Contribution to the Liber Amicorum in honour of Sijbren Cnossen. of the European Community ever since the signing of the Treaty of Rome in 1957.

Transcription

1 DO WE NEED TAX CO-ORDINATION? Contribution to the Liber Amicorum in honour of Sijbren Cnossen by Peter Birch Sörensen, University of Copenhagen Issues of tax co-ordination and tax harmonisation have been on the policy agenda of the European Community ever since the signing of the Treaty of Rome in The founding fathers of the EEC feared that large differences in national tax systems would distort conditions of competition in a common European market with free mobility of goods, services, capital and people. Over the years, economists and policy makers have debated whether and how national tax policies need to be co-ordinated to secure a proper functioning of the common European market. For decades, Professor Sijbren Cnossen has been in the forefront of this debate, contributing many insights and creative policy proposals 1. Written in his spirit - although he may not agree with everything that I say - the present paper is a modest attempt to summarize the case for European tax co-ordination. It goes without saying that I will have to be very selective in restating some of the arguments from this great European debate. 1 Professor Cnossen's contributions are far too numerous to mention here. Let me just draw the reader's attention to a few of the most important ones: "Harmonization of indirect taxes in the EEC", British Tax Review, vol. 4, 1983; "The case for tax diversity in the European Community", European Economic Review, vol. 34, 1990; "Company taxes in the European Union: criteria and options for reform", Fiscal Studies, vol. 17, no. 4, 1996.

2 The limits to competition The launching of the euro is likely to promote further integration of European capital markets. The formation of monetary union has therefore increased the perceived need for some form of intra-european co-ordination of taxes on capital income. After years of hard bargaining, the EU summit meeting in Feira in June 2000 established a fragile agreement among EU member states to work towards a systematic exchange of information ensuring that each country can tax the interest earned by its residents on savings invested in other member states. EU ministers also pledged to adhere to a Code of Conduct aimed at preventing harmful tax competition in the sphere of business taxation. Many economists believe that such tax co-ordination is undesirable. Even within international organizations which are officially in favour of tax co-ordination, there are sceptics. For example, in a report on the recent OECD project to counter harmful tax competition, the Economist quoted an anonymous OECD staff member who regretted the term harmful tax competition and asked how we as economists could ever consider competition in any field to be harmful? Here is a first answer: the rationale for government activity is that competition cannot be used as a mechanism for allocating resources in areas where market failures are pervasive or where the implications for income distribution would be intolerable. If governments do what they are supposed to do, they will only intervene where competition in the marketplace is unsuitable for allocating resources. The claim that fiscal competition among governments is desirable - because competition per se must always be a good thing - reflects a failure to understand that the essence of government activity is to guide resource allocation in areas where competition is bound to fail 2. 2 This is essentially the point made by Hans Werner Sinn: "The Selection Principle and market failure in systems competition", Journal of Public Economics, 1997.

3 Harmonisation versus co-ordination Debates on tax competition often fail to distinguish between tax harmonisation and tax co-ordination. Tax harmonisation means international equalization of effective tax rates, implying a serious loss of national fiscal autonomy. With cross-country differences in economic structures and political preferences, complete harmonisation of the most important taxes may involve a substantial welfare loss. By contrast, tax coordination aims to prevent distortions and inequities arising from overtaxation or undertaxation of cross-border economic activities, while respecting as far as possible the desires of nation states to choose their own preferred size of the public sector and their own preferred structure of taxation. Preserving national tax autonomy in an integrated world The recent EU agreement to initiate international exchange of information is a good example of tax co-ordination which does not involve harmonisation. Indeed, if all countries in the world could agree to exchange information to enforce residence-based taxation of worldwide income, each country would be able to choose its own preferred level of income taxation without having to fear capital flight, at least as long as the individuals owning the capital are not too mobile internationally. When governments can enforce the residence principle of international taxation, the taxpayer faces the same marginal tax rate on foreign source income and on domestic source income. His decision on where to invest his wealth will then be unaffected by the tax system, and capital will flow to those countries which offer the highest pre-tax rates of return, since this will also ensure the investor the highest after-tax return. This is exactly what is required for an efficient international allocation of capital: even though investors residing in different countries will face different national tax rates, they will choose the same pattern of investment as they would have chosen in the absence of taxes.

4 The analogue of the residence principle in income taxation is the destination principle of indirect taxation. Under the destination principle all goods and services are taxed in the country of final consumption. Goods which are exported are allowed to leave the origin country free of domestic tax, since they will be taxed in the foreign destination country, and imported goods are subject to the same indirect tax as similar goods produced domestically. Despite differences in national tax rates, domestic and foreign producers in each national market will therefore face the same indirect tax rates on their products. Hence the producer with the lowest cost of production will always be able to sell his product at the lowest price to consumers (abstracting from the costs of border tax adjustments which may leave foreign producers at a competitive disadvantage). This is just what is needed for the efficient functioning of the EU single market. Why the residence and destination principles may fail The trouble is that residence-based and destination-based taxation are very hard to implement in practice. For example, consistent destination-based taxation requires a system of border controls, but these have been abolished in the EU Internal Market, partly because they were seen as an unacceptable barrier to trade, and partly to enforce the principle of free mobility of individuals within the EU. Hence consumers can engage in border trade to take advantage of lower indirect tax rates on the other side of the border. In the area of income taxation, consistent residence-based taxation requires that foreign-source investment income be taxed on a current basis in the country of residence, even if the income is retained and reinvested abroad. For a number of reasons, this is seen as impractical and controversial. Basically, the problem is that residence-based taxation relies on the willingness of source countries to assist in collecting revenues which end up in the coffers of foreign residence countries. This raises an obvious incentive problem, so one has to wonder whether the EU member states will really manage to establish an effective system of information exchange.

5 The trouble with tax competition In the absence of extensive international information exchange, capital income can usually only be taxed in the country of source where the capital is invested. This opens the door to tax competition, as governments try to attract investment from abroad by lowering their source taxes on capital. Tax competition is likely to drive capital income taxes to inefficiently low levels. Here is the reason: for the world economy as a whole, the elasticity of capital supply is given by the interest elasticity of saving. From a global viewpoint, this is the elasticity which ought to form the basis for evaluating the welfare cost of capital income taxation. But under source-based taxation and free capital mobility national governments face the possibility of capital flight to other jurisdictions. Hence they perceive a much higher elasticity of capital supply to the domestic economy, and consequently they set a lower capital income tax rate than a policy maker adopting a global perspective. The reason why tax competition generates a suboptimal level of capital taxation may also be explained in another way. Suppose the domestic government decides to incease its source-based capital income tax, inducing a flow of capital from the domestic to the foreign economy. This capital flow will generate an increase in the value of foreign output equal to the pre-tax marginal rate of return on investment in the foreign economy, but it will only have to be rewarded by the prevailing after-tax return obtainable in the international capital market. The difference between the pre-tax and the after-tax return to the added inward investment represents a net gain to the foreign economy, taking the form of an increase in foreign tax revenue. The domestic government will tend to neglect this fiscal spillover effect on other countries, since this positive external effect does not benefit its own citizens. Hence there is a tendency for each government to set its sourcebased capital income tax at a level which is too low from the international perspective.

6 An offsetting factor: tax exporting In principle, the tendency for tax competition to drive capital taxes below their optimal level may be offset by the incentive for governments to export part of the domestic tax burden to foreign owners of domestic firms. If capital taxes fell only on the normal return to capital and if capital mobility were perfect, the government of a small open economy could not impose a real burden on foreign capital owners via a domestic source-based capital income tax, since such a tax would be fully shifted onto immobile domestic factors via an outflow of capital. However, in practice taxes on capital like the corporate income tax tend to fall on pure profits as well as on the normal return to capital, and some forms of international investment like foreign direct investment are less than perfectly mobile. This leaves scope for governments to impose a real tax burden on foreign owners of domestic firms through instruments such as a source-based corporate income tax. As economic integration proceeds, the level of capital income taxation is thus exposed to two offsetting forces. On the one hand increasing capital mobility puts downward pressure on source-based capital taxes by strengthening the incentive for tax competition. On the other hand integration also tends to increase the international crossownership of firms, and the increased presence of foreign owners generates an upward pressure on capital taxes by stimulating the incentive for tax exporting. Perhaps these offsetting tendencies explain why average effective tax rates on capital income have been roughly constant in the OECD area in recent decades, despite the sharp rise in capital mobility. However, in Europe the rough constancy of effective tax rates on capital has been accompanied by an increase in effective tax rates on labour income, so in relative terms there has been some shifting of the tax burden towards the

7 more immobile factor of production 3. Thus it seems that tax competition has protected capital from sharing in the burden of financing the expansion of the public sector. Increased inequality or underprovision of public goods? If tax competition reduces the ability of governments to raise revenue by taxing capital, will this lead to an underprovision of public goods, or will it reduce the level of redistributive public transfers, generating a more unequal distribution of income? Most of the academic literature is concerned that tax competition causes an underprovision of public goods. This literature assumes that source-based capital income taxes are the marginal source of finance of public consumption goods. As increasing capital mobility raises the elasticity of capital supply to each individual country, the perceived distortionary cost of capital taxation goes up, inducing governments to cut back on public goods provision, the argument goes. Personally I am not convinced that underprovision of public goods is the main problem with tax competition. If governments face falling capital income tax revenues, they may react by cutting back on transfer payments rather than public goods provision. Many governments pay out transfers which are close to being lump sum in character. For example, the Scandinavian governments offer a basic flat rate old age benefit to all citizens above the retirement age. At the margin, governments have the option to cut such benefits rather than cutting public service provision. In this sense it is as if the marginal supply of public goods is financed by a lump sum tax. With lump sum finance, we know from economic theory that public goods supply will not be distorted. At the margin, the government will trade off the benefits from increased public goods provision against the benefit (in terms of an improved distribution of income) from an increase in transfer 3 These trends are documented in Peter Birch Sørensen: "The case for international tax coordination reconsidered", Economic Policy 31, 2000.

8 payments. If the demand for public goods is highly inelastic w.r.t. to the tax price (including the excess burden of taxation), increased difficulties of taxing capital income will mainly be reflected in lower redistributive transfers. There is another reason why I believe that increased inequality rather than underprovision of public goods is the main problem caused by tax competition: as we have seen, tax competition tends to shift the tax burden away from mobile capital towards immobile labour. Since the distribution of capital income is more skewed than the distribution of labour income, a smaller reliance on capital taxes will tend to reduce the degree of income redistribution via the tax system. Regional co-ordination and the problem of capital flight If tax competition is such a problem, how come that many economists and policy makers remain sceptical of tax co-ordination? Let us briefly consider some of the most common arguments in favour of tax competition, starting with the problem of capital flight. Even if EU member states were to engage in a regional exchange of information to enforce the residence principle of capital income taxation within Europe, the source principle of capital taxation would still prevail in tax relations between the EU and the rest of the world, since EU governments cannot effectively monitor capital income from non-eu countries in the absence of exchange of information with these countries. Hence sceptics have argued that if the EU member states were to undertake a co-ordinated increase in their capital income taxes, they would generate a capital flight from Europe which would erode the potential welfare gain from co-ordination. This is a legitimate concern which must be taken seriously. However, there are two reasons to believe that the EU may have some scope for regional co-ordination of capital taxes, even if tax competition continues to dominate relations with the rest of the world. The first reason is that, since the EU countries have closer economic links and

9 most of them share a common currency, the degree of capital mobility tends to be higher within the EU than between the EU and the rest of the world. The frictions generating less than perfect capital mobility vis á vis non-european countries will help to contain the outflow of capital from Europe. Second, since the EU is a large player in the international capital market, a co-ordinated rise in capital taxes in Europe would significantly weaken the forces of tax competition at the global level, inducing other countries to raise their capital taxes, or at least preventing them from undertaking further reductions. This will also help to stem the flight of capital from Europe. Recent quantitative analysis based on an applied general equilibrium model suggests that these factors are sufficient to ensure that the European Union could reap a non-trivial welfare gain from regional co-ordination of capital income taxes 4. Taming Leviathan Perhaps the most popular argument againt co-ordination is that tax competition serves to constrain the rent-seeking activities of budget-maximizing bureaucrats and politicians serving special interest groups. By raising the marginal cost of public funds, international tax base mobility and the associated tax competition is said to provide a healthy check on excessive government spending. Just as the market does not always work perfectly, it would be silly to deny the existence of imperfections in the political system. Yet it always seemed to me that fostering tax competition is an odd second-best response to rent seeking. If rent seeking is a big problem, we should concentrate on institutional reform to eliminate the relevant political distortions rather than relying on tax competition which creates distortions of its own (by shifting the tax burden from mobile to immobile factors). 4 See the reference mentioned in footnote 3.

10 Of course, it may be politically infeasible to implement the institutional reforms which would eliminate wasteful government spending, but even in that case it is not clear that tax competition will be welfare-improving. A few years ago Jeremy Edwards and Michael Keen showed that - under certain simplifying assumptions - tax competition will be welfare-increasing only if the fraction of marginal public spending which is pure waste is greater than the elasticity of the capital income tax base with respect to the tax rate 5. It does not seem very likely that this condition is met in a world of high capital mobility where the capital income tax base is quite elastic. Tax competition and political commitment Another argument for tax competition is the alleged time inconsistency of the ex ante optimal capital income tax policy. In principle, governments have an incentive to levy an unanticipated (and hence non-distortionary) tax on preexisting wealth. Unless it imposes a very high capital income tax rate ex ante, the government may therefore be unable to convince the voters that they may safely save some of their income without having to fear confiscatory capital income tax rates ex post. On the other hand, if governments allow some amount of tax evasion through international capital flight, they provide the private sector with an escape route from - and so reduce the government s incentive to impose - an ex post capital levy. This might strengthen private investor confidence and stimulate wealth accumulation by lowering the time-consistent capital income tax which the government may credibly impose - or so the argument goes. I have always thought that the analytical models which are typically used to illustrate this time inconsistency problem are too simplistic. Unless the government is extremely shortsighted, it will realize that it has an interest in securing that wealth 5 See Jeremy Edwards and Michael Keen: "Tax competition and the Leviathan", European Economic Review, 1996.

11 accumulation will continue the day after tomorrow. Governments are engaged in a repeated tax policy game with the private sector, and they will realize that they will be punished via dangerously low future savings levels if they succumb to the temptation to impose a confiscatory capital levy. As a parallel, consider why patent systems protected by law can survive. Clearly, once an innovation has been produced in the expectation of patent protection, it would be optimal from a short term perspective for the government to withdraw the patent, since national income would be boosted if the new technological know-how were dispersed as quickly as possible throughout the economy. Thus an ex ante promise of patent protection does not seem time consistent. Yet patent laws are maintained, presumably because governments realize that withdrawal of patent protection of pre-existing inventions would undermine the credibility of any promise to grant patent protection to new innovations. Furthermore, people in democratic societies are not powerless in opposing unwanted government actions. As Walter Hettich and Stanley Winer have pointed out 6, the legal system in most developed nations makes it difficult for governments to unilaterally expropriate private property. Organization of political opposition is another way in which taxpayers can make themselves more difficult targets than the time inconsistency argument suggests. I do not wish to deny that time inconsistency may sometimes imply credibility problems for governments, but it seems to me that the time consistency problem of confiscatory capital income taxation has been exaggerated by many economists. 6 See Walter Hettich and Stanley Winer: "Rules, politics and the normative analysis of taxation", Working Paper, California State University and Carleton University, August 2000.

12 Is the optimal tax rate on capital zero after all? Some economists have argued that the optimal long run capital income tax rate is zero and that tax competition is useful by helping to drive capital income tax rates towards this efficient level. In most cases the claim that a zero capital income tax rate is optimal is based on the standard Ramsey infinite-horizon model where long-run equilibrium requires the after-tax real interest rate to equal the consumer s exogenous rate of time preference. This long-run constancy of the after-tax interest rate implies a horizontal long-run capital supply curve which in turn implies that the long-run interest elasticity of saving is infinitely high. This means that the burden of a capital income tax would be fully shifted to labour in the long run via a welfare-reducing fall in capital accumulation, making it futile to tax capital. But do we really believe in an infinitely high savings elasticity? I, for one, do not, since most empirical studies suggest that the interest elasticity of saving is quite low. Yet, in an intriguing analysis of an overlapping generations economy with a finite savings elasticity, Thomas Renström has shown that if today s government could commit all future governments to an optimal second-best tax plan stretching into the indefinite future, such a plan would involve a zero capital income tax rate in long-run equilibrium 7. The reason is that, in a social optimum, policy makers must be indifferent between alternative ways of transferring resources from one generation to another. If the government wanted to redistribute resources towards future generations, it could do so by inducing a rise in physical capital investment. The amount of resource transfer would be determined by the marginal productivity of capital, that is, by the interest rate before tax. Alternatively, the government could redistribute towards future generations by cutting the 7 See Thomas Renström: "On optimal taxation in overlapping generations economies", paper presented at a CEPR/Tilburg University conference on Dynamic Aspects of Taxation, Tilburg, 8-10 September 2000.

13 level of public debt. The gain to future generations would then equal the future tax savings due to lower government interest payments, and these tax savings would be given by the after-tax interest rate payable on government debt. In the social optimum the two ways of engineering a resource transfer across generations must be equally attractive. Clearly, this can only be achieved if the pre-tax interest rate equals the after-tax interest rate, that is, if the capital income tax rate is zero. Against this background one might argue that tax competition could help governments to commit to the zero capital income tax rate which seems to be optimal in the long run. The problem is that during the (long) transition to the steady state, it will generally be optimal to have a non-zero capital income tax rate, according to Renström s analysis. Unless the social rate of time preference is very low, the introduction of tax competition may therefore generate a social welfare loss by causing an undertaxation of capital during the transition to long run equilibrium. Indeed, in the type of overlapping generations economy considered here it is not clear why society should care about future generations, given that the households currently alive are assumed not to care about future generations (including their own children) at all. The ultimate case for co-ordination There are many distorting effects of capital income taxes, and it may be that the optimal capital income tax rate is lower than the optimal tax rate on labour. But in an unemployment-ridden Europe where labour is already burdened with very high tax rates, we have to ask ourselves whether we want to foster tax competition which will almost surely imply a further shift of the tax burden from mobile capital to less mobile labour? And will the present liberal economic regime with free trade and open borders be politically viable if the gains from economic integration are perceived to accrue mainly to the more mobile factors? I don t believe so, and this is another reason why I believe that some form of European tax coordination is desirable.