Tax competition between subnational governments - Université de ...

14 downloads 490 Views 81KB Size Report
to attract mobile tax bases and firms in order to boost economic development in ... Keywords: fiscal federalism, regional and local development, decentralised tax.
1

Tax competition between subnational governments: theoretical and regional policy issues with reference to Switzerland Sergio Rossi and Bernard Dafflon University of Fribourg, Switzerland publisched in Hein E., Heise A. and Truger A., Finanzpolitik in der Kontroverse, Metropolis-Verlag, Marburg, 2004, pp. 227-250.

Abstract From a theoretical point of view, tax competition is seen as an economic policy strategy to attract mobile tax bases and firms in order to boost economic development in terms of employment and output growth within the political jurisdiction implementing it. In the literature it is further argued reduces tax revenues and hence puts a pressure on the general government sector to increase efficiency in public spending. This paper focuses on a third, neglected issue, namely that the dynamics of tax competition between subnational governments is such that a local authority will not improve its relative position within the country and none will gain any long-lasting competitive advantages.

Keywords: fiscal federalism, regional and local development, decentralised tax competition

2

Tax competition between subnational governments: theoretical and regional policy issues with reference to Switzerland*

1 Introduction and stylised facts Tax competition between regional or local authorities within a decentralized, multilevel government system has long been an issue for policy makers. This is so much so in a federal system like Switzerland, where the differences in business taxation between (and within) cantons have been rather important during the last thirty years (for a survey over the period 1970–1990 see Administration fédérale des contributions, 1990: 63–8). In the last decade these differences have neither disappeared nor diminished, as one might think on account of the federal law on the harmonization of direct taxation by the cantons and the communes that was passed on 14 December 1990.1 Empirical evidence shows, for instance, that in the 1980s the tax burden on the firms’ profit varied considerably between the 26 Swiss cantons, with maximum/minimum ratios ranging from 2 to 4 depending on the amount of profits and the formula used to compute them for fiscal policy reasons. Similar empirical evidence seems also to exist, though to a lesser extent, for communal business taxation within the same canton (Dafflon, 1991: 55–65). The 1999 OECD Economic Survey indicates that in Switzerland the average tax rate on business profits varies from 13 to 31 per cent when the three-level government taxation is considered altogether (Organisation for Economic Cooperation and Development, 1999: 109; see also Feld and Kirchgässner, 2003: 132). Such a relatively big discrepancy on business taxation between subnational governments provides a framework where tax competition and reduction of the tax burden should, theoretically, attract (or retain) firms within the jurisdiction implementing this policy. This should promote local economic development in terms of employment and output growth. In the literature it is further argued that tax competition is ‘good’ because: (i) it limits public expenditure and thus the State power, since it cuts back on the revenues accruing from business taxation and hence curbs government spending; (ii) it is an efficiency-enhancing policy that forces the general government sector to strive towards an optimal resource allocation. Yet, some argue that tax competition is ‘bad’ because tax arrangements lead to the erosion of government revenues and may therefore put the sustainability of public spending at stake. This paper focuses on another, neglected issue, namely, that the dynamics of tax competition between subnational governments is such that a local authority will not improve its relative position within the country and none will gain any long-lasting competitive advantages from such a policy, when considered from a closed-economy stance. In fact, when community A decides to cut corporate tax rates in order to gain a relative advantage on neighbouring authorities, the latter will retaliate through decreasing their own taxation, so that any differential will be levelled out *

An earlier version of this paper was presented at the 7th Workshop of the Research Network on Alternative Macroeconomic Policies held in Berlin, Germany, October 24–25, 2003. The authors would like to thank, without implicating them, Gerd Grözinger, Jochen Hartwig, Eckhard Hein, Gerhard Leithäuser, Waltraud Schelkle, and Andrew Watt for their helpful comments.

3 rapidly. Switzerland is a case in point. In addition, the dynamics of tax competition between Swiss subnational governments is such that all local authorities (cantons as well as communes) lose a part of their revenues from lowering taxation. This form of competition may eventually impinge on the implementation of the required, or planned, fiscal policies at the subnational level, where governments may indulge in granting ad hoc tax advantages to particular firms and/or their managers. This is so much so when a balanced budget requirement is taken into account. The remainder of the paper is structured as follows. In section 2 we focus on tax competition from a public choice point of view. We investigate this issue from the perspective of a decentralised government, and ask whether it is a priori better to engage in tax competition or to stand up for basic tax harmonization, information exchange, and cooperation. We show that, while there may be some arguments in favour of tax arrangements with respect to international competition and State power, this kind of competition leads, in fact, to a double loss when considered from a regional policy stance in a closed-economy setting, which we illustrate referring to Switzerland. First, owing to the dynamics of tax competition, any subnational government can take advantage from its lower tax burden only for a short period of time before other subnational units react. Secondly, this race-to-the-bottom tax policy reduces the tax yield of the targeted group of taxpayers, possibly giving rise to fiscal imbalances and/or an inequitable treatment of social groups of taxpayers. Section 3 expands on these issues. It points out that in Switzerland there actually exist some firms seeking to negotiate particular tax arrangements with a number of competing governments under the threat of moving (a part of) their activities to the fiscally most advantageous jurisdiction. To be sure, the outcome of tax competition is strongly affected by asymmetric information favouring those firms that seek specific tax advantages. Today, these ad hoc tax arrangements not only concern particular firms but also their managers, which is blatantly against the existing regulations and the federal law. To control tax competition, and hence to avoid its present shortcomings at the subnational government level, section 4 puts to the fore two policy proposals. First, a real harmonization of business taxation between subnational governments should see the light, with a single corporate tax system and a single tax rate across all Swiss cantons. Secondly, to solve the problem of asymmetric information an independent observatory should be set up for the general government sector, in order to collect all subnational government decisions granting any tax incentives or other tax advantages to particular firms. This public agency will automatically intervene if it notices or gets information about ad hoc tax arrangements, and might also be in charge of examining the effects of actual practices in terms of resource allocation and distribution.2 2 Tax competition in a prisoner’s dilemma framework Tax competition may be considered as a regional (that is, cantonal) or local (i.e. communal) strategy implemented in order to ‘bid for firms’ (Wilson, 1999: 293–4). Decentralized fiscal authorities may seek to attract, or to retain, within their territorial jurisdiction the most interesting business activities for a number of macroeconomic reasons – basically, to enhance regional economic growth and development, as well as to curb unemployment. The underlying idea is that tax cuts create conditions that eventually enlarge the tax base so that there will be a net fiscal gain at the end of the process. This rationale provides some arguments for tolerating, or even promoting, tax competition between and within countries. As the Organisation for Economic Cooperation and Development (1998: 15) notes in its report on Harmful Tax Competition, “countries with specific structural disadvantages, such as poor

4 geographical location, lack of natural resources, etc., frequently consider that special tax incentives or tax regimes are necessary to offset non-tax disadvantages, including any additional cost from locating in such areas. Similarly, within countries, peripheral regions often experience difficulties in promoting their development and may, at certain stages in this development, benefit from more attractive tax regimes or tax incentives for certain activities. This outcome, in itself, recognises that many factors affect the overall competitive position of a country. Although the international community may have concerns about potential spillover effects, these decisions may be justifiable from the point of view of the country in question.” From a theoretical perspective, another contention often raised in the literature is that tax competition constrains the Leviathan’s appetite for more State power and more public money (Pommerehne et al., 1996; Feld, 1999). As a result, the general government sector is put under pressure to increase its public policies’ efficiency, because competing governments have to carry out their spending plans with less tax revenues. In fact, this strategy may also help a country, say Switzerland, improve, or consolidate, its competitive ranking with respect to the rest of the world. For the national economy as a whole this may have a positive effect on employment and economic growth. In fact, a particular canton, say Zoug or Geneva, may be able to attract a number of business activities from abroad, without this implying a loss for another canton in terms of forsaken income and workplaces. Let us address the first of these aspects, namely, the regional/local growth argument in a closed-economy framework. In Switzerland, several cantons have recently lowered (and are still planning to lower) their taxation on business profit and capital. Further, eleven cantonal tax laws have been amended, to switch from a progressive tax rate system to a system where the tax rate on business profits is unique (see Office for Fiscal Information, 2001: D3–D18). In most cases, this structural change in the form of corporate taxation includes a significant reduction of the tax burden (between 20 and 30 per cent in four cases).3 As noted above, the rationale lying behind these amendments is to consider tax competition and the interaction of tax systems as a policy instrument to stimulate new investment and to boost regional and/or local economic development in terms of output growth and labour market clearing. As Feld and Kirchgässner (2003: 135) put it, “[d]ue to the small size of the country and its subfederal units, corporate taxpayers can easily move to places with low tax burdens and should respond to cantonal tax differentials accordingly.” Failure to coordinate cantonal tax policies at the national level, however, leads to a prisoner’s dilemma situation. Incidentally, this situation exists also between countries. In fact, at the international level coordination against fiscal havens and harmful preferential tax regimes is almost non-existent and seems even to be utopian for the time being. Yet, as pointed out by public economic theorists in the early 1980s already (Tulkens, 1985), tax competition between subnational governments within the same country does not represent an optimal solution from a political economy standpoint (Wilson, 1999). To be sure, over the long run tax competition decreases the tax burden and, therefore, the fiscal revenues of the subnational governments as a whole – unless one is ready to admit that a cooperative equilibrium can be reached between the fiscal authorities involved.4 Even if tax competition enables to control, and perhaps to limit, the expansion of the public sector, empirical evidence at the EU level shows that this kind of competition between regions exacerbates structural budgetary deficits and might put the sustainability of government debt at stake (Commission of the European Communities, 1997). Further, according to the 1998 OECD report on tax competition, globalization creates a framework where the number of offshore places (the so-called ‘fiscal havens’ imposing zero or minimal taxation) keeps increasing. To counterattack,

5 governments are thus stimulated to implement tax strategies in order to attract, or simply to retain, within their precincts those business activities that are highly mobile. At present, tax competition between governments aims indeed at limiting the moving of firms – both in terms of plants and capital – towards offshore places, by decreasing the tax burden on business profits as well as on their managers’ incomes. “Thus, tax competition for mobile capital in Switzerland may take place either by relocation of real capital leading to subsequent changes in economic activity or by profit shifting among cantons” (Feld and Kirchgässner, 2003: 135). However, this kind of competition between subnational governments (but the argument applies to nation-States as well) encompasses a series of drawbacks, namely, (i) the risk of diverting profitable trade and investment from the region, (ii) the risk of reducing the revenues of the subnational government sector considerably, and (iii) the risk of transferring part of the tax burden on less mobile factors or activities (i.e. labour and consumption), a transfer that may impinge on both employment and fiscal equity (Rossi and Dafflon, 2002). Now, since in Switzerland each canton seems ready to engage in reducing the tax burden on business profit and capital in order to attract new firms or to assist an existing firm’s expansion, any economic benefits that might result in terms of regional income or employment growth are bound to be nullified by the tax competition of other cantons. This process leads to a double loss. First, any canton can take advantage from its lower tax burden only for a short period of time – in fact, this period depends on how quickly the other cantons react in adjusting to the bottom their own tax rates, and this may take less than a fiscal year. In this framework, any tax advantage soon disappears, and the ‘after adjustment’ situation levels out the relative position of each canton to a situation very similar to the ‘before adjustment’ situation. Secondly, since such a race-to-thebottom tax policy reduces the tax yield of the targeted group of taxpayers, a series of problems may arise, namely: (i) fiscal imbalances in the current budget and/or in the current account; (ii) with a balanced budget constraint, problems of equity through reporting the tax burden on other (less mobile) groups of taxpayers; (iii) a reduction of public expenditures, most often affecting the less organized (and thus most vulnerable) groups of citizens. The self-reinforcing, downgrading process of tax competition between subnational governments can be illustrated with respect to the empirical evidence on corporate taxation available over the period 1985–2001 for the 26 Swiss cantons. Let us consider a numerical example assuming a business capital of 2 million Swiss francs and a profit of 320,000 Swiss francs (Table 1).5 [insert Table 1 around here] The example illustrated in Table 1 includes cantonal, communal and church taxation, and abstracts from inflation. It shows that all the cantons, except Geneva, have reduced their tax burden on business profits over the period considered. From 1985 to 2001, the highest rates of decrease reached well above 35 per cent in six cantons (Zoug, Appenzell R.-E., Schaffhouse, Grisons, Tessin, and Appenzell R.-I.). The cantons of Jura, Neuchâtel, and Valais, which suffer from their geographic situation at the periphery of the main economic centres, have tried to attract (or to retain) firms via an important decrease in business taxation (–32, –35, and –28 per cent respectively). Also the cantons of Schaffhouse and Berne have considerably reduced their tax burden on firms (–41 and –32 per cent) in order to keep up, or to catch up, with the huge economic development of the Basle–Zurich area during the last decade.

6

Now, as Table 2 shows, these reductions in cantonal tax burdens on firms are such that, despite a considerable decrease in business taxation in the cantons at the periphery, business activities are still attracted by the more central places, particularly by the Basle–Zurich area, where the cantons have also been able to reduce corporate taxation in a way that preserves their comparatively favourable position. This result is consistent with recent econometric analysis by Feld and Kirchgässner (2003: 151), who use a panel data set from 1985 to 1997 to show that “tax rate differences between the Swiss cantons matter, but do not exert major effects on employment although their impact on the regional distribution of taxpayers is considerable.” As a matter of fact, if the average index in Table 2 is set equal to 100 for Switzerland as a whole, over the period 1985–2001 only six out of the 25 cantons that lowered corporate taxation have passed below the 100 points threshold: Berne, Tessin, the two Appenzell, Nidwald, and Schaffhouse. Starting each with an index lower than 100 in 1985, only three cantons, Saint-Gall, Schwyz, and Zoug, have been able to maintain, then to improve an already very favourable position. In twelve cantons the index movements show that these cantons still have tax burdens much higher than average despite lower taxes in nominal terms. Vaud is a good example summarising the case in point: despite a 12 per cent decrease in corporate taxation between 1985 and 2001, its relative position rose from a favourable index of 97 points to a not too competitive position (107 points) above national average. Clearly, this outcome of tax competition is not enough to outcompete the rest of Switzerland for attracting business activities and thus enhancing regional economic development. On the contrary, as shown in Table 1, tax competition is such that, on the whole, all subnational governments lose a considerable part of their revenues because of the lowered tax burden on corporate profit and capital. [insert Table 2 around here] In short, when one considers tax competition between decentralized governments from a public choice point of view, one can notice that each canton merely reacts to the contingent situation, instead of taking the lead and obtaining a durable benefit in macroeconomic terms. In the end, firms are therefore not attracted to a specific location by this kind of competition, since the ongoing process of tax competition between subnational governments removes any competitive advantages these authorities might have one over the other in the short run by engaging in such a downgrading strategy. Note in passing that this result is also in line with several studies on business location (see Calzonetti and Walker, 1991, for a survey), which show that “the tax burden is just one factor for the location decision of private firms; many other factors can play a much more important role” (Feld and Kirchgässner, 2003: 152). 3 Tax arrangements, individual cantonal strategies, and asymmetric information The prisoner’s dilemma pointed out in the previous section exists for any canton considering the possibility of entering into tax competition with other governments at the same level. From a canton’s perspective, it implies analysing the potential costs and benefits of changing the canton’s tax regime6 and reducing the canton’s tax rates on business profit and capital for any firm that is, or will be, situated within the cantonal jurisdiction over a particular fiscal year. In a nutshell, this dilemma has a general application to the whole business sector of a canton. By contrast, there is also a specific, i.e. case-by-case, analysis that each canton must carry out when it considers tax competition. It concerns those firms that actively engage in negotiating particular tax arrangements with a number of competing governments

7 under the threat of moving (a part of) their activities to the fiscally most advantageous political jurisdiction. If successful, the bargaining strategy of these firms allows them to obtain more generous tax reductions than those already available to the business sector as a whole as a result of interjurisdictional tax competition. In other words, this competition between cantons acting individually and with no coordination at all opens the way for some firms to bargain ad hoc tax arrangements. Whenever a canton’s tax strategy is successful in attracting a particular firm by granting it an additional tax bonus, another canton will soon respond by doing the same, but to a larger extent (in terms of value and/or duration of the tax incentive), until all cantonal tax burdens are eventually reduced to a minimum owing to a lack of coordination between decentralized governments. This situation is not altered but strengthened by asymmetric information. In fact, when a firm is looking for tax advantages (basically, a reduction of its tax burden), it often starts negotiations with more than one subnational authority. This allows firms to know, and to compare, the various sorts and amounts of tax incentives offered by competing governments, whereas any of the latter does not know what the competitors are actually offering. Now, a more desirable result for the fiscally competing governments could be obtained through cooperation and an ongoing exchange of information between subnational fiscal authorities. To be sure, in Switzerland the 1948 intercantonal agreement regulating tax competition – which entered into force on 6 October 1949 and is still legally binding for the 26 cantons that signed it – forbids ad hoc tax arrangements such as those noted above (see article 1, paragraph 1, in Table 3). [insert Table 3 around here] Unfortunately, the 1948 agreement does not expand on the notion of tax arrangement. It indicates, however, that the only tax incentives a canton and/or the communes7 within its jurisdiction may offer to firms, without contravening the agreement, concern tax advantages available to those newly founded industrial firms whose development the canton is concerned with promoting (see article 1, paragraph 3, letter b).8 If we contrast the letter of this article with articles 5 and 23, paragraph 3, of the 1990 federal law on the harmonization of direct taxation by cantons and communes – which was elaborated in the spirit of the 1948 intercantonal agreement and entered into force on 1 January 1993 –, we notice nevertheless a shift in the argument for acceptable tax arrangements. In the 1990 federal law, in fact, the business taxpayers entitled to claim, and legally obtain, particular tax arrangements are those “newly founded businesses that serve the canton’s economic interest” (see articles 5 and 23, paragraph 3, in Table 3). Yet, the letter of these two articles offers a wider room for interpretation than the letter of the 1948 article: the ‘economic interest of the canton’ is indeed a much broader concept than ‘the development of the activity of industrial businesses’. For instance, fears of an upsurge in regional unemployment may justify a particular tax incentive that a canton grants to a firm (or to a group of firms) ‘in the canton’s economic interest’, as the 1990 law stipulates. This tax arrangement, however, cannot be passed in conformity with the 1948 intercantonal agreement, since the latter allows only those incentives promoting the development of an industrial activity that the canton aims to attract within its boundaries. Further, the 1990 federal law is also less rigorous than the 1948 intercantonal agreement on two additional points: (i) any reference to ‘industrial activities’ made in the latter has disappeared from the former, and (ii) article 23, paragraph 3 of the law stipulates that an ‘important extension’ of a firm’s activity may be assimilated to a newly founded firm as regards the possibility for the canton to grant it an ad hoc tax regime.9 The first point means, in fact, that small and

8 medium-sized firms may also benefit from particular tax arrangements. The second point consents to already existing firms, beside newly created businesses, to qualify for such arrangements.10 There is therefore a substantial extension of actual practices as regards tax competition tolerated by the 1990 law in respect of what is allowed by the 1948 agreement – which, incidentally, is still in force and applies cumulatively with the federal law. As public choice literature suggests, in fact, those mechanisms designed to enhance cooperation are really effective only in so far as the sanctions associated with them are credible enough to represent a serious threat for those intending to free ride or to break the law. In this respect, the 1948 intercantonal agreement does not seem up to the task. According to article 4, paragraph 4, those cantons granting tax advantages that contravene the agreement will have to pay a penalty amount into a fund administered by the Conference of the cantonal Ministers of Finance. The penalty fees enshrined in the agreement are rather nominal: they range from a minimum of 1,000 SFr. to a maximum of 10,000 SFr., with a ceiling of 50,000 SFr. for recidivists. If these amounts might have been considered high enough to disincline cantons to pass tax arrangements with taxpayers at the time when the intercantonal agreement was signed, they surely are not any more so today in an affluent country like Switzerland. Moreover, the institutional process designed in the 1948 agreement to set off a pecuniary sanction suffers from a naïve understanding of cooperative games. In fact, some might even claim that these sanctions have been purposively designed in order to be inoffensively bypassed by a ‘silent-law’, off-the-record cooperation between the cantons. As a matter of fact, although the cantons must exchange all relevant fiscal information according to the 1948 agreement, the letter, if not also the spirit, of its article 4, paragraph 3, is crystal clear as regards the working of the sanction mechanism: “The canton noticing that another canton, or one of its districts, circles, or communes, does not treat a taxpayer according to the rules approved by this agreement, or does not provide the information set out in this agreement, will make a complaint to the agreement Commission. The latter, after a contradictory process, will determine whether or not the agreement has been breached” (Commission intercantonale d’information fiscale, 1998: A16, our translation). In fact, the so-called agreement Commission – which has the legal mandate to ensure that the intercantonal agreement is correctly implemented – is elected by the Conference of the cantonal Ministers of Finance, that is, by the very fiscal authorities that are controlled by it. As a result, the sanction mechanism enshrined in the agreement lacks an authoritative, independent control body of the cantons’ tax incentives and actual practices. In this respect, the step from the lack of political independence to a lack of credibility is easily done. One may therefore conclude that the practical outcome of the 1948 intercantonal agreement is a cooperative equilibrium where all cantons are better off if each of them does not denounce the others’ tax arrangements that can be noticed, thus giving rise to a sort of silent-law situation where all the cantons are led to behave strategically. It is therefore no exaggeration to claim that, in addition to an observable general trend of decreasing business taxation on the grounds of tax competition between subnational governments, firms may obtain ad hoc tax arrangements for circumstances that are not always clear and never made explicit. To be sure, these situations are the outcome of a negotiation process, involving a firm and one or more subnational governments, that is affected by asymmetric information in favour of the former. Yet, this is not the end of the story. In fact, off-the-record tax arrangements should also be considered, because it is plain that when they are generalised beyond tax

9 competition, tax differentials as a means for attracting economic activities within the jurisdiction simply vanish. This does not imply that tax competition or tax incentives are of no significance in business location. As regional economic theory explains, in fact, the location of firms depends on a number of factors other than tax advantages (see Frey, 2002: 215–28, and Feld and Kirchgässner, 2003: 138–46). However, if one jurisdiction were to refrain from engaging in the practice of tax arrangements, it would price itself out of the market. Since “[p]ersonal tax rates in Switzerland are more important than corporate tax rates for the location of business because they are crucial for the attraction of highly skilled employees” (Feld and Kirchgässner, 2003: 146), subnational governments have recently engaged in targeting tax reliefs for individual taxpayers. As a matter of fact, empirical evidence shows that in Switzerland, at both cantonal and communal levels, there exist now a number of particular tax arrangements going beyond, or circumventing, the 1948 intercantonal agreement as well as the 1990 federal law on tax harmonization. These ad hoc tax arrangements no longer concern corporate taxpayers but individuals. They offer (for instance, by granting tax reductions to the firm’s managers) a somehow negotiable tax base or, technically speaking, an artificial definition of the tax base. Clearly, tax rates cannot be changed on a case-bycase basis – although tax coefficients might be negotiable at the local level. There exist, nevertheless, several income categories that may be removed from a manager’s taxation: they range from the firm’s subsidies to its foreign managers designed to smooth their higher cost of living in Switzerland than in other countries, to the amounts top managers earn for representing the firm abroad and that are not taxed. As noticed by the Organisation for Economic Cooperation and Development (1998: 30–31), actual practices include: (i) income deductions going beyond the various legal provisions that narrow the tax base; (ii) practices that allow costs to be deducted even though the corresponding income is not taxable; (iii) deductions for deemed expenses that are not actually incurred or that are generously estimated. These practices introduce a discriminatory bias between ‘normal’ and ‘privileged’ taxpayers and, as such, are inequitable.11

4 Conclusion and policy proposals To sum up, tax competition between Swiss subnational governments is nowadays carried out at three distinct levels, which are not fully consistent with the existing regulations or the federal law. 1. The first level of tax competition concerns the business sector as a whole. Empirical evidence over the period 1985–2001 suggests, however, that tax incentives cannot attract firms to a specific location eventually, because the ongoing process of tax competition removes any competitive advantages subnational governments might have one over the other in the short run by implementing such a downgrading policy. This policy represents a non-cooperative solution of the prisoner’s dilemma. As a result, all fiscal authorities are bound to lose a part of their revenues from lowering taxation (ceteris paribus).12 This involves a resource allocation problem. Further, a race-to-the-bottom tax policy gives rise to a fiscal equity problem when associated with a balanced budget requirement: it forces subnational governments to transfer

10 part of the tax burden to other, less mobile factors or activities, such as labour and consumption. 2. The second level of tax competition applies to particular firms being granted further, ad hoc tax advantages. It explains how the 1990 federal law harmonizing cantonal taxation extends, in fact, the 1948 intercantonal agreement about ad hoc tax arrangements for businesses. The mere fact that such arrangements are still permitted, and extended, provides an irrefutable argument that tax competition did not and does not reach its proclaimed goals. 3. The third level of tax competition, and a detestable one, is that in order to gain a competitive advantage in a disruptive situation, some cantons are ready to accept offthe-record practices that affect taxation of the firms’ managers. These practices introduce a discriminatory bias between ‘normal’ and ‘privileged’ taxpayers and, as such, are inequitable. On the whole, the lack of cooperation in direct taxation between subnational governments as well as tax competition are overall detrimental for the general government sector, in terms of forsaken fiscal revenues and economic development (Organisation for Economic Cooperation and Development, 1998: 32–3) – unless the goal of such a policy is to yoke the Leviathan.13 This kind of competition alters resource allocation and may reduce social welfare. Moreover, on the assumptions of unchanged policies and a balanced budget requirement, any tax bonus granted to a firm (or to a targeted group of firms) is bound to be compensated by an increased tax burden for other firms or individuals. This compensation introduces a system of implicit grants that threaten fiscal justice. To avoid these shortcomings, two proposals may be put forward at the policy level, one concerning tax competition as such and the other addressing the problem of asymmetric information within that framework. Let us expand on both, situating them in the Swiss framework. With respect to tax competition, a real harmonization of business taxation between subnational governments should be put into practice. In fact, tax harmonization between the Swiss cantons provided by the federal law of 14 December 1990 has no practical effect on tax competition because, according to article 1, paragraph 3, tax rates, on the one hand, and the determining and maximum amount for any tax deduction, on the other hand, are excluded from the harmonization process. If an effective tax harmonization is to see the light, a structural reform ought to be implemented, without going so far as centralising taxation powers at the federal level because of the drawbacks of such a policy (see Rossi and Dafflon, 2002: 24–25). Since in Switzerland, as in most OECD countries, statutory corporate tax rates have shrinked substantially in the last decade with respect to personal tax rates for high-income earners, conventional wisdom would suggest that both tax rates should be reduced at the same time in order to limit domestic tax avoidance. In other words, a 10 per cent decrease in business taxation, for example, might occur if, and only if, direct taxation on personal income and wealth would also be reduced by 10 per cent at the same time. Janeba and Smart’s (2003: 275) two-country model suggests that (international) agreements linking corporate tax rates to top personal tax rates are likely to limit corporate tax competition and increase revenues overall. In its most radical form, such an agreement might give rise to a single corporate tax system and a single tax rate.14 Fiscal federalism would be kept safe by distributing tax revenues between the federal, cantonal, and communal government levels according to a financial equalization formula sanctioned by the Constitution. This formula would respect the present rule for sharing the federal direct tax revenues, and would not be negotiable during annual budgetary discussions (Council of Europe, 1998: 23).

11

As regards the information problem, an independent observatory may be established for the general government sector, in order to collect all regional (that is, cantonal) and local (communal) decisions granting any tax bonus or other specific advantages to particular firms. This public agency may automatically intervene if it notices (or gets information about) ad hoc tax arrangements, and might also be in charge of examining the effects of actual practices in terms of resource allocation and distribution. It could therefore elaborate policy guidelines aiming at: (i) avoiding the prisoner’s dilemma that up to now has been characterising tax competition between subnational governments; (ii) making effective the policy relevance of both the 1948 agreement and the 1990 law on harmonization of direct taxation in the cantons and communes, particularly in respect of ad hoc cantonal arrangements for individual taxation; (iii) putting in force the relevant sanctions when needed,15 and (iv) consolidating the ranking of the national economy with respect to international competition, by an overall tax policy that is both consistent and coordinated at all the government levels (in Switzerland the federal, cantonal, and communal levels altogether). The concept of local autonomy ought to be revised accordingly and adapted to an integrated economic system within the national boundaries.

Notes 1

This law was enforced on 1 January 1993, granting however an eight-year transition period to the cantons in order for them to adjust their own tax laws accordingly. This period ended on 31 December 2000. 2 For instance, a reduction in local business taxation might increase housing prices and rents within the jurisdiction, in so far as the reduced tax burden attracts firms and their managers. This increase in land prices and shelter costs might push up the cost of living, and thus shifts income distribution in those subnational units engaging in tax competition. These issues, however, lie outside the scope of this paper. 3 Over the period 1991–2001, ten cantons have adopted a proportional corporate taxation following the canton of Jura in 1989. These cantons include: Lucerne (in 1991), Obwald (1994), Appenzell Ausser-Rh., Appenzell Inner-Rh., Nidwald and Tessin (1995), Geneva (1999), Fribourg, Schwyz and Vaud (2001). Only in three cantons, the two Appenzell and Geneva, the amount of corporate tax has increased in nominal terms, by 2, 3 and 7 per cent respectively; but taking inflation into account, there has been no increase in real terms, except for Geneva (see Table 4 at the end). 4 An example pointing towards intercantonal cooperation is the recent merging of the two offices for the promotion of regional economic development in the cantons of Vaud and Neuchâtel respectively. At present, there is a single agency (located in the canton of Vaud and headed by the former director of the Neuchâtel bureau) dealing with issues in economic development that concern either canton. 5 Generally speaking, the Swiss cantons have a system of business taxation with tax rates increasing with respect to the firm’s rate of return (i.e. the ratio between the firm’s net profit and its capital). This system is being superseded by a system where the tax rate is unique and proportional (see footnote 3 and Office for Fiscal Information, 2001). 6 Subnational fiscal authorities may, in fact, decide to switch from a tax system where the tax rate is based on the firm’s rate of return to a system with a single proportional tax rate. See the previous footnote and Table 4 at the end. 7 In most cantons, direct taxation at the communal level is a piggy-backing system: communes have to accept the basic structure of the canton’s direct tax (namely, the tax base, tax deductions, and the tax rate schedule), but can add a surtax in the form of an annual coefficient aimed at balancing the current budget. Hence there is no leeway for business tax arrangements at the communal level. However, at this level, fiscal autonomy is somewhat larger in setting user charges – communal regulations, in fact, may contain

12

selective features for which only specific users/taxpayers qualify – or in setting non-tax incentives for local industrial development. 8 As Feld and Kirchgässner (2003: 134) note, “‘[n]ewly founded’ may mean anything from the construction of new firms or affiliates to the relocation of companies traditionally having been located in other cantons for years.” 9 In both cases, the maximum duration of these particular tax regimes extends over ten years. 10 Cantons may therefore grant two ten-year periods of tax reduction, or exemption, to a particular firm: the first period on account of its being newly created, the second period arguing from an ‘important extension’ of the firm’s activity. 11 At a 2001 meeting of roughly hundred cantonal Ministers, including the Finance Ministers, in Thun, Switzerland, a number of them have pointed out the fact that these practices are blatantly against the law. As several daily newspapers reported, some of them went as far as claiming that these practices are a form of corruption by the public sector. 12 Of course, a reduction of corporate taxation might theoretically lead firms to increase wages and thus income tax earnings. This issue, however, is not germane to the argument of this paper, which focuses on the net loss of financial resources that tax competition entails for the decentralized government sector as a whole. Moreover, in practice it is doubtful whether a reduced business taxation generates higher wages for local workers and hence higher tax revenues from these workers’ taxation. Quite the contrary may happen: as a result of the reduced tax burden, firms may be led to increase their profits so as to pay out more dividends to their shareholders and be able, therefore, to attract new capital to invest in them and generate further profits. 13 Even if this were the ultimate goal, a proper method for the parliament to limit public sector growth would consist in examining the merits of the proposed public expenditures, deciding priorities, and then pruning the budget accordingly, within a balanced budget constraint. The method of reducing tax revenues first and then cutting public expenditures on the grounds of logrolling and opportunistic political behaviour would indeed harm, rather than benefit, the subnational jurisdiction. 14 Alternatively, a less rigorous solution would allow cantons to have a limited fiscal choice, say a margin of ±10 per cent with respect to the tax rate. The European Commission (1997) argues in a similar vein. 15 If the solution to set up such an independent control body with sanction powers does not seem to be democratical, or politically feasible, the alternative solution to sanction those local authorities that do not reach their proclaimed goals – say, by dismissing those politicians in charge of them, as put to the fore by Walsh (1995) in his now famous contractual approach for central bankers –, does not seem up to the task. In fact, as recognized by Walsh (1995: 152) himself, ‘[a] dismissal threat is credible [only] when the government can choose a supply of identical potential [fiscal policy makers]’. Owing to the severe institutional and legal constraints existing within any political system, this solution seems very far from reality and, moreover, would not solve the problem: politicians would be induced to put forward statecontingent reasons that might explain why they have not been able to abide by the politically (democratically) decided fiscal policy goals. Of course, if the independent control authority were created according to the proposal sketched out here, it would still be possible for the firms’ top managers to circumvent any form of control, and benefit from off-the-record tax arrangements. Nevertheless, these practices would be more difficult to implement, and indeed might be reduced to a considerable extent, because today’s silent-law framework that leads fiscal authorities to behave strategically would be disposed of by the setting up of such an independent institution.

13 References Administration fédérale des contributions (1990), Charge fiscale en Suisse 1989. Berne: Administration fédérale des contributions, Statistical Series 18. Calzonetti, F.J. and R.T. Walker (1991), ‘Factors Affecting Industrial Location Decisions: A Survey Approach’, in H.W. Herzog, Jr. and A.M. Schlottmann (eds), Industry Location and Public Policy. Knoxville: University of Tennessee Press, pp.221–240. Commission intercantonale d’information fiscale (1998), ‘Concordat entre les cantons de la Confédération suisse sur l’interdiction des arrangements fiscaux, conclu le 10 décembre 1948’, Informations fiscales, Berne: Bureau d’information fiscale, Vol.2, pp.A1–A17. Commission of the European Communities (1997), A Package to Tackle Harmful Tax Competition in the European Union. Luxembourg: Office for Official Publications of the European Communities, COM(97) 564 final. Council of Europe (1998), ‘Limitations of Local Taxation, Financial Liquidation and Methods of Calculating General Grants’, Local and Regional Authorities in Europe, Vol.65. Dafflon, B. (1991), ‘Local Business Taxation in Switzerland: Assessment, Strategies and Conflicts’, in G. Pola (ed.), Local Business Taxation: An International Overview. Milan: Vita e Pensiero, pp.41–66. Dafflon, B. (2000), ‘Faut-il baisser ou réformer la fiscalité directe?’, Revue Économique et Sociale, Vol.58, No.1, pp.19–36. European Commission (1997), ‘Code of Conduct (Business Taxation)’, Bulletin of the European Union, No.12, December, point 2.1.1. Feld, L.P. (1999), Steuerwettbewerb und seine Auswirkungen auf Allokation und Distribution: eine empirische Analyse für die Schweiz. Ph.D. dissertation, University of St. Gall and Bamberg, Vol.222. Feld, L.P. and G. Kirchgässner (2003), ‘The Impact of Corporate and Personal Income Taxes on the Location of Firms and on Employment: Some Panel Evidence for the Swiss Cantons’, Journal of Public Economics, Vol.87, No.1, pp.129–155. Frey, R.L. (2002), Wirtschaft, Staat und Wohlfahrt: eine Einführung in die Volkswirtschaftslehre am Beispiel der Schweiz. Helbing and Lichtenhahn: Basle and Münich. Janeba, E. and M. Smart (2003), ‘Is Targeted Tax Competition Less Harmful than its Remedies?’, International Tax and Public Finance, Vol.10, No.3, pp.259– 280. Office for Fiscal Information (2001), Informations fiscales de la commission intercantonale d’information fiscale. Berne: Office for Fiscal Information, Vol.2. Organisation for Economic Cooperation and Development (1998), Harmful Tax Competition: An Emerging Global Issue. Paris: Organisation for Economic Cooperation and Development. Organisation for Economic Cooperation and Development (1999), Economic Survey. Special Feature: Tax Reform in Switzerland. Paris: Organisation for Economic Cooperation and Development. Pommerehne, W., G. Kirchgässner and L.P. Feld (1996), ‘Tax Harmonization and Tax Competition at State–Local Levels: Lessons from Switzerland’, in G. Pola, G. France and R. Levaggi (eds), Developments in Local Government Finance: Theory and Policy. Cheltenham: Edward Elgar, pp.292–330. Rossi, S. and B. Dafflon (2002), ‘The Theory of Subnational Balanced Budget and Debt Control’, in B. Dafflon (ed.), Local Public Finance in Europe:

14 Balancing the Budget and Controlling Debt. Cheltenham: Edward Elgar, pp.15–44. Thierstein, A., Abegg, C. et al., (2003), Les différences spatiales des charges fiscles et compétitivité régionale, Société suisse d’études pour l’organisation de l’espace et la politique régionale OEPR, Zürich. Tulkens, H. (1985), ‘Analyse économique de la concurrence entre juridictions fiscalement souveraines’, Brussels: Institut Belge des Finances Publiques et CORE, Université de Louvain, Reprint No.676. Walsh, C.E. (1995), ‘Optimal Contracts for Central Bankers’, American Economic Review, Vol.85, No.1, pp.150–167. Wilson, J.D. (1999), ‘Theories of Tax Competition’, National Tax Journal, Vol.52, No.2, pp.269–304.

15

Table 1 Business profit taxation in Switzerland, 1985–2001, in Swiss francs, for firms with a profit of 320,000 SFr. and a capital of 2 million SFr. Canton 1985 1995 2001 1985–2001 1 2 3 4 5=(4–2)/2 Appenzell R.-E. 65,262 57,709 38,062 –42% Appenzell R.-I. 52,717 53,546 33,970 –36% Argovie 66,395 56,591 51,853 –22% Bâle-Campagne 59,159 57,247 58,467 –1% Bâle-Ville 62,854 61,070 60,591 –4% Berne 72,028 51,710 48,955 –32% Fribourg 61,614 59,149 55,029 –11% Genève 59,641 59,607 63,567 7% Glaris 71,023 67,434 59,115 –17% Grisons 87,105 78,192 54,173 –38% Jura 85,582 58,723 58,549 –32% Lucerne 65,157 49,151 49,134 –25% Neuchâtel 90,180 88,897 58,240 –35% Nidwald 57,225 47,222 38,453 –33% Obwald 59,280 48,181 46,740 –21% Saint-Gall 59,059 50,087 45,480 –23% Schaffhouse 69,863 58,770 41,538 –41% Schwyz 61,359 59,078 44,521 –27% Soleure 90,924 59,386 59,869 –34% Tessin 84,740 70,233 52,388 –38% Thurgovie 68,304 52,676 48,538 –29% Uri 67,880 59,936 58,226 –14% Valais 74,704 65,325 53,638 –28% Vaud 66,914 64,091 59,079 –12% Zoug 46,736 34,517 26,613 –43% Zurich 72,041 61,150 59,746 –17% Sources: Charge fiscale en Suisse 1985, respectively 1995 and 2001, Administration fédérale des contributions, Berne: Statistical Series, 18, 1986: 67; 1996: 69; 2002: 69.

16

Table 2 Index of fiscal burden on firms’ profit and capital in Switzerland, 1985–2001 Canton 1985 1995 2001 1985–95 1995–2001 1985–2001 1 2 3 4 5=3–2 6=4–3 7=4–2 Appenzell R.-E. 102.1 109.1 71.7 7.0 –37.4 –30.4 Appenzell R.-I. 107.0 95.1 62.0 –11.9 –33.1 –45.0 Argovie 111.5 103.0 105.2 –8.5 2.2 –6.3 Bâle-Campagne 109.6 108.8 117.4 –0.8 8.6 7.8 Bâle-Ville 107.9 113.4 118.5 5.5 5.1 10.6 Berne 118.1 95.3 91.9 –22.8 –3.4 –26.2 Fribourg 106.8 101.8 105.0 –5.0 3.2 –1.8 Genève 101.3 117.5 122.1 16.2 4.6 20.8 Glaris 131.3 145.6 106.0 14.3 –39.6 –25.3 Grisons 159.8 144.3 128.8 –15.5 –15.5 –31.0 Jura 111.8 104.7 107.9 –7.1 3.2 –3.9 Lucerne 103.7 89.3 100.2 –14.4 10.9 –3.5 Neuchâtel 138.6 138.6 111.8 0.0 –26.8 –26.8 Nidwald 106.4 86.7 70.2 –19.7 –16.5 –36.2 Obwald 94.3 87.9 98.6 –6.4 10.7 4.3 Saint-Gall 99.6 94.2 90.2 –5.4 –4.0 –9.4 Schaffhouse 109.6 101.7 85.7 –7.9 –16.0 –23.9 Schwyz 87.3 91.9 53.5 4.6 –38.4 –33.8 Soleure 120.5 97.1 107.1 –23.4 10.0 –13.4 Tessin 122.9 121.9 95.9 –1.0 –26.0 –27.0 Thurgovie 102.9 96.9 103.8 –6.0 6.9 0.9 Uri 114.1 113.6 113.6 –0.5 0.0 –0.5 Valais 118.6 114.2 109.5 –4.4 –4.7 –9.1 Vaud 97.1 105.9 106.7 8.8 0.8 9.6 Zoug 67.2 57.6 49.7 –9.6 –7.9 –17.5 Zurich 101.1 103.3 105.2 2.2 1.9 4.1 Switzerland 100.0 100.0 100.0 Sources: Charge fiscale en Suisse 1985, respectively 1995 and 2001, Administration fédérale des contributions, Berne: Statistical Series, 18, 1986: 74; 1996: 76; 2002: 76.

17

Table 3 Tax arrangements: a comparison of existing Swiss regulations The 10 December 1948 intercantonal agreement The 14 December 1990 federal law on the forbidding tax arrangements harmonization of direct taxation in the cantons and communes enforced on 6 October 1949

enforced on 1 January 1993 and definitively on 1 January 2001

article 1, paragraph 1: “The cantons agree not to conclude tax arrangements with taxpayers, and not to make use of their legal power to enter into such arrangements.”

article 1, paragraph 3, letter b:

article 5 and article 23, paragraph 3:

“It is legally allowed to grant tax facilities to those newly founded industrial firms whose development the canton has an economic interest to encouraging, for the year when the firm is created and the nine years thereafter.”

“The cantons may decide by law to grant tax advantages to newly founded businesses that serve the canton’s economic interest, for the year when the business is created and the nine years thereafter. An important extension of a business activity may be assimilated to a newly founded business.”

Sources: Concordat entre les cantons de la Confédération suisse sur l’interdiction des arrangements fiscaux; Loi fédérale sur l’harmonisation des impôts directs des cantons et des communes (LHID) (RS 642.14). Own translation.

Table 4 Cantons with proportional corporate profit taxation (fixed tax rate in per cent of taxable profit)* before the change after the change proportional Canton reference amount in reference amount in variation rate since year SFr. year SFr. 1 2 3 4 5 6 7 Jura 1989 1988 84,173 1991 58,703 –30% Lucerne 1991 1990 62,220 1993 48,096 –23% Obwald 1994 1993 62,010 1996 48,181 –22% Appenzel R.-E. 1995 1994 57,036 1997 58,360 +2% Appenzell R.-I. 1995 1994 51,349** 1997 52,663 +3% Nidwald 1995 1994 53,067 1997 47,222 –11% Tessin 1995 1994 86,580 1997 68,952 –20% Genève 1999 1998 59,481 2001 63,567 +7% Fribourg 2001 2000 58,721 2002 not avail. not avail. Schwyz 2001 2000 53,744 2002 not avail. not avail. Vaud 2001 2000 63,976 2002 not avail. not avail. * For a 320,000 SFr. profit and a 2 million SFr. capital before the tax system changed. ** Corporate businesses can choose between two tax regimes: on individual income or on corporate profit, that is, 77,421 SFr. in the first case and 51,349 SFr. in the second case.