scenario deteriorates in the second half of the eighties. 17 For details on the historical evolution of the state development programs and of the fiscal war, see the complete work that this article draws from: PRADO, S./CAVANCANTI, C.E.(2000). In this context, in which the economy tends to operate predominantly through production cycles18, without any strong resumption of investment, a relative stagnation of the locational profile of the productive capacity, mainly industrial, might have taken place. The aggregate data available on the trends of regional deconcentration of the Brazilian industry, as we saw above, suggests that the strong process of “centrifugal” movement of industrial production, away from the metropolitan region of São Paulo, was virtually interrupted in 1985, and the nearly stagnated distribution persisted for the remainder of the decade, at least regarding the distribution among states. It appears that a significant movement of the industrial production towards the inland occurred in each state, and mainly in São Paulo. If this hypothesis is correct, it would explain how the apparent disappearance of the most aggressive initiatives of regional policies of attraction would stem much more from this virtual stagnation of private investment and temporary stabilization of the spatial distribution of production, than from the effectiveness of governmental controls. It should be noted that, since the beginning of the eighties, the movement of reversion of the authoritarian centralism was gaining strength, as well as the increasing financial and decision-making autonomy of the SNGs, which culminated in the 1988 constitutional reforms. Even in this context, the eighties went by without any record of conflict around investment decisions. Drawing from the argument above, the decisive factor for the re-emergence of the phenomenon was the resumption of private investment in Latin America in general since the late eighties, and in Brazil as from 1993, and the resumption of internal investment as from the 1994 stabilization. Direct foreign investment in Brazil during the eighties presented a typically restrictive behavior. After sharp reduction of the entry flows, the net result was negative. As from 1992, however, it accumulated a net result of more than US$ 18 billion. Although a large share of these resources is associated with the process of privatization of infrastructure sectors, there is still a clear process of resumption. The most obvious side of these flows has been the entrance of large companies in the sector of hard consumption, mainly in the automotive sector. The combination of the relative success in the stabilization of the main economies of the continent – all supported by aggressive policies of trade and financial opening – with the creation of MERCOSUR has induced a reorganization of the presence of foreign capital in the region, with the entrance of new agents and expansion of those already located here. 18 That is, demand oscillations primarily attended by the variation of the level of occupation of the installed productive capacity. As for internal investment, there are still strong doubts regarding the possibility of a cycle of expansion of productive capacity – the investment has been, so far, much more focused on specialization processes, reduction of bottlenecks and modernization aimed at productivity gains19. Nevertheless, there is no doubt that the adaptation to the new context generated by stabilization lead to a cycle of expansion in the sectors that benefited most from the increase of internal demand, apparently changing the situation of relative stabilization of the profile of spatial location of production. The sector of production of consumer goods, in general, is seeking to increase productive capacity through new plants. For the purpose of this paper, there is a particularly relevant aspect in this process: there was apparently, in the 1994/1997 period, at least in some sectors, a strong temporal concentration of investments, determined either by the requirements of new levels of consumption after the Real Plan, or by the search for rapid inclusion in highly competitive markets that began to be viewed, in a short period of a few years, as more attractive, due to factors such as the stabilization and the MERCOSUR. In this case, there was a “round” of new investments resulting from the entrance of new agents and adaptation of the existing ones to the new evaluations of the private sector. From this point of view, the position of the SNGs in attempting to interfere in locational processes of this circumstantial flow of investment through aggressive autonomous policies could be seen as rational. Even if stabilization is sustained and the good image of the country among the so-called emerging markets is maintained, it is clear that, in many sectors, investment intentions already announced will cause a significant leap in productive capacity, notably in the automotive sector. Once new large plants are established, it would be reasonable to assume that a large share of future expansion would occur through their expansion or, at least, through a strong locational pull around them, as the case of FIAT in Minas Gerais demonstrates. 19 For a recent and not very optimistic assessment of these trends, see Editorial de INDICADORES IESP no. 50, March 1997, by Mariano Laplane and Fernando Sarti. 2.3. Limits of the fiscal war: the impact of incentives on the local economic space Now let us address the second set of issues from the perspective of the relationships between SNGs and private investment decision. In recent work, Prof. Ricardo Varsano addressed rigorously and competently the core issue involved in the fiscal war: the combination of a strong negative impact at the aggregate level with the generation of undeniable positive effects on the local economy (VARSANO (1996)). The Author provides a thorough explanation of the basic conditions that, from the standpoint of the theory of public finances, would justify the granting of incentives and demonstrates that, from the viewpoint of the local economy and society, most incentives meet these conditions. This is, no doubt, the core issue in question, since it removes the more simplistic assumptions on the perversity of the fiscal war, and presents it as a logical regional development policy alternative within a system that is not, in principle, intrinsically cooperative. Varsano’s formulation suggests an approach that can be useful in the unfolding of the issue: to try to identify the basic conditions that would allow advocating fiscal war-type policies (i.e. a model supported by the total autonomy of the SNGs to decide upon the allocation of their fiscal resources in attracting investments, without any interference from the central power) as a second best for addressing the issue of regional development at the national level. The argument here is that, even though it is possible to come up with theoretical situations where these policies would present some positive aspects, the necessary assumptions for this are heroic. In a more realistic assumption, policies such as the ones in course in the country today maximize the fiscal cost of an aggregate volume of investment that is highly inelastic to incentives, therefore, being the worst possible alternative for the intervention of the public sector in the process of private investment. The case for the fiscal war Let us generically consider any situation involving a specific state (endowed with certain natural and locational resources, infrastructure, fiscal and financial capacity, etc.) and a formal business decision to establish a plant of a certain profile and size, in search of the best location alternative. Let us take for granted that the company has previously done a careful analysis of the locational alternatives20 that resulted in something similar to a map of what we call above equivalence areas. In the assumption that its preferential allocation is not in the state in question, there is necessarily a cost involved in changing the preferential location. Here an important point emerges. The worse the position of a state – in general, of its more attractive municipalities – in the company’s ranking, the greater the overall cost involved in tax benefits necessary to get the company to move away from the “optimum” location determined by its previous assessment. At the extreme, absolute restrictions emerge in the simple example that there would not be incentives big enough to make a large assembly plant move to a small state in the north region. It should also be noted that, given the preferential private choice of the company, the tax benefits have not only to cover the additional cost involved in moving, but has in fact to surpass them in order to offer a prize. If not, the company would be exchanging a “natural” optimum situation for an equivalent one, but dependent on public sector support, which obviously entails additional risks. Although it cannot be taken for granted, let us also assume that the government of the state puts into practice a detailed analysis of the impacts generated by the investment in question, carefully assessing the cost-benefit elements involved. In a simplified manner, this process involves assessing two major groups of effects. On one hand, the net fiscal cost of the incentives from an inter-temporal perspective. This is crucial, because typically in all regional development policies – and the fiscal war is not an exception – there is a concentration of negative fiscal impact in the initial phase, with possible and desirable future partial recovery of the resources invested. On the other hand, it is necessary to assess the overall benefits generated by the new investment: direct and indirect creation of jobs, induction of complementary investment (automobile parts, for example), induction of expansion and diversification of the tertiary level, etc. 20 In principle, the bigger the sunk costs involved in the investment the bigger the weight of structural factors in the selection and the higher the compensation amounts that the company will demand to move away from the preferential choice. As an example, compare the establishment of an automobile plant in a region where the sector does not exist, with all the requirements of development and consolidation of a whole supporting industrial structure (case of FIAT in Mines) with the establishment of a typical camouflage company, operating with low density of capital and supported by importation of most of the final added value. From a strictly technical point of view, the government’s assessment should confront the expected benefits of the investment with the basic cost involved in the actual moving. Under the strong assumption that the state decision is guided by the most rigorous technical assessment possible of these factors, the decision would be to grant the benefits in the amount required to change the decision of the company,, provided the result of the inter-temporal cost-benefit assessment is positive. An essential component of these assumptions is that the state government has some possibility of estimating, at least approximately, the effective cost of removal established by the private company’ assessment. It is possible, in an effort to defend the case for the fiscal war, to indicate some positive aspects of this model, mainly as opposed to a centrally planned and controlled industrial policy. It maximizes the use of the information resources and expertise of the local technical bureaucracy regarding its own regional economy. It has, in principle, one of the basic advantages of federate decentralization, which is more proximity and subordination of the decisions to the interests of the local society. It induces greater development of local institutional mechanisms of support and interaction between the government and the private initiative, and can thus increase the efficiency of the integrated action of the state and municipal governments (there is no doubt that this is taking place in all the states that have adopted aggressive fiscal war initiatives). It should also be observed that, in a situation where the central government of a country with high regional disparities abandons any concern with integrated regional development policies, the state governments, particularly those relatively behind economically, are legitimized in operating policies with this format. It is our assumption that, if a careful assessment of many of the projects currently involved in the fiscal war were possible, in the terms above, no doubt the conclusion would be that the net result for the local economy and the country is potentially very favorable. The limits of the fiscal war The essentially weak point of this construction is that all the assumptions necessary to guarantee the efficiency of such decision process on allocation of fiscal resources is heroic and unreal. This results basically from three factors. 1) The state governments do not know the profile of the allocative preferences of the companies. It is an essential element of the private strategy in the fiscal war to build the image of a “draw” between alternative locations, so that the final decision will depend on competition between incentives. As a result, in the absence of a higher level with political capacity to intervene, the initiative in the interactive process that leads to the decision will belong to the private sector. This in general reduces the possibilities that a state, relying on its “natural competitive advantages”, will give up participating aggressively in the incentives auction. 2) It is unrealistic to presume that the decision to abandon the dispute will be taken if the strictly economic cost-benefit relation is unfavorable. The point, as noted by Prof. Varsano, is that the benefits generated by the investment often include factors that are hard to weigh, such as regional prestige according to popular image, expressed in increased political prestige for governments that manage to bring major industrial plants to the state 21. Thus, the limit for the granting of incentives tends not to be based on the cost-benefit assessment, but rather on an assessment based on very broad and generic “well-being” parameters, incorporating the particular interests of the political groups in power. Finally, the enormous technical difficulties involved in the prospective identification of the benefits and the technical incapacity of bureaucracies also hamper this analysis. 3) Finally, the competitive nature and absence of any higher level of arbitration, combined with the two previous elements, leads necessarily, in the strong sense that it is the most efficient rational position on the part of the agents, to an interactive dynamics where the final result is the maximization of the benefit by successive exclusion of participating states, as the bet surpasses what is allowed by their “expanded cost- benefit assessment ” and/or their budgetary limits. 21 In this case it should be remembered that this is the same story that was found in the case of the development of some of the major blocks of the State Productive Sector in Brazil. In the fifties and sixties, the steel, telecommunications and electric sectors were completely or predominantly implemented by state initiatives, in general supported by political campaigns that emphasized the “industrial vocations” of each state. (For a reconstitution of these aspects, see PRADO, S. 1995). Therefore, even in the cases (probably very frequent, even though absurdities exist) where the net benefit for the local economy is positive, it is highly likely that these investments are being allocated at a maximum fiscal cost, without the additional spending resulting in any additional gain for the country. There is another basic aspect to be added to the previous comments. The validity of the usual argument in the rhetoric of fiscal war operators, that the incentives “create investment” that otherwise would not occur, is highly questionable. Our assumption is that, in almost all relevant cases, the investment decision is taken previously, derives from long-term business strategies and would tend to be carried out even in the absence of incentives. The case of the automotive sector is enlightening. The block of recent investments results from the situation of the markets in developed countries and of the need for major world producers to consolidate their position in view of the growth potential of Latin American markets. Likewise, the major internal investments are determined by the recent expansion of the markets and by the search, on the part of the companies, of more favorable labor costs. We find it very likely that, in most cases, the investments would be carried out even in the total absence of incentives. Obviously, a distinct locational profile would result, probably more concentrated. The whole paraphernalia of tax incentives and their maximum fiscal cost tends to have, in fact, the basic effect of changing the locational profile of a given volume of investment. The arguments above require a qualification. The wide use of state incentives, in a random way, within a competitive context, and at an increasing scale, can, from a certain point, have the actual effect of marginally expanding the volume of investment, particularly in a context of opening of the economy and significant flow of direct foreign investment. In fact the fiscal war would reduce the costs of the investment in the national space. There would be some sense in the argument, often brandished by state technical experts in the defense of their programs, that the programs affect external investment decisions and that, in their absence, other countries of the Mercosur would be sought. On the other hand, even if this were true for a share of the investments, it is clearly the most perverse and least efficient way to achieve results that could equally be achieved22 under a system which is more orderly and subject to the minimum criteria of priorities and regional comparative advantages. In fact, this effect is closer to the one that would be achieved if the incentives policy were controlled by the federal government, and the federal government decided to expand the benefits to attract more investments. This is a relevant factor when we consider the relative omission of the central government in Brazil today. Since random and deregulated state initiative occupies the “empty space” left by the federal omission, the result is the most perverse of the systems . These aspects raise the issue of distinguishing, in this context, two types of incentive programs. In the cases above, where the investment decision is taken on the basis of internal or international strategies of expansion of major groups (and that we assume to be widely dominant), it is obvious that, essentially, the incentives policy will essentially affect the allocation of the investment to alternative points of its equivalence curve. There are other cases, however, where it can reasonably be admitted that the program can determine the investment decisions that would not occur without it. An example can be the case of sector incentive programs that explore comparative local advantages and have the effect of strengthening (although not exclusively) local investment initiatives, such as incentives in the tourism industry in the northeast. The general conclusion that can be drawn from these arguments is that the potential of the fiscal war for leveraging new investment basically depends on the nature of the economic sectors involved and, mainly, on the relative size of the beneficiary companies. The more the state policies aim at attracting the investment of the large national and international conglomerates, the higher the probability of the policies having their reach strictly limited to the spatial location of the previous amounts of investment. 22 It is hard to assess the overall net result of the advantages and disadvantages that exist today for the large international capital in the choice between Brazil and Argentina. If, from the private point of view, some aspects favor the neighboring country – for example, the more advanced stage of the liberalization policies – the fact remains that the Brazilian economy, for its size, defines the paths of this regional system. Regardless of the progress of the Argentine reforms, if the Brazilian stabilization fails, there is a huge possibility of Brazil dragging with it the countries that depend on its market and currency exchange policy. Item 3. FISCAL WAR, STATE FINANCES AND TAX REGIME The arguments of the previous item clarify the generalized adoption of fiscal war practices by a large share of Brazilian states, even though, as proven before, these programs have a truly perverse and negative net result in their overall effect on the national economy.. This has been leading to a growing debate on the need and the alternatives of intervention by the central powers – Congress and the Executive – to neutralize their impact. One of the recurring issues concerning the fiscal war is related to its impact on the budgetary situation of the state governments. The relevant questions are: 1) Does the granting of incentives lead to a deterioration of the fiscal situation of the state that practices it? If true, is this impact immediate or medium/long term? Wouldn’t the main form of deterioration occur with the loss of ICMS collection? 2) If the fiscal war involves a negative fiscal impact, would it not be a phenomenon doomed to disappear? 3) Does the granting of incentives affect the finances of the states that do not practice it? One of the main instruments used in the fiscal war is the ICMS collection. Since this source of resources is so important in the financing of the states – which applies precisely to the states that we analyze in this paper – the conclusions regarding the perverse consequences of the fiscal war for their fiscal balance are straightforward. This discussion is only a limited aspect of the more general issue that was already addressed above, that the fiscal war leads, in general, to a negative aggregate net effect although, for each individual state, it can have positive net results. In other words, the fiscal war is a process of alteration in the allocation of productive capacity, employment and tax revenue among federated units, where the whole of the country loses, but there is nothing to prevent some from benefiting individually. This issue requires some thoughts on the current foundations for the intervention potential of the instruments in the hand of the states, on their reach and on how long they will remain strong. To address this theme we will first present a typology of intervention instruments, based on which we will be able to discuss the characteristics of the programs that employ them and their relations with the fiscal organization of the SNGs. 3.1. A typology of intervention instruments The generation of different incentives that increase the flows of investment to a unit of the federation can be achieved by means of action focusing on the following factors: Systemic aspects 1. Processes of deregulation and privatization that reduce costs of transaction, inputs and services, and restrictions of access to markets. 2. Public institutions dedicated to regional development, applied research, centers of support to private activity. 3. Investment in economic and social infrastructure. Selective benefits 1. Specific locational advantages for a company: local infrastructure improvements, guaranteed supply of raw materials and inputs, land, port facilities, etc. 2. Reduction or deferral of tax burden by company, with or without implicit credit subsidy. 3. Granting of subsidized credit. 4. Shareholding. As for the systemic incentives, the presence of a policy to attract investment would be expressed in the existence of a bias in the processes of allocation of public spending on investment, as well as in the institutional reforms that would guide them towards the creation of an environment favorable to private investment in areas and sectors regarded as having greater potential in the state/region. In this sense, it should be noted that there is a whole significant dimension to regional development policies – focused on attracting federal and international resources for investment in economic and social infrastructure – that involve strong medium and long-term impact on the regions’ systemic competitiveness conditions. The fact that some states of the federation have been more successful in accessing the resources of multilateral organizations such as the IBRD, for example, has sizeable impacts on their possibilities of attraction. However, these systemic factors, due to their diffuse nature, tend to have a slower impact on the private allocation decision process. They constitute the defining elements of the dynamic competitive advantages, and define, in fact, together with other structural factors, the level of possibilities based on which the use of specific instruments can maximize the volume of investments seeking space at the state level. It is possible, in principle, to ground an investment attraction policy entirely on factors of a systemic nature. In the current Brazilian case, this would mean focusing on the expansion of infrastructure with a view to local comparative advantages, improvement of efficiency of utilities that are still public or their privatization, development of more efficient forms of integration between state and municipal governments in the creation of information systems and support to private investment. The model adopted during most of the recent period by the state of São Paulo is very close to this ideal type23. The main “negative” aspect of this option is, obviously, that it cannot achieve short-term results. In situations – as seems to be the current one – where there is a sharp increase of investment with strong indications of high spatial mobility, these less aggressive policies can lead to losses for the state/region, most certainty in the short run, with uncertain likelihood of long-term recovery, given the synergic effect generated by the differential allocation carried out in the other regions. The specific instruments: fiscal capacity and tax structure The use of specific factors of attraction can be broken down analytically into two typical-ideal categories, which can be very useful particularly for the observation of their historical evolution. A first ideal format would be the organization of instruments into formal broad programs, with clear definition of access conditions to each benefit, available to a broad range of companies and without any provision allowing discrimination among them. This pattern corresponds, roughly, to what was in place in Brazil in the seventies and eighties. Its main aspect is that it involves, in principle, prior definition of the type of benefit involved, as well as clear definition of the criteria that link the awarding of benefits to a range of sectoral and sub-regional allocation priorities. 23 For a detailed analysis of the position of São Paulo in the context of the fiscal war, see PRADO, S. and CAVALCANTI, C.E. (2000) CAP 3. The second format, concerning ideal types, could be called fiscal-financial operations dedicated to specific projects. This is the case when the local government designs a set of benefits agreed with the individual private agent, involving an interactive process between the characteristics of the investment and the benefits granted24. The main aspects of this option are the maximization of flexibility and response time, coupled with little transparency in the process. It should be noted that this format tends to be suitable in cases where large investments are negotiated, as is the case of the recent expansion of the automotive sector. A basic aspect of the evolution of the fiscal war in the recent period, as we will see below, is undoubtedly the gradual transition from the typical traditional program model, to an increasing predominance of dedicated operations. This is clear in some cases, where programs are created with the whole typical format of the traditional ones, but their basic purpose is to serve specific large projects. At the same time, there is a clear trend towards protecting these programs with all the legal guarantees that allow a reduction of the political risk of subsequent interruption of the benefits due to changes of government. In the case of large projects, which are difficult to reverse without high losses, the agents tend to demand from governments the formulation of detailed contracts that entail binding commitments. From the point of view of the recent process, it is important to assess the existing potentialities in the use of the specific instruments. The power of these instruments depends, basically, on the following factors: 1) budgetary conditions to reallocate state revenue towards expenditures with benefits. This involves not only state budgets, but also the situation of the states’ fiscal accounts. Even states with high participation in national public expenditure and revenues can have circumstantially some limited autonomy to operate incentives, due to a situation of strong fiscal imbalance and high indebtedness. 24 As an illustration, companies in Brazil often reformulate the size of their investments – for example, anticipating longer-term investments – according to the volume of benefits granted at the present moment. 2) Existence of, and capacity for, control of the SNGs over public agencies and companies in infrastructure and basic input services. This factor has proved to be particularly important in the case of ports, agencies involved in road construction and state-owned electricity companies. It is obvious that the recent movement extending privatization processes to the SNGs requires a reduction or even cancellation of this type of instrument, forcing the budgetary reporting of these benefits, which today are diluted in the investment programs of the state-owned companies. 3) Direct availability of funding sources – state-level public financing sector – or access to the federal and international credit sources. The same remark made in the previous item, regarding state privatization, applies to the state-level financing sector. 4) Tax competence over relevant indirect taxes. Most of these factors can be reduced to the current spending power of the SNGs, either through central budgets, or through state agencies and companies. It is clear that, although in a mediated way, the financial leverage capacity is also associated with that main factor. This would lead to the conclusion that, the bigger the spending power of the state, subject to some financial order, and the bigger its availability and control on infrastructure generators, the greater its power to induce investment. The other side of this proposal is that, as SNGs go through periods of budgetary imbalance and are, at the same time, submitted to restrictions of access to additional credit, their power in the use of incentives tends to fall. Many observers see in this aspect one of the factors that can lead to the fiscal war losing strength naturally, as revenue waivers begin to affect budgetary situations that, in general, have been far from balanced in the country, and increase the alternative cost, also in political terms, of privileging spending on incentives. This conclusion would be flawless were it not for the peculiar nature of the phenomenon that we are analyzing, and that derives from the aforesaid ICMS state competence and the configuration adopted in what is known as Tax Adjustment at the Border (AIF): the prevailing origin principle. The crucial point is that, in these conditions, the necessary result of the investment that is the object of the incentive is an immediate and perfectly measurable increase in tax collection by the government that grants the incentive. Thus, the budgetary base pertaining to the relevant period of the incentive is not established in advance: it is affected by the object of the incentive, as the local industrial production increases. In these conditions, the SNG has a powerful instrument of incentive through the waiver of future revenue generated by the plant in question, which does not affect the level of its current revenue. Secondly, depending on the size of the new plants established, and particularly in the case of large industrial complexes such as automobile assembly plants, there will be inducing effects over new investments by suppliers and service providers. As long as the government does not subsidize these sectors as well, a rise in collection will occur in relation to the level that would exist without the subsidized project. That is where one the most peculiar and perverse aspects of the current process of fiscal war resides. It can be claimed that, if the project as a whole is successful, there should be no loss of collection (compared to the previous situation) during the period in which the company is subsidized and, moreover, after this period, the fiscal situation will probably improve, from the strict point of view of tax collection capacity. . The graph above, indicating the evolution of ICMS collection of a hypothetical state, serves as a schematic reference to assess the issue, in the case of a large investment project. The bottom curve (a) projects the collection that would be achieved if the investment did not occur. In the case where it occurs, from the moment (time 8) when the company initiates its operations and sales, the potential revenue rises to another level, represented by curve (c). Assuming, as in the current practice of incentives, that all the incremental collection is deferred through subsidies, the actual revenue would continue to be represented by curve (a). As the driving effects associated with the project begin to occur (time 10), they would cause additional revenue gain. This is a gain derived from the main project, under the assumption that the state will not subsidize these new investments as well (typical examples are the suppliers of parts of large assembly plants). cba From the point of view of the national economy, the fiscal cost is defined by the difference between the potential curve and the actual curve. From the perspective of the state economy, in comparison with its previous situation, strictly regarding its tax revenue, there is no direct loss. On the contrary, the state will have a revenue increase due to the effects of the resulting investment. In other words, from the point of view of its previous potential revenue, there is a gain that results from the additional revenue that the investment brings with it, represented by the difference between curves (b) and (a). Let us assume that, as from time 16, the deferral benefits are exhausted and the company returns to the normal level of ICMS collection. In this case, the actual revenue equals the potential revenue, at a level higher than the revenue that is possible without the investment. This example demonstrates a basic point: from the perspective of the awarding state, there is no medium or long-term loss of revenue compared to the hypothetical situation of the project not being carried out. During the period of subsidy provision, from the point of view of the temporal behavior of the collection-spending sequence, there is evidently a reduced revenue elasticity in relation to the local product, as a result of the tax waiver. In addition, there is the set of direct fiscal expenses required by the project (infrastructure, land, etc.) If the demand for current expenses and investment expenses of the SNG is correlated, to some extent, to product level, this would indicate a worsening of the fiscal situation in the medium run, or, alternatively, the deterioration of the public services provided. In the first case, the period of provision of subsidies, mainly in the cases of more aggressive state policies, might result in a deterioration of government’s finances, through increased public debt. Since ICMS deferral and credit operations are subsidized, the future collection and return of these resources will not necessarily cover these liabilities. This conclusion, however, must be hued by some factors. First, for many states the share of resources derived from the overall tax base, through participation funds and other transfers, is significant, which reduces the vulnerability of the revenue to loss through incentives (in fact, as many critics point out, some states rely on revenues from transfers to expand their granting of incentives). Second, there is the important aspect of the local impact of successful projects. From a dynamic point of view, a successful program of incentives should generate a set of direct and indirect impacts on local income, through subsidiary companies, suppliers of parts and equipment, rendering of services, etc. All this has consequences that are hard to assess, but are certainly positive, on tax collection levels, since it affects a wide range of sectors that do not rely on tax reduction. It is true that some states (Paraná, for example) have extended the benefit of deferral to suppliers that move to the state trailing behind the main company. This, however, has not been the rule. In most cases, an immediate increase in tax collection is associated with the direct (deferred) additional revenue of the subsidized project. The main implication of these facts is that each state’s capacity to grant incentives becomes, to a large extent, independent from its spending power and even from its conjunctural fiscal situation. This leads to the following extreme hypothetical situation, which is highly enlightening. In principle, disregarding other conditions surrounding the private allocation decision, the state of Rondônia would have, as regards manipulation of the tax burden, the same power as any other state to offer incentives to a company. 3.2. Alternative mechanisms for the use of tax competence As analyzed in item 1, the use of ICMS reductions through the individual initiative of each state is, in principle, vetoed by current legal provisions. It is important to note, however, that the only truly necessary condition for the states to be able to use the instrument is the maintenance of the origin principle, that is, ensuring the appropriation of the revenue generated through the operation of the new plant by the state that harbors the investment. There is no legal impediment for a state to grant credit or financial benefits to a company that establishes in its territory. Thus, there is nothing to prevent, for example, the granting of subsidized credit for support and complementation of the company’s working capital in the first years of its operation. In this case, the company would actually pay the ICMS, and the requirements provided in CONFAZ law would not be disrespected. It is obvious, however, that this opens a wide possibility for the ICMS reduction to be camouflaged by a triangular operation involving, for example, the company, the state Treasury and an incentives fund for state development. The basic pattern of this operation would be the following. An incentives fund is created, fed by budgetary resources and other usual sources, to grant funding for, say, working capital. The amount of credit is defined so as to correspond, roughly, to the expected amount of ICMS to be paid by the company25. In the case of a new enterprise, the generated tax is usually entirely refunded. In the case of expansion, the credit reimburses the so-called incremental ICMS. Thirdly, it has to guarantee the necessary transfer of the resources obtained through tax collection to the incentives fund. Thus, the final result is the same as that of granting the incentive, but it is achieved indirectly in a way that contemplates the actual collection of the tax. This model of incentives would have as a basic characteristic the elimination of the explicit use of the state competence over the ICMS, and the transition to the use of: 1) traditional fiscal funds, whose basic funding is made up of budgetary resources and loans. In actual fact, ideally, the fund would be fed by the incremental tax collection generated by the project. This leads to safe and automatic operation. 2) use (not explicit in the budget of the funds and of the government) of the spending capacity of state agencies and state-owned companies, to create infrastructure facilities and benefits. The general base of the programs could be described as composed of two types of flows of fiscal resources applied in benefits. One part, that could be called rotating, is in fact a disguised ICMS deferral operation, inbuilt in the granting of financing for the working capital. The final result is that the company ceases to exist from the perspective of ICMS collection. The state ceases to collect the new tax that is generated, thus not incurring in any reduction of current revenue. As long as the mechanism for transfer of the ICMS revenue from the budget to the fund is guaranteed, the mechanism does not require any credit support from the state. The possible losses in this scheme, from an inter-temporal point of view, are restricted to the establishment of subsidized interest rates. The other part can be considered, at least in the short run, as non-returnable expenditure: direct benefits and donations such as land, infrastructure and other facilities. Moreover, it is not possible to consider all these expenditures as costs of the enterprise for the state, since many of them (transport infrastructure, for example) generate benefits and positive externalities for the community. What are the intrinsic limits to these two basic sources of subsidy provision? 25 This explains some “magic figures” found in certain contracts of the automotive sector. In general the subsidy is granted over a percentage of the invoicing that coincides with the attribution of the ICMS rate to an estimate of the value added of the company. The first has a “natural” limit in the ICMS collection and, in fact, since it is not characterized by a high degree of subsidy provision, actually represents a deferral of the tax, which can mean a financial relief for the company in its implementation stage, but can also mean a financial overburden in the future. On the other hand, if the working capital is financed without price level adjustment and, as often occurs in the current processes, without interest, the loss in the period of subsidy provision will be even greater. The second depends directly on the state government’s spending capacity, the financial situation of its decentralized agents and the possibility of alternative sources of financing for the other necessary expenditures of the state, allowing reorientation of budgetary resources to the fund. By way of conclusions Now let us go back to the more general issues with which we initiated this item. First of all, attention should be drawn to the fact that we are discussing exclusively the relation between fiscal war and budgetary balance. Thus, we are completely overlooking the problem of whether the fiscal war brings extra-fiscal benefits in terms of employment generation, modernization of local productive structures, etc. Even if the impacts of this kind are very positive, the point is whether the process to achieve them introduces a deficit bias in public accounts. There is no actual evidence, either from the wider historical point of view, or in more recent events, that the fiscal war necessarily involves the worsening of the fiscal situation of the individual states that practice it, although it must be acknowledged that it is still too early for this assessment. However, there does not seem to be, from this perspective, any inexorable trend towards weakening incentives policies. With this, all we mean is that, as long as extreme benefit granting policies are avoided – as has been occurring in some recent cases involving the automotive industry – it is possible that the medium and long-term net result will be positive for individual states. This makes criticism to the process based on the individual situation of the states particularly problematic. Otherwise let us consider the following. We indicated above two aspects to be considered. First, whenever the incentives are decisive in changing the allocation decision, much of the fiscal cost involved in them refers to incremental revenue manipulation that, without the benefits, would not exist in the state coffers. It should also be noted that the fiscal expenditures related to infrastructure and similar projects, even if they are prioritized because of a specific company, generally have a positive overall impact for the region, even if, from the strict view of the marginal benefit generated, they were situated at a lower level on the scale of alternative uses of the resources. From this standpoint, the higher burden in the short run, in the models currently practiced, results from the immobilization of public credit resources. Secondly, and more relevant, there is the important aspect of the medium and long-term strengthening effects that the establishment of industrial complexes can bring about. If it were possible to estimate and bring to date the values of the incremental tax revenue to be generated by the new plants, it is likely that the simplistic calculations presented in the media lately would take on a different character. Generally, it should be pointed out that, depending on the nature and size of the incentives granted, and, mainly, depending on whether the initiative is successful in enabling the constitution of industrial districts or of productive complexes (automotive sector, for example), the optimum alternative in terms of policy outcome can be obtained: a behavior of the “J curve”26 type for the fiscal impact associated with the programs. In the initial stage – which can be extended for some years – the costs would predominate. From a certain point, either due to the beginning of repayment of debts, or through the emergence of the strengthening effects, a balance would be reached, possibly to a higher expenditure-revenue level than would be possible without the presence of the new industrial complexes and districts. From this point of view, the essential conditions for a positive fiscal result would be: 1) appropriate initial sizing of the benefits vis-à-vis the expected strengthening effects on the revenue. 2) the subordination of the benefits to a wider concept involving the creation of conditions that enable the generation of strengthening effects. Another aspect takes on particular importance in the recent cases that took place in the automotive sector. As already pointed out, there is a huge difference between the state attracting an assembly plant and attracting three assembly plants. The possibilities of bringing into the regional space a whole sectoral productive chain considerably increase the present value of the future revenue gains. In this sense, the government’s success in attracting supplying companies to the state can be decisive in achieving that positive effect. (It is worth remembering a historical case that can be very useful in assessing the most recent experiences: the establishment of FIAT in Minas Gerais, in the seventies. The successful constitution of the entire park of suppliers turned the assembly plant into an extremely dynamic factor as regards tax collection. The municipality of Betim ranks first in tax collection in the state, even ahead of the capital). On the other hand, in the current situation, the existence of the MERCOSUR increases the risk of the assembly plants established in the south region maintaining a high level of importation from neighboring countries, notably Argentina, significantly reducing this type of impact that, in the case of Minas, was repressed by the transport costs. Let us now examine another issue, referring to the impact the fiscal war has on the states that do not practice it. In the recent Brazilian case, this is the nearly exclusive situation of the state of São Paulo, even though it is true that, if we take a specific sector (textiles, clothes, meat) other states of the south-southeast have been affected by the attraction policies implemented by the northeastern states. 26 The notion of “J curve” is used in foreign trade and currency exchange analyses to indicate the fact that an exchange depreciation can, under certain conditions, have an immediate unfavorable impact on the trade balance, and the favorable effect takes some time to occur. The answer to the question is necessarily positive, since the basic outcome of the process is, as made clear from the whole discussion above, the interregional reallocation of the productive activity. In the event of complete absence of incentives, the trends of industrial allocation, and consequently of the distribution of ICMS collection, should be closer to the previous profile of concentrated distribution, given the underlying trends of private allocation decision as regards any processes of deconcentration or reconcentration, as discussed above. With the incentives, and provided they are successful – in the strict sense that the allocation decisions are not reversed at some point in the future – the states that benefited in that situation began to actually lose revenue and, therefore, will suffer some impact on their fiscal situation. The issue can be illustrated by the case of São Paulo. Given the size of its markets, the production of the companies that leave the state to settle in neighboring states tends to be largely imported by São Paulo. Leaving aside, in this case, the loss of jobs and other internal impacts resulting from the company’s moving, the state ceases to collect 18% of ICMS that it was entitled to when the company was based in the state, and collects only 6%, after deduction of the 12% charged by the state of origin in the interstate transaction. If the ICMS is the object of incentive in the state of origin and – let us assume – is not collected, the company brings to São Paulo what could be called a spurious credit, that the state is forced to acknowledge in the graphic account of the importing company. The tax is not collected, but São Paulo is forced to acknowledge it as legitimate tax expenditure of the company27. 27 A comment on the side. It is often heard, from the technical experts of the government of São Paulo, that the state “loses” the 12% referring to the credit. This has already led, in the past and again in the end of 1999, the government of São Paulo to try to annul the subsidized production credits in other states. The source of the argument is understandable, but it is incorrect. The state loses 12% of collection when the company moves its headquarters to another state. After this, under the current rules, São Paulo is no longer entitled to this tax, therefore there is no loss of credit. Even though it is hard to accept being forced to abate “false” credits, it is the state of origin that is “losing legitimate revenue”. Likewise, if a state decides to grant incentives to companies that have always been located in its territory, São Paulo does not lose revenue. This results from the current rule for the tax, the same that guarantees to the state of São Paulo a profit of about 20% in the collection for being a net exporting state. The change in the rule would eliminate the “loss” of these credits, but the net result for São Paulo would not be favorable. It also seems certain that the attempt to annul the subsidized production credit in the origin would mean that São Paulo would be collecting the tax waived by the state of origin. The only point where the argument has some validity is in the cases where the company uses the tax reduction in the state of origin to re-pass to the prices and to practice dumping in the São Paulo market, as has occurred some times. In short, the aggregate result of this process is, first, an overall loss of collection, affecting all the states for the duration of the incentives. Second, an alteration in the tax revenue appropriation profile, as a result of the changes in the locational profile of the productive activity. It is obvious that this whole discussion is disregarding other changes that might occur in the system, such as, mainly, the change to the destination principle. To conclude, some clarification should be provided to prevent misinterpretation of what is argued above. We attempted to show that, in contrast to what is usually postulated, the fiscal war does not result in an inexorable and immediate deterioration of the fiscal situation of every SNG that practices it. The importance of this conclusion cannot be underrated: it indicates that the fiscal war does not tend to “die out on its own”, and therefore requires some political action to control it. However, this is not meant to minimize the problem or, in any way, to suggest the legitimacy of current practices. First and foremost, because the overall cost for the whole of society is absurdly high. Second, because the absence of any public control on a process that is ultimately dominated by the private sector, has led to unreasonable expansion of benefits, way beyond any rational criterion. Item 4 – RESOLUTION ALTERNATIVES FOR THE FISCAL WAR We understand that there are two possible approaches to address the perverse effect generated by the fiscal war. We will present them in a polarized way, in order to make it easier to understand. In fact, they are not mutually exclusive, and can be combined. The first, which we will call, for want of a better word, impositive, is based on the assumption that these initiatives are condemnable and undesirable as a whole, and attempts to neutralize their manifestations completely. The hesitant attempts by the federal government since 1995 to restrain the process, can, without excess, be classified in this category. The second, which we will call ordering, admits that there is room for the positive action of regional policy, and essentially aims at preventing them from being part of a context where the benefits generated by them are annulled by excessive fiscal costs. First approach: guidelines for elimination of state development policies The first approach would admit, in our view, three basic alternatives for the elimination of the fiscal war practices, which have been discussed in the recent period. 1. To establish a comprehensive legal restriction, if possible at constitutional level, that vetoes the main mechanisms employed, or at least those seen as most harmful; 2. To expand the institutional arrangement of the CONFAZ in order to turn it into an effective forum of inspection and control; 3. To seek restriction through direct tax mechanisms, that is, the adoption of the destination principle. The alternatives can be assessed, in our view, through the use of three criteria. 1. Efficacy – in the sense of capacity to neutralize all the main instruments employed. Its importance derives from the fact that the fiscal war today relies on multiple instruments, not only tax-based ones. 2. Efficiency – in order to involve speedier legal and institutional procedures, capable of responding quickly to the demands. 3. Political viability – with the obviously purpose of maximizing the possibilities of achieving a minimum consensus in the federate political sphere regarding its implementation. Before going into the analysis, attention should be drawn to a recurring point in this paper. The fiscal war mechanisms in the recent period have gone way beyond the limits of simple provision of tax subsidy. While in the past it was basically a tax war, today it is clearly a fiscal-budgetary war. This poses more general problems for restrictive initiatives, with complex implications regarding the delimitation of SNG autonomies. It is one thing, based on the appropriate conception of the ICMS as tax collection, to reject its manipulation for the purpose of discretional policy. It is quite a different thing to question the power of the state governments to allocate their resources freely, through their budgets, to whatever end they wish. This would very likely be seen as a return to the budgetary centralism of the period of the authoritarian regime. The first alternative listed above would choose to introduce general restrictions, preferably in the Constitution. An example is the apparent attempt of the Federal Government to introduce a strong restriction in article 155 of the tax reform bill, of August 1995. The basic merit of this alternative is efficacy: it would allow, in principle, restraining a wider share of the instruments used today, mainly the “financial triangulation” that brings the ICMS deferral into play. Its main problem is that it would send all the disputes to the common justice, with the inevitable implications in terms of long legal proceedings and loss of agility, making it less suitable from the point of view of the efficiency criterion. From the latter point of view, any alternative based on a more restricted forum and that combines enforcement capacity and room for negotiations among federate agents can be more efficient. Even the main merit of this alternative – its scope – would be limited by the existing difficulties, as mentioned above, in restricting the typical budgetary and credit actions without generating conflicts concerning the autonomy of the SNGs. It should be pointed out that it would have little political viability. The second alternative would rely on the use of a more restricted forum, the CONFAZ. The hypothesis that apparently guided the most recent attempt to inhibit the fiscal war, included in the Kandir Bill, was that its improvement, with the creation of an Ethics Commission, would enable it to regulate the excesses. This alternative is limited in scope. It can be effective only at the strict level of provision of tax subsidies, but has little chance of inhibiting other instruments. Concerning efficiency, it would have the merit of belonging to a more restricted forum, which would in principle speed up procedures. However, given the current format of the CONFAZ and given its legal base, the Council would only be able to point at the irregularity and act indirectly, through the Attorney General of the Republic or the Court of Accounts28. 28 The most recent proposal to send to the Federal Senate the judgment of contracts made between the private sector and the state governments has some points in common with this discussion. The Federal Senate would be elected as a special forum to take the necessary measures to inhibit the fiscal dispute. Similarly to the case of the CONFAZ, this proposal wins in terms of political viability, but loses mainly in effectiveness and efficiency. The third alternative would be to explore the possibility of altering the ICMS regime, implementing the destination principle. This could be done directly, by changing the rules for establishing the interstate rates29 , or indirectly, through the adoption of the shared ICMS with the federal rate being used as a transfer mechanism in the interstate transactions, which entails significant additional advantages30. This alternative is, on the one hand, highly efficient in interrupting the flows of deferral presently in course, but should not be viewed as a panacea, as seems to be the opinion of some observers. First, it affects only the tax benefit, and does not reach the fiscal-budgetary forms of subsidy provision; therefore, its effectiveness is restricted. Moreover, there is a vast supply of deferral agreements, both explicit and disguised, that would be neutralized by it. This, in itself, would establish an additional factor of rejection to the measure, in addition to other existing factors31. From the more systemic point of view, it should be noted that this measure is as capable of neutralizing current forms of tax subsidy provision as it is of reversing the current relations of relative power in the use of the instrument. The destination principle makes it impracticable to provide the benefit to “internal platforms of exportation” – states with their own small markets and close to large centers – and increases the subsidizing capacity of the states that have the largest markets. For many sectors, viable production scales will be optimal and can be close to the share of local market controlled by the company. This would reintroduce, only for the large states, the possibility of negotiating ICMS collection in exchange for investments, even with the adoption of the destination principle. From this perspective, therefore, the destination principle represents not only an advantage to the smaller non-producing states, since it increases their share in collection, but also a loss of relative potentiality in the use of tax subsidy provision. As to the criterion of political viability, some considerations are required. From this point of view, only the second and third alternative would apparently be feasible. The second, for making use of an already existing agency and proposing a surveillance process restricted to the states themselves, without the mediation of other powers. The third, because it involves a significant revenue increase for the majority of the states32. If we consider that today the great majority of the Brazilian states use policies to attract investment, it is unlikely that they will accept the pure and simple elimination of this possibility without any counterpart, such as in the second alternative. We are left with the third, which has obvious advantages. It remains to be seen if the states, mainly the small ones, whose tax collection apparatuses are inefficient and have no advantage in the fiscal war other than the manipulation of the ICMS, will accept without resistance the substitution of the instrument for an additional potential revenue (which depends on their own tax collection effort). 29 For a comprehensive discussion of the problems involved, see RESENDE, F. (1993). 30 The operation of the shared tax is described didactically in VARSANO, R. (1995). 31 The recent case (end of 1999) that occurred in the discussion of the tax reform is illustrative. In face of the proposed implementation of the destination principle, the strong opposing reaction was only bypassed by the presentation, headed by the Liberal Front Party of Bahia, of a proposal by which the federal government would take on the costs of the projects already approved. Second approach: the limited autonomy of the SNGs in the regional development policies The second approach is based on the assumption that it is not possible, perhaps not even necessary and probably not desirable to summarily cancel the initiatives that, in their extreme version, constitute the fiscal war. It is not possible, within the limits of the current federate arrangement, for the basic reasons that we have already pointed out above. The central government’s incapacity to operate discretional policies does not lead to their disappearance. On the contrary, it induces their creation at other levels of government, as we have seen, presenting specific and new deficiencies and biases. The economic and political interests that previously attempted to gain space in the national policies, now focus on creating their spaces at the regional level. Moreover, at no time in the past, not even during the authoritarian period, were these practices eliminated. It might not be necessary, because the problem does not derive from development policies per se, but rather from their uncontrolled and perversely competitive format. In essence, the problem is that, in the absence of a state coordination alternative, the process is handed to the control of the private sector, which seeks to maximize everything that is possible, but certainly not public benefit in the use of fiscal resources. Finally, it might not be desirable due to the various advantages that ordered federate autonomy offers, at a time when the reach of policies that are exclusively supported by the central state seems to be limited (see below). 32 It should be noted that, in the event of strict enforcement of the application of the destination principle, the increase of collection of the smaller states would depend on the improvement of their tax collection apparatus, which would be hindered by the inevitable increase of the number of companies to be inspected. This would only be prevented by collection based on the origin and later distribution. For further details, see RESENDE (1992). All this leads to another approach that seems more coherent, which is to combine a minimum level of restrictive normatization with the reconstitution of an institutional space for organizing policies at the national level. We will restrict ourselves here to indicating basic guidelines, as an initial contribution to a debate that has been notably lacking in alternatives. 1. Principle of transparency. Any alternative depends on the primary and essential condition that political and legal arrangements be put in place to guarantee the impossibility of private agreements being made between governments and companies. It is the only way for the incentives to be both the object of scrutiny by society and the object of evaluation and comparison with any rules and criteria that are developed. There are no easy solutions for this, since the desirable democratic alternative, namely, the scrutiny of the state governments by the local societies, does not necessarily take place. It is possible, however, that even the local societies, albeit strongly committed with the investment attraction policies of their governments, can be drawn to the thesis of transparency, and start supporting some mechanism to enforce it. Various measures can be implemented, such as the development of tax waiver budgets at the level of the states, which already occurs for the federal budget, or the requirement of public reports at certain periods, such as was done in the case of the federal privatization program. From the point of view of the SNGs, the acceptance of the transparency requirements can be compensated by the continuation of the operation of regional policies in a legitimate manner. 2. Coordinating the federal government’s action to achieve the formulation of criteria to define the limits within which the granting of benefits can be practiced. Essentially, this would require, that the problem of industrial deconcentration be taken up again; regions and sub-regions (and not necessarily state geopolitical spaces) would have to be ranked according to the relevance of the industrial density;; regions and sub-regions would have to be correlated with preferential industrial sectors, according to natural economic advantages, infrastructure availability, etc. This seems to be the direction of the proposal made by the governor of the state of Rio Grande do Sul, based on the definition of allowed “bands or brackets” of subsidy provision, according to each state’s degree of economic development. 3. Development of mechanisms and institutions for the monitoring, evaluation and scrutiny of the benefits. From the institutional point of view, the most promising alternative seems to be a state forum, provided it is free of the inhibiting rule of unanimity, and under the coordination of the federal government. The CONFAZ, with its current structure, would not be the most suitable agent for this. Another institution would be required, with a new type of interaction with the federal government. This institution would be formed by a technically equipped and strengthened executive secretariat and majority-based decision criteria. Final remarks Two final remarks are required on the position advocated here, which is based on the ordering approach described above. First, this option does not exclude the use of the restrictive alternatives presented in the beginning. In some cases, those measures can be implemented regardless of the aim of inhibiting the fiscal war, as is the case of the adoption of the destination principle. The essential point, in our view, is that the possibilities of success are limited in any process of control that does not encourage, in some way, the cooperation and the compliance of a significant share of state governments. Recent developments have demonstrated that, presently, in the absence of a better arrangement and in face of the possibility of losing degrees of freedom in this type of policy, the state governments will react radically to any attempt at control, as was demonstrated in the fragile attempt to rewrite Complementary Law n. 24/75 and the creation of an innocuous Ethics Commission in the CONFAZ. Secondly – and this is a point of absolute relevance in the understanding of the whole process – this option takes into consideration some basic aspects of the transformations currently in course, both in the domestic and the international spheres. Starting with the domestic evolution, the process of dismantling the strong interventionist apparatus that characterized the central power in Brazil in the last decades is clearly defined and probably irreversible. The federal government has led all the privatization and deregulation initiatives on several fronts, drastically reducing the practices of discretional intervention that had been the rule in the past. In particular, as we argued in the first part of this paper, the central government has adopted a clearly remote position regarding active industrial policies. This has meant, among other effects, the dismantling of a broad and complex institutional apparatus that articulated and organized the interests of regional groups, economic sectors, national and foreign economic agents, through the action of the ministries, public funds, large state-owned companies and government agencies. In parallel, the process of institutional and political reforms that culminated in the end of the authoritarian regime resulted in an effective renewal of federate autonomy that is typical of the post-1946 period, probably expanded by an even more decentralized appropriation of resources. A spontaneous, unplanned movement is taking place, in which the sub-national governments increasingly take on burdens that previously belonged to the federal government. From another perspective, it is obvious that the mechanisms of articulation between state governments and the private sector are, today, radically different from those that characterized the previous stage, commanded by the federal government. Instead of an investor/producer state, whose agents were public technocracies, based in the state-owned companies and large government agencies, dragging along a complementary private sector or encouraging the emergence of economic sectors dependent on their spending power, we now see state governments that seek to operate in a complementary manner and, essentially, dependant on private investment. If, as we point out above, it seems that state governments have replaced the federal government in the role of implementing regional policies, no doubt the relationship model has changed deeply. In the international sphere, there is increasing evidence that, in the context of what is known as “globalization”, one of the main developments is the relative weakening of the national states, which are increasingly incapable of operating macroeconomic policies with minimum autonomy, and most of which are tied to highly restrictive situations of fiscal deterioration33. At the same time, international researchers increasingly focus on strictly regional economic, political and social issues. Many believe that the weakening of the national states would place on the agenda the importance of other sources of governability (Hirst & Thompson (1986); Rosenau.& Czempiel (1992)). These aspects seem to suggest the possible emergence of what could be a new pattern of relationship between governments and society in the globalized economy, in which the dynamic agents of a new model of state intervention would probably be the regional political powers. It is clear that the problems and specificities of an economy such as the Brazilian one impose major obstacles to a process of this scope. It is also true that much can be said against an external movement of autonomization of regional powers. In fact, this study addressed, before anything else, this fact. And, even more obvious, this is a subject that requires additional research, which exceeds the limits of this paper. After considering all this, there is still evidence that the Brazilian state has shifted from an extreme position of centralized decision-making, which coincided historically with the authoritarian regime, to an process of growing political and fiscal decentralization, advancing much more quickly than the necessary construction of an institutional arrangement capable of regulating it. The fiscal war emerged, clearly, as an expression of this gap. The crucial point is perhaps that it is not possible, or even desirable, to reconstitute the old mechanisms of regulation. Therefore, the challenge is to adjust the organization of the state to this new reality. 33 At the extreme, leading to positions of the type represented by authors such as OHMAE (1990), who defend the end of economic borders and the end of the so-called national interest. BIBLIOGRAPHY AFFONSO, Rui de B.A.(1995a) “A Crise da Federação no Brasil ”. FEE Essays, year 15, no 2, pp.321-337. ____________________1995(b) “Os Estados e a Descentralização no Brasil.” São Paulo. mimeo, 54p. AFONSO, José Roberto R. (1989) “Evolução das Relações Intergovernamentais no Brasil entre 1968 e 1988: transferências e endividamento”. Master’s Dissertation Inst. Ec. Industrial, UFRJ, RJ, September. _____________________(1994) “Descentralização: um estudo de caso sobre o Brasil”. ( Textos Discussão CEPP, no. 20), January. _____________________(1995) “A Questão Tributária e o Financiamento dos Diferentes Níveis de Governo”. Paper presented at the International Seminar on Impasses and Perspective of the Federation in Brazil. São Paulo, FUNDAP/IESP, 21p, maio. CANO, Wilson (1981) “Raízes da Concentração Industrial em São Paulo”. SP, T.A Queiroz. _____________(1991) “Concentracion, Desconcentracion, Descentralizacion en Brasil.” Campinas, mimeo, 22p. ______________(1995) Auge e Inflexão da desconcentração econômica regional no Brasil. ANPEC Annals – ENCONTRO NACIONAL DE ECONOMIA, Salvador. DINIZ, Clélio C. (1991) “Desenvolvimento poligonal no Brasil – nem desconcentração nem contínua polarização”, 53p, mimeo. ________________(1995) A dinâmica regional recente da economia brasileira e suas perspectivas. Discussion Texts IPEA, no. 375, June. DINIZ, C.C.e CROCCO, M. A . (1996) Reestruturação Econômica e impacto regional: o novo mapa da indústria brasileira. Nova Economia, Belo Horizonte, v. 6 no. 1 pa.77-103, July. HIRST, P. & THOMPSON, G. (1986) Globalization in question. Cambridge, U.K.: Polity Press. LONGO, Carlos A. (1979) “Uma contribuição para a reforma do ICM: o caso dos ajustamentos de impostos na fronteira.” Research and Economic Planning, Rio de Janeiro, v.9, no 3, pp.803-818, December. _________________(1980a) “Replica: o caso dos ajustamentos de impostos na fronteira: uma contribuição para a reforma do ICM.” Research and Economic Planning, Rio de Janeiro, v.10, no 1, pp.327-332, April. _________________(1980b) “Os ajustamentos de fronteira do ICM, o comércio interestadual e internacional e a autonomia fiscal dos Estados – comentário.” Research and Economic Planning, v.10, no 2, pp.659-660, August. __________________(1981) “Controvérsias sobre o ICM no comércio interestadual: uma resenha.” Research and Economic Planning, Rio de Janeiro, v.11, no 1, pp.229-250, April. OHMAE, K. (1990) The Bordless Word. London: Harper Collins. PRADO, S. (1995) Intervenção Estatal, Privatização e Fiscalidade: um estudo sobre a constituição e crise do Setor Produtivo Estatal no Brasil e os processos de privatização a nível mundial. Doctorate Thesis, Economics Institute of the State University of Campinas, December, 302 pages. PRADO, S /CAVALCANTI, C.E.(2000). Aspectos da Guerra Fiscal no Brasil. FUNDAP/IPEA Edition, São Paulo. RESENDE, Fernando (1996a) “O processo de reforma tributária. Brasília,” Instituto de Pesquisa Econômica Aplicada – IPEA, 18p, January. (Text for discussion no. 396). __________________ (1996b) “Propostas de reforma do sistema tributário nacional.” Brasília, Instituto de Pesquisa Econômica Aplicada – IPEA, 26p, May. (Text for discussion no. 420). ROSENAU, J.N.& CZEMPIEL, E. (eds.) (1992) Governance without government: order and change in world politics. Cambridge: Cambridge Un. Press. VARSANO, Ricardo (1979) “Os ajustamentos de fronteira do ICM, o comércio interestadual e internacional e a autonomia fiscal dos Estados.” Research and Economic Planning, Rio de Janeiro, v.9 no 2, pp.315-350, August. _________________(1980a) “Comentário: o caso dos ajustamentos de impostos na fronteira: uma contribuição para a reforma do ICM?” Research and Economic Planning, Rio de Janeiro, v.10, no 1, pp.319-326, April. _________________(1980b) “Os ajustamentos de fronteira do ICM, o comércio interestadual e internacional e a autonomia fiscal dos Estados – réplica.” Research and Economic Planning, Rio de Janeiro, v.10 no 2, pp.661-662, August _________________(1995) “A tributação do comércio interestadual: ICMS atual versus ICMS partilhado.” Brasília, Instituto de Pesquisa Econômica Aplicada – IPEA, 14p, September (Text for discussion n. 382). __________________(1996) “A guerra fiscal do ICMS: quem ganha e quem perde.” Text presented at the International Seminar on Decentralized Fiscal Policies , Brasilia, DF, 16p, November.