involve compromise; not everyone can have the outcome they wish. Changes in the structure of govemment will thus typically result in gainers and losers, and this will usually be true of decentralisation. Only where decentralisation manages to separate the population into homogeneous sub-groups is it possible to prove that decentralisation leads to a Pareto-optimal outcome; in all other cases, the case for decentralisation must contend with the fact that the greater diversity in government decisions which will result from decentralisation may not be to everyone’s liking (see Hughes and Smith, 1991, for further discussion).
1. Co-ordination of indirect Taxes
The extent of economies of scale in tax administration is likely to vary between taxes and between the various aspects of tax administration and enforcement, depending on the type of operations involved. Administrative economies of scale arise from the interaction of two opposing influences. On the one hand, centralisation allows more specialisation, and hence, gains from the division of labour. On the other hand the problems of communication and control will generally increase more than proportionately with the size of the organisational unit. (Arrow, 1974). Generally speaking, the larger the information content of policy implementation - for example, about local needs and circumstances - the greater will be the case for decentralisation, often to levels below those of national governments (Helm and Smith, 1987). Although evidence from the US suggests that tax administration is an activity characterised by some economies of scale (Netzer, 1974), it is unlikely that this can be directly transferred to the EC situation, where both language differences and differences in legal systems and corporate structures are likely to reduce quite sharply the potential for administrative savings from EC-wide tax administration.
Considerations of economies of scale would seem to indicate that tax administration should remain at the national level; the gain in going further to the EC level is likely to be small, whilst many of the costs of greater scale arise at this point. Whereas in moving from local to national level the scale of operation may increase by a factor of hundreds or thousands, in moving from national to Community level the scale of operation only increases, on average, by a factor of twelve. It is unlikely that the unit cost savings of an increase in scale of this order will be sufficient to warrant the costs of diversity in languages and legal systems that EC-wide administration would have to confront.
This is not to deny that there may be some considerable potential for beneficial administrative cooperation between the tax authorities of EC member states. Much tax evasion is able to take place precisely because the information and powers of national tax authorities decline sharply beyond national frontiers; information exchanges and other forms of administrative coordination may be an important source of gains from EC tax policy (for example, in the taxation of investment incomes), and the coordination of VAT administration on transactions which cross the Community’s internal frontiers is one of the main issues in EC VAT policy.
2.3 "Spillovers"
Externalities or policy spillovers provide the main economic efficiency argument for higher-tier assignment (Oates, 1972). Where the actions of each national government have consequences which are felt outside its territory, it is unlikely that policy decisions taken by national governments will fully reflect all the costs and benefits of particular policy choices; in particular the interests of non-residents are unlikely to be given adequate weighting. Olson (1969) argues that assignment of policy responsibilities between levels of government should reflect a principle of "fiscal equivalence"; the government unit responsible for a particular government function should cover a sufficiently large area to include within its boundaries all those likely to be significantly affected by its policy decisions.
Inter-country spillovers in taxation may take a number of forms.
1. Co-ordination of indirect Taxes
(i) the structure and rates of tax of a member state may have effects on private sector decisions
which extend beyond its national boundaries. In some cases the spillover effects on other member states may be clearly beneficial or damaging, as for exampie, when a particuiariy generous scheme of investment incentives attracts new investment away from other member states. In other cases the balance of costs and benefits associated with the spiliovers may be less clear-cut. As an analogue of the concept of fiscal neutrality which has proved useful as a yardstick for evaiuating the tax policies of nationai governments we may define a concept of "cross-country fiscai neutraiity" - where the differences between member states’ tax systems do not induce an ailocation of economic activity and resources between member states that differs from that which would be encountered in the absence of tax differences. As with other notions of neutrality, it is clear that cross-country neutrality shouid be regarded as a yardstick or tool for policy analysis and not as an objective in itself. Whether it is desirable that the tax system should have a neutral impact on any particular set of transactions may be affected by the existence of distortions eisewhere in the economic system; "first best" solutions may be inappropriate in a "second best" world.
(ii) inter-country spillovers may also be encountered in revenue allocation. To the extent that private sector decisions are affected by differences in tax rates, the distribution of the tax base between member states may change, ieading to a redistribution of revenues between states. For example, smaller states may be able to benefit substantially from the inflow of economic activity if they reduce their tax rates, and these gains may more than offset the revenue iosses from iower taxation on their domestic residents.
In addition, where the tax systems of member states give credit for taxes already paid elsewhere, the tax rates levied by one member state would be liable to affect the net tax receipts of other member states. Such inter-country spiliovers in revenue aliocation can arise in various systems of corporation tax and VAT ; depending on the interface between the tax systems of different member states, a rise in corporation tax or VAT rates in one state can increase the amount of credit that another has to give.
(iii) these cross-country linkages in tax revenue may also mean that there are inter-country spiilovers from the faxen/orcemenf activities of individuai member states. Better enforcement by one member state may give rise to externalities benefiting others, as a result of more effective enforcement of ciaims to be credited for taxes paid elsewhere, preventing iow-tax states becoming "tax havens" for those seeking to evade taxes in their own country, and by preventing taxabie economic activity evading tax aitogether.
The scale of cross-country externalities arising from the effects of taxation on private sector decisions are iikely to be a function of, the one hand, the extent to which member states’ tax systems differ, and, on the other, the extent of integration of different markets (for iabour, goods and services, capital, etc.) within the Community economy.
There are marked differences between member states, both in absoiute per capita tax revenues, and in tax revenues as a percentage of GDP. These differences refiect different underlying poiicies towards the ievel of government spending (and, in the short term at ieast, different attitudes towards
1. Co-ordination of indirect Taxes
State borrowing). For as long as national governments have different priorities in public spending, or economies differ in levels of income and other aspects of the tax base, differences in taxation will be inevitable. Thus, whilst a completely uniform tax system would eliminate the potential for tax-induced distortion of industrial competitiveness and location decisions within the Community market, this conclusion is of little interest. For the forseeable future, the question is not how to avoid all sources of tax-induced distortion within the Community, but simply to reduce the most serious distortionary effects whilst causing the least disturbance to member states' revenue raising powers.
The degree of market integration is a key factor determining the location of significant cross-country effects from taxation. Where large numbers of economic agents are making decisions across a range of alternatives located in different member states, the potential arises for taxation to influence the pattern of outcomes across countries. The process of European integration in different markets has been very uneven, and even after 1992 will remain so. Thus, the market for goods is already highly integrated within the Community, and, after 1992, the remaining institutional barriers to goods market integration should, it is intended, have been eliminated. Major steps will also be taken towards full integration of the capital market. However, even with the removal of a significant number of institutional barriers to labour mobility, it is likely that the labour markets of member states will remain only weakly integrated, segmented by language barriers and cultural differences that may take many years to erode. This differential integration determines where the cross-country effects of taxation are likely to be of significance, and where, over time, new cross-country issues are likely to arise.
2.4 Assignment versus Coordination
However, assigning responsibility for taxation to the Community is not the only possible response to policy spillovers, and may not be the appropriate choice in many cases. Often it may be possible for the interests of other countries in the policy decisions taken by national governments to be reflected in an agreement between the countries concerned, specifying the way in which policy will operate.
Indeed, the EC plays a mixed role - part government, part forum for international negotiation. It therefore provides scope for a range of possible degrees of constraint on national policy-making, ranging from ad hoc negotiation over the key distortionary aspects of member states’ tax systems, to the comprehensive takeover of member states’ responsibilities for a particular area of taxation. The current emphasis on subsidiarity and decentralisation implies that the former route is to be preferred wherever possible. When, however, is minimal coordination, of this form, likely to suffice? Gatsios and Seabright (1989) discuss the circumstances in which agreements on policy coordination or harmonisation will be adequate to deal with the problem of cross-country policy spillovers. They argue that the main gains from actually handing over responsibility for an area of policy with large cross-country policy externalities to the Community arise in the form of greater credibility of central operation. Where it is difficult to monitor compliance with international agreements (for example, where their implementation requires a large amount of judgement, based