Nicholas Aroney Fiscal Competition (Work Sessions 11 and 23) 1. Introduction The work sessions on the subject of fiscal competition focused on questions of tax competition versus tax coordination and the re-distributive incidence of fiscal competition. Such questions suggest a problem of conflicting values: between competition and efficiency on one hand, and coordination, redistribution and equity on the other. Can the efficiency benefits of fiscal competition be maintained, without leading to sub-optimal re-distributive outcomes? The economic rationale for federalism assumes the existence of territorially defined local public goods, and claims that federal structures enable responsive variation, experimentation, competition and efficiency in the provision of such goods. However, some experts believe that fiscal competition can lead to “skewed” outcomes in revenue, regulation and the provision of public goods between different regions. Competition may encourage sub-national units to reduce taxes in order to attract investment and to reduce welfare to deter immigrants with lower incomes. It is therefore argued that central governments must retain coordination and re-distributive powers to overcome these problems. But in response it is said that centralised solutions foreclose the efficiency gains and local choice entailed in fiscal decentralisation. Centralised solutions may lead to excessive public sector growth – the so-called “Leviathan” state. A tension between opposed policy outlooks is thus recognised, and attempts are made to secure the benefits of fiscal competition while avoiding the problems associated with it. However, the design of such systems is vexed by significant theoretical and practical limitations. Firstly, the empirical literature seems to be inconclusive; it appears that different variables and models produce different outcomes. Secondly, the literature is beset by the normative problem of identifying an optimal revenue and expenditure mix. Thirdly, there is the practical problem of implementing idealistic fiscal reforms in the unique political, economic and social contexts of each country. Economists may agree on the goal of utility maximisation, but formidable empirical, normative and practical obstacles lie in the way of implementing schemes that maximise the benefits of fiscal competition while minimising the detriments. 2. The work sessions The three cases considered – Brazil, India and the United States – offer very different examples of fiscal organisation, each within a relatively large federal system, but with different levels of economic development. In the United States, federal, state and local governments exercise their capacities to levy a wide range of taxes independently, enabling them to engage in fiscal competition. Given the substantial autonomy and fiscal competencies recognised by the federative pact of 1988, Brazilian states also engage in extensive horizontal competition. India, by comparison, has a vastly more centralised fiscal system, which substantially eliminates fiscal competition. However, both Brazil and India adopt re-distributive formulae to allocate federal government funds among the states, whereas there is no formal redistributive mechanism in the United States. 2.1. Tax competition versus tax coordination Fiscal competition in Brazil is largely conducted through reductions in the incidence of state value added tax (VAT) at the point of origin. Critics argue that the ensuing “tax war” diminishes revenue, reduces expenditure and concentrates firms within the more developed states. One case statement maker argued that fiscal competition in Brazil was neither effective nor benign. Infrastructure, rather than tax incentives, is the strongest determinant of business relocation and investment. Paradoxically, tax incentives enable firms to “blackmail” the state for continued incentives. Tax wars therefore encourage distorted and unsustainable investment, while entailing a decline in tax revenues in a competitive “race to the bottom”. As a solution, a state destination tax collected at source, together with a system of inter-governmental transfer, was proposed. Expenditure competition and competition for federal funds could be maintained, the latter being made conditional on performance in the provision of public goods. The success of such reforms would depend, however, on the negotiation of a new federative pact and the development of robust local and regional government, with strengthened civic participation. As was later pointed out, reform would also have to satisfy the interests of the larger and wealthier states, and maintain a politically sustainable level of state revenues. Another case statement maker drew attention to efficiency improvements among the Brazilian states resulting from competitive incentives to best practice in the provision of public goods. However, negative side effects were also observed, such as localities “free riding” on superior health services provided in adjacent localities. It was agreed that tax competition is only a secondary factor in investment decisions, but politicians behave as if competitive policies were a decisive factor. Indeed, firms can respond unpredictably, sometimes undergoing corporate restructuring in order to take advantage of concessions. It was also agreed that fiscal competition entails a degradation of the tax system in general. A youth participant further observed that fiscal competition can be in tension with regional development policy, particularly in countries where there are substantial socio-economic disparities. Government negotiations with businesses are also likely to undermine political transparency. Other participants agreed with and elaborated these points. Can problems such as these be avoided through better institutional design? An analysis statement maker explained the limitations of the various kinds of economic models, but argued that dynamic, repeatable “game theory” models show that fiscal coordination is possible. Agents may free ride on services provided by others, but agreements to coordinate fiscal policy can be drafted to prevent such problems. Another analyst observed that it is difficult to reform origin-based service taxes to make them destination-based; it would be better to avoid service taxes in general. In the case of state VAT, a destination tax could eliminate the incentive to fiscal competition, but enforcement would be inefficient. It would thus be better to levy a destination-based tax at source. It was observed that vertical fiscal competition is also a problem in Brazil: the federal government now relies on sub-optimal turnover and labour taxes. The third case statement maker pointed out that the centralised fiscal system of India substantially eliminates harmful competition. The rationale for federalism in India is cultural and social, rather than economic, and the federation is highly centralised in order to counterbalance centripetal social forces. This includes fiscal centralisation in order to address the economic weaknesses of many of the Indian states. The states assert their autonomy for cultural and political reasons, but remain financially dependent upon the centre. Inter-governmental fiscal relations are generally prescribed by the Indian constitution. Some central taxes are automatically assigned to the states, others are collected and appropriated by the states, while others are distributed by the centre to the states. The Finance Commission makes general distributions, the Planning Commission allocates funds for special development projects, and the various central ministries may make additional transfers. However, the funds available to the Finance Commission are dependent on decisions of the central government, and the states complain that the scheme is overly centralised. Yet ironically, the states are politically reluctant to exercise the taxing powers they do have. They are also reluctant to transfer funds to localities, and it is argued that the centre should fund localities directly. The small amount of fiscal competition that does occur in India is largely ineffective because the main determinant of investment decisions is the availability and quality of infrastructure. Another factor is that the financial health of the centre and the states is poor. Moreover, there is a lack of fiscalpolicy coordination. Governmental bodies often plan for projects for which the necessary funds do not materialise. Fiscal reform is therefore on the political agenda. One analysis statement maker argued, however, that the elimination of state taxing powers may be a remedy that is worse than the disease. Fiscal competition is probably inevitable in fiscally decentralised systems, and the goal is to structure the system to secure the benefits of competition while avoiding the problems. In this respect, tax competition must be assessed in the wider context of competition in the provision of infrastructure, services and regulatory policies. Despite its problems, Brazil has adopted a clear allocation of taxes to each level of government, which makes the system more transparent and encourages accountability. Adopting a destination principle for state taxes may help to alleviate the problems that do exist. It was also noted that there is little empirical evidence for a race to the bottom. For example, there is no evidence that the devolution of environmental policy in the United States has resulted in reduced environmental protection. Models that measure “distortions” as a result of fiscal competition usually report only minor effects, in the order of 3% of revenues. And an assessment of fiscal competition must take account of the “Leviathan” tendency of governments. Lower revenues may reflect efficiency gains, not merely welfare losses. It was observed that capital mobility in the United States entails a high level of investment competition. Poorer regions make bids to induce investment, with a kind of “balancing” tendency. Problems of harmonisation within the European Union (EU) were also discussed. Destination-based taxes in the EU were compared to origin-based taxes in the former Soviet Union. In the latter case, the structure of products may be significant. The production of goods with high elasticity of demand can depend on products with low elasticity, reducing the effectiveness of fiscally competitive policies. It was also noted that the allocation of taxes to each level of government may not fully resolve the issue. Firstly, the characterisation of particular taxes can be disputed and judicial determination may be necessary. Secondly, executive discretion in enforcement can be used as a mechanism of tax competition notwithstanding the legal position. Recent fiscal responsibility legislation in Brazil and Argentina was also discussed. The effectiveness of the Brazilian scheme, it emerged, may have been enhanced by lower interest rates and the restructuring of state debts. In Argentina, fiscal reforms had only been implemented effectively at a national level. In this respect, it was considered that the best sanction to encourage fiscal responsibility is for the central government to resist pressure to bail out improvident regions. However, national economic credibility may also be at stake and a federal government may be politically or constitutionally obliged to intervene in major regional fiscal crises. Finally, it was noted that the dialogue had focused on the “problems” of fiscal competition in Brazil. It would have been better, it was argued, to identify the general principles of beneficial competition, and to work out specific policy proposals in that light. 2.2. Fiscal competition: re-distributive incidence In the United States, the federal and state governments levy a wide range of taxes. This allows a wide variety of revenue policies and extensive fiscal competition. Some federal conditional grants have equalising characteristics, but there is no systematic attempt to redistribute finance among the states. One case statement maker argued that fiscal decentralisation in the United States significantly undermines incentives for re-distributive policy among local and state governments. In California, state taxes on property, income and retail sales leave little scope for horizontal competition. However, there is a high degree of vertical competition as well as a kind of lateral competition within governments among various projects and services. Much of this competition is the result of the democratisation of fiscal policy. Citizens have repeatedly voted for constitutional guarantees of fiscal responsibility, weakening the capacity of governments to provide welfare services. Democratic approval of new taxes has been limited to popular programs, such as increased spending on roads and police, but not health. It was concluded that central governments must not abandon their responsibilities to engage in re-distributive policy. Another case statement maker pointed out that within countries belonging to the Organisation for Economic Cooperation and Development (OECD) there are significant levels of fiscal competition. For example, most Swiss cantons have significant tax competition schemes. Low tax rates seem to be more effective than special incentives to particular firms. There is a correlation between low tax regions and high personal income, but it is not clear which causes which, and the effect of low tax policies can be diminished by high property prices. The experiences of Switzerland and Canada, it was said, also suggest that decentralisation does not necessarily entail a break down in progressive taxation and welfare expenditure, although this is complicated by disparities between rural and urban areas. Rural areas require less infrastructure and can more easily engage in fiscal competition through lower tax rates, whereas urbanisation can entail concentrations of poverty and high welfare costs. To counteract “harmful” fiscal competition, it was proposed that central governments retain re-distributive functions, horizontal cooperation be encouraged and, more radically, poor and rich localities be merged. Whether harmonisation can be achieved will depend, however, on the size, territorial organisation and constitutional framework of each country. If sub-national units are prevented from imposing externalities on other units, fiscal competition can be efficient, but principles of tax neutrality and transparency are necessary to avoid discriminatory policies. The third case statement maker noted that the four significant categories of government in the United States – federal, state, local and tribal – are each expected to be self-sustaining. The federal government in fact raises 100% of its revenue, the states raise around 75% (with 25% in federal grants), local governments receive small grants from the federal government, and tribal administrations receive federal monies and raise additional revenues through various enterprises. Some federal expenditures, such as social security and regional tax breaks, have a re-distributive effect, and particular federal grants provide funds for social services. However, in the State of Washington for example, fiscal policy has been deeply influenced by a strong “anti-tax” mood, entailing under-funding of various public services. By contrast, it was noted that in Germany changes in social policy are translated into changes in revenue and resource allocation. It was also observed that in Canada there are serious fiscal imbalances, with long-term trends to deficit budgets in the provinces, and problems for vital public services such healthcare for the elderly. Federal intervention may resolve such inequities, but with negative consequences for provincial “sovereignty”. The discussion then turned to vertical fiscal competition in the United States, noting that the states usually resist un-funded and partially funded federal mandates. Inter-state fiscal competition in South Dakota was also discussed. The state is seen as a “low tax state”, with no income tax, and must therefore rely on regressive taxes, such as sales taxes. However, even those adversely affected appear to prefer this. Recent popular rejection of an estate tax also had a competitive effect (encouraging retiree immigration), but voters were not apparently motivated by the competitive aspects of the policy. It was argued that fiscal competition can be inefficient, with high costs in foregone revenue and low returns in terms of job growth. Also, the firms liable to be induced by tax benefits are likely to be financially “weak” in any case. However, competition has not produced an observable race to the bottom, and the states have engaged in voluntary tax harmonisation, such as attempts to prevent sales tax avoidance through Internet sales. It was therefore concluded that fiscal competition is mostly beneficial. States with high unemployment and poverty seek to attract investment, with a long-term tendency to “balance out” inequalities. It was noted that in the EU the “destination principle” resolves uncertainties about catalogue and Internet sales. It was also noted that in India, fiscal and legislative centralism prevents any comparable fiscal conflict from occurring. At first glance, the United States is a curious choice of case, given the absence of equalisation measures at an inter-state level, although there is equalisation at an intra-state level. However, generalisation is difficult – there are remarkable differences among the various states, towns and counties. The system appears chaotic, but functions reasonably well, despite the problems. It was agreed that block grants had altered the incentive structure of state revenue arrangements, and one might expect a race to the bottom. However, the evidence is not forthcoming. Concern with welfare dependency had led to fiscal decentralisation and the imposition of outcome conditions on federal grants to encourage innovation, experimentation and employment growth. Federal conditions have included a requirement that historical levels of state welfare spending be maintained. The scheme appears to have been very effective, with many people finding work. It was concluded that fiscal competition certainly has an impact. Infrastructure is the most important factor, but there remains a correlation between low taxes and increased investment. However, calling fiscal competition a “fiscal war” places an unfortunate spin on the evidence. Fiscal competition can increase efficiency and need not entail a race to the bottom. Economists generally applaud the efficiency of market competition and it is somewhat incongruent to then critique competition in the provision of public goods. In the work sessions, it was generally agreed that fiscal competition can entail a competition for funds and public goods, and that taxpayers may be prepared to raise taxes for enhanced service provision. Governments have to establish that they are being efficient, and fiscal competition can have a positive impact. Some participants, however, observed that competition could still have a negative impact on poor regions, especially in less developed countries – although it was countered that corruption and weak infrastructure were often the significant factors. One analysis statement maker observed that the Brazilian states are legally required to obtain the agreement of all the states before implementing tax incentives, but that this law has not been enforced. The implication was that economic incentives are a better means to regulate fiscal relationships. It was then asked whether the absence of any race to the bottom in the United States could be explained by the use of “eligibility” requirements (e.g. residency periods) for welfare entitlements. It was observed, however, that such requirements had been struck down as unconstitutional, except for schemes where migrants are initially entitled only to a level of benefits equal to those available in their state of origin. It was then asked whether fiscal competition encourages spending on infrastructure rather than welfare? It was said that despite the devolution of welfare responsibility, welfare payments are not a large proportion of state budgets. Health expenditure is more significant, and here low government reimbursements have encouraged the avoidance of the provision of medical care for the poor. However, significant variations among the states were also noted. It was observed that healthcare has both a re-distributive and a local public good dimension. Efficiency and accountability may be strong at a local level, but local revenues are relatively weaker. It was also observed that historically, the Catholic Church had provided welfare throughout Europe, but that Protestants often made welfare a civic responsibility, so that communes tended to make citizenship a criterion of entitlement. As welfare responsibilities were shifted to cantonal and national levels, “club membership” continued to be a criterion. Recent fiscal decentralisation means a return to localised responsibility, with renewed affirmations about limited entitlement and payment for services. At this point, one participant pointed out that the discussion had lost sight of the human dimension: the impact of financial policy on actual living standards. Attention was also drawn to the background issue of trade-offs between efficiency and equity involved in arguments between “club” goods and public goods. Another observation was that too much political attention is often given to the role of large firms, when small businesses constitute the largest sector in many Western economies. However, governments are reluctant to negotiate with small firms, and larger firms have more political influence. It was also observed that a range of intangible factors, unrelated to government policy or fiscal design, were responsible for the success of many emergent commercial regions, such as Silicon Valley and the Seattle region. The general consensus was that fundamentals such as infrastructure and social capital are the key determinants of economic development. Why, then, do political leaders continue to engage in tax competition? It was noted that politicians tend to make short-term decisions, whereas fiscal design and redistributive incidence are long-term issues. But then how do we account for the divergent directions taken in the United States and the EU, as well as significant variations between similarly organised countries, such as Switzerland and Germany? An analysis statement maker concluded that genuine fiscal decentralisation is bound to produce fiscal competition, whereas harmonisation forecloses the benefits of diversity. The basic problem, therefore, is how to prevent the most destructive forms of tax competition, while retaining its benefits. For example, two governmental entities in Bosnia and Herzegovina adopted different retail and import taxes, and businesses restructured their processes to avoid paying tax in both jurisdictions. One response to this was that such situations justify reform, but not necessarily tax harmonisation. Another was that tax uniformity, as in India, completely avoids problems such as these. It was pointed out, moreover, that different social, cultural and political backgrounds of countries such as India and the United States correlated with very different attitudes to fiscal competition, harmonisation and re-distributive incidence. 3. Conclusions The choice of cases was a significant determinant of the discussion. Focusing attention on transitional economies in the first work session oriented the dialogue towards problems of fiscal competition, rather than problems of fiscal centralisation. By comparison, the focus in the second work session on a highly developed national economy brought the case for fiscal competition more clearly to the fore. In developed nations, fiscal competition is often viewed in comparatively benign terms. In such countries, infrastructure levels are relatively uniform, so that competition can be effective. Also, poverty and regional economic disparities are less serious, so that any race to the bottom is less ominous in its implications. Thus even those who drew attention to the regressive re-distributive incidence of fiscal competition in the United States tended to see competition as a natural and mostly healthy aspect of economic and political life. It was widely agreed that factors other than tax competition are the key determinants of firm location and investment decisions. Infrastructure, social capital and other intangibles appear to be the most important economic success factors. The short-term orientation of politicians who behave as if tax incentives were a decisive factor was frequently noted. However, proponents of fiscal competition did not rest their case on a “competition” for business, but on the opportunities for variation and experimentation in the provision of public goods, and the incentives to efficiency and effectiveness that this entails. The underlying economic and political values of participants from different countries were an important factor in the discussion. Here, many of the usual dichotomies – such as those between equality and freedom, solidarity and diversity, and between absolute and relative measures of utility or welfare – seemed to lie beneath the surface. It also appeared easier for participants to agree on abstract generalisations concerning trade-offs between fiscal competition and coordination, just as it was easy to agree on the specific technicalities of particular tax regimes. However, it was more difficult to find common ground on “middle-level” questions concerning the evaluation and reform of economic policies adopted by particular countries. As noted at the outset, the inconclusiveness of the empirical evidence, and differences in normative perspective are compounded by practical obstacles imposed by the historical, social and political circumstances of each country under consideration. One general conclusion was that the design and operation of federal economic systems is heavily dependent on the unique political, social and historical factors of each country. For example, in India cultural diversity is countered by a high degree of fiscal and political centralisation, whereas the terms of Brazil’s federative pact make any concession of state fiscal powers difficult to procure. Fiscal arrangements in both countries are thus shaped by both economic considerations and historical circumstances. It was concluded that economic reforms can be effectively implemented only if they are sensitive to these factors. So, how far did the work sessions advance beyond the existing literature? Perhaps the real benefit of the dialogue was that it brought together a range of participants who contributed a diversity of national and professional experiences, analysed from an array of scholarly disciplines and a variety of political and economic perspectives. Perceptive commentators might have anticipated the general results of the discussion, but the forum brought home the complexity of the issues and the variety of frameworks from which fiscal competition can be assessed. As mentioned in the introduction, a pragmatic path may seem to beckon: one which unites meticulous research with careful design and cautious, gradual implementation. However, the orientation to design, it seems, must always be balanced – particularly in federal systems – with principles of autonomy, negotiation and consensus. And thus it may be that federalistic decisionmaking processes (which give regions and localities a distinct voice in national decisions) are an appropriate and effective way to facilitate the negotiated arrangements that can resolve issues of fiscal policy as they emerge.