Towards an Internal Market without tax obstacles - Communication from the Commission - Hoofdinhoud

 

COUNCIL OFBrussels, 5 November 2001

THE EUROPEAN UNION

13365/01ADD1

FISC 224

COVER NOTE

from :Mr Bernhard ZEPTER, Deputy Secretary-General of the European Commission

date of receipt :30 October 2001

to :Mr Javier SOLANA, Secretary-General/High Representative

Subject :Towards an Internal Market without tax obstacles Communication from the Commission

Delegations will find attached a Commission Staff Working Paper on Company Taxation in the Internal Market (document SEC(2001) 1681).

________________________

COMMISSION OF THE EUROPEAN COMMUNITIES

Brussels, 23.10.2001SEC(2001) 1681

COMMISSION STAFF WORKING PAPER

Company Taxation in the Internal Market

{COM(2001)582 final}

Preface

Origin of the study

This study has been prepared by the services of the European Commission in compliance with an official mandate by the Council of Ministers. It is accompanied by a communication from the Commission to the Economic and Social Committee, the European Parliament and the Council.

Mandate to the Commission

for a study on company taxation in the European Community

The Commission is invited to present an analytical study of company taxation in the European Community. This study will be undertaken in the general context of the Vienna European Council conclusions emphasizing the need to combat harmful tax competition whilst taking into account that cooperation in the tax policy area is not aiming at uniform tax rates and is not inconsistent with fair tax competition but is called for to reduce the continuing distortions in the single market also in view of stimulating economic growth, and enhancing the international competitiveness of the Community, to prevent excessive losses of tax revenue or to get tax structures to develop in a more employment-friendly way. This study will also be undertaken on the basis of the ECOFIN Council conclusions asking to illuminate existing differences in effective corporate taxation in the Community and the policy issues that such differences may give rise to. This study should also highlight remaining tax obstacles to cross-border economic activity in the Internal Market.

The study will analyze differences in effective levels of corporate tax in Member States, taking into account, inter alia, the results of the report of the Ruding Committee (1992). Attention should be given to the influence of corporate tax bases on effective levels of taxation. Moreover, the study should also identify the main tax provisions which may hamper cross-border economic activity in the Single Market. On this basis, an assessment should be undertaken of the effects on the location of economic activity and investment. The Commission should highlight the tax policy issues involved in reducing tax induced distortions and examine possible remedial measures, taking account of the respective spheres of competence of the Member States and the Community.

Background and history of the mandate

The mandate for this study goes back to the call by the EU Ministers of Finance at their informal meeting of 26 September 1998 in Vienna for a comprehensive study on company taxation in the European Community. At that meeting, among other things, the Ministers discussed the Code of Conduct for business taxation and some suggested that further measures in the field of company taxation might be necessary in the future and asked the Commission to examine this question. The ECOFIN Council of 1 December 1998 in Vienna, in approving the first progress report of the Code of Conduct group, formally agreed to ask the Commission for this study on company taxation in the European Community. Moreover, it asked the Permanent Representatives Committee to define the concrete terms of the mandate for the study and requested the Taxation Policy Group to be consulted thereon. The European Council in Vienna on 11 and 12 December 1998 explicitly confirmed this agreement of the ECOFIN Council.

After preparatory discussions in the Taxation Policy Group and in the Council Financial Questions Group of 14 June 1999 and 24 June 1999, the Permanent Representatives Committee agreed on 22 July 1999 the official mandate to the Commission for a study on company taxation in the European Community.

Two panels of experts

In preparing the study, the Commission has been assisted by two specifically created panels of experts. The task of the first panel was to advise the Commission services on the choice of methodology for the evaluation of the effective tax rates in Member States as well as the interpretation of the qualitative and quantitative results of the analysis. The task of the second panel was to advise the Commission services on the remaining company tax obstacles to the proper functioning of the Single Market and to analyse these taxation obstacles from the point of view of the European business community and social partners.

The first panel was composed of academics and experts who have previously been involved in theoretical and empirical work related to the evaluation of effective level of company taxation. They were chosen on the grounds of their outstanding reputation and proven ability in this area. The members of panel I were

The second panel was composed of experts from among the business community and social partners at the Community level. The Commission services contacted a variety of leading business associations, trade unions and accountancy associations and invited them to designate a member of the panel. The members of panel II were:

Dr. Carlo-H. Borggreve and Roland Walter for CEEP (European Centre of Enterprises with Public Participation)

Prof. Bruno Gangemi for CFE (Conféderation Fiscale Européenne)

Dr. Piergiorgio Valente for EFFEI (European Federation of Financial Executives Institutes)

RA Alfons Kühn for Eurochambres (Association of European Chambers of Commerce and Industry);

as from April 2000 Eurochambres was represented by Dr. Harald Hendel

Wilfried Rometsch for Eurocommerce

Philip Gillett for ERT (European Round Table of Industrialists)

Prof. Sven-Olof Lodin for IFA (International Fiscal Association)

Prof. Sylvain Plasschaert for TEPSA (Trans European Policy Study Association)

Dr. Fidelis Bauer for UEAPME (European Association of Craft, Small- and Medium-Sized Enterprises); as from May 2000 UEAPME was represented by Dr. Peter Zacherl

Jos W. B. Westerburgen for UNICE (Union of Industrial and Employer's Confederations of Europe)

Christophe Quintard and Marina Ricciardelli for ETUC (European Trade Union Confederation)

Madeleine Lindblad Woodward from the Fédération des Experts Compatbles Européens took part in one meeting of the panel. The secretariat of the panel was ensured by Dr. Rolf Diemer (European Commission).

Acknowledgements

The Commission is indebted to the members of the two panels and their very helpful oral and written contributions.

The Commission is very grateful to Dr. Joann Martens Weiner (former economist, Office of Tax Policy,

U.S. Department of the Treasury) and John Neighbour (OECD secretariat) who made very valuable presentations to panel II.

The Commission also wishes to acknowledge the very helpful submissions by the Fédération des Experts Comptables Européens .

In accordance with the mandate, the Commission services bear the sole responsibility for the study and its contents. Therefore, the present report does not necessarily represent the views of all or individual members of the panels of experts.

COMPANY TAXATION IN THE INTERNAL MARKET

Executive Summary

Introduction

(1) The conclusions of the ECOFIN Council in December 1998 requested the Commission to carry out

an analytical study on company taxation in the European Union. This study should illuminate differences in the effective level of corporate taxation and identify the main tax provisions that may hamper cross-border economic activity in the Single Market. On this basis an assessment should be undertaken of the effects on the location of economic activity and investments. In July 1999 the Permanent Representatives Committee (COREPER) refined this request into a formal mandate for the Commission asking for a factual analysis and a policy assessment with a view to EU company taxation.

(2) The Commission has been assisted by two specifically created panels of experts one focussing on

the method for calculating the effective tax rates in Member States and the other on the remaining tax obstacles to the proper functioning of the Single market. The first panel was composed of academics with appropriate experience and scientific reputation in relevant theoretical works. The second panel included experts from among the business community and social partners at the Community level. The individual members of the second panel were designated by the respective organisations.

The Ruding report and the impact of the Internal Market

(3) This study takes the report of the Committee of Independent experts on Company Taxation into

account that was asked by the Commission in 1990 to determine whether differences in business taxation and the burden of business taxes among Member States lead to major distortions affecting the functioning of the Single Market and to examine all possible remedial measures (Ruding Committee). The underlying analysis of this earlier study is mostly still topical. In this context it has to be noted that little progress has been achieved in the field of company taxation as a result of its findings and recommendations. However, the context for studying company taxation in the EU has since then changed in various ways. Moreover, the mandate given to the Commission by the Council for the present study is broader than that given by the Commission to the expert committee in 1990 as it explicitly requests the analysis of tax obstacles in the Internal Market.

Company Statute), the tax impediments to cross-border activities within the Internal Market are becoming increasingly important. These elements describe important specific EU dimensions on company taxation which did not exist in the same way in 1990.

(6) EU businesses are presently confronted with a single economic zone in which 15 different company

tax systems apply. This causes losses of economic efficiency, generates specific compliance costs, and contributes to a lack of transparency. The Internal Market and Economic and Monetary Union also strongly impact on the way EU companies carry out business in the Community and set the - intended - incentive to create effective pan-European business structures. This is because EU companies increasingly no longer define one Member State but rather the whole EU as their "home market". The resulting structural changes lead to the EU-wide re-organisation and centralisation of business functions within a group of companies, many of which were traditionally present in many or even all Member States. Such re-organisation can be achieved via internal realignments, via mergers and acquisitions or through the creation of foreign branches. These tendencies, in turn, impact on the taxation of these companies. EU companies argue that their perception of the EU as their "home market" generally does not correspond to a tax reality, unlike the USA for US companies. Thus, a variety of legal and economic factors define a specific "EU dimension" for analysing company taxation.

The effective level of company taxation in the EU

(7) From the point of view of economic efficiency, tax systems should ideally be "neutral" in terms of

economic choices. In such an analytical framework, the choice of an investment, its financing or its location should in principle not be driven by tax considerations. From this perspective, and in an international context, similar investments should not face markedly different effective levels of taxation purely because of their country location. Differences in the effective levels of corporate taxation may in fact imply welfare costs because economic activity may not take place in the lowest (pre-tax) cost location by the lowest cost producers. If the impact of differences in tax regimes favours one location over another, or one producer over another, then goods may be produced at a higher pre-tax cost. Therefore, the size of these tax differentials and dispersions deserves attention.

(8) However, a full welfare cost assessment of differences in effective corporation tax rates would

require a broader analysis, taking into account the existence of other taxes and other economic parameters, as well as national preferences for equity and the provision of public goods. Moreover, to the extent that there are pre-existing distortions and/or imperfections in the market economy (market failures), taxes may be used to internalise these externalities (e.g. pollution), thereby enhancing economic efficiency. It is impossible to precisely quantify the size of tax differentials needed to correct or mitigate market failures. However, the larger the tax differentials, the larger the market failure must be unless there is to be a loss of efficiency and welfare. It should be stressed that this study has not attempted to quantify the size of any efficiency loss or welfare cost that might be associated with existing differences in effective corporation tax rates in the European Union.

provided by each country's tax law to undertake various types of investments at home or in anotherEU Member State. Second, it identifies the most important tax drivers influencing the effective tax burdens, that is the weight of each of the most important elements of the tax regimes in the effective tax burden.

(11) The analysis does not provide evidence of the impact of taxation on actual economic decisions.

Although empirical studies show that there is a correlation between taxation and location decisions, because of the weaknesses of the existing methodologies and their limitations due to lack of available data, it has been considered that none of the existing approaches could have been usefully adopted in the current study without considerably extending the range of the work.

(12) Taxation is, of course, only one of the determinants of investment and financing decisions. The

existence and quality of economic infrastructures, the availability of qualified work, as well as the short and medium-term outlook in different markets and countries are among the other important determinants of investment behaviour. The geographical accessibility of markets, transport costs, environmental standards, wage levels, social security systems and the overall attitude of government all play an important role too. Which of these factors are relatively the more important very much depends on the individual type of investment decision. Nevertheless, as economic integration in the EU proceeds in the context of the Economic and Monetary Union and the Internal Market, in an environment where capital is fully mobile, the pattern of international investment is likely to be increasingly sensitive to cross-border differences in corporate tax rules.

(13) The study presents estimates of effective corporate tax rates on domestic and transnational

investments in the 15 EU countries (as well as the US and Canada in certain cases) taking the tax systems in operation as of the year 1999. In addition, it presents estimates of effective corporate tax rates on domestic investments for the EU Member States in 2001. In view of the structure and magnitude of the German tax reform approved in 2000, the effects of this reform, as of the 1

st

January 2001, are separately analysed. The calculations consider primarily corporation taxes in each country, but also include the effect of personal income taxation of dividends, interest and capital gains.

(14) The most commonly used indicators for analysing the impact of taxation on investment behaviour

are based on forward-looking approaches which permit international comparisons and are especially tailored to provide an indication of the general pattern of incentives to investment that are attributable to different national tax laws as well as on the most relevant tax drivers that influence the effective tax burdens. In this study, the main body of the computation of the effective corporate tax burden builds on the methodology involving calculating the effective tax burden for a hypothetical future investment project in the manufacturing sector. In technical terms, the analysis relies on a revised and extended methodology of the so-called King & Fullerton approach, set out by Devereux and Griffith (1998). This computation is supplemented by data arising from the application of the "European Tax Analyzer" model which utilises the model-firm approach set out by the University of Mannheim and ZEW (1999). Considering that each methodology is based on different hypotheses and restrictions, the comparison of the results of these approaches permits the testing and, possibly, confirmation of the general trends arising from the computations.

for example, like the Ruding report, calculates effective tax rates at a given post-tax rate of return, whereas other studies

1 compute the effective tax rate for a given pre-tax rate of return. Differences

in the assumptions underlying the hypothetical investment and the economic framework can give rise to somewhat different numerical results.

(16) These approaches do not permit, for methodological reasons, taking into consideration in the

computation all the relevant features linked to the existence and functioning of different tax systems. For instance, the effects of consolidating profits and losses throughout the EU are not included because the model assumes all investments are profitable. Neither is it possible to quantify or include compliance costs. However, the most important features of taxation systems such as the rates, major elements of the taxable bases and tax systems are included. The results produced should therefore be understood as summarising and quantifying the essential features of the tax system.

(17) Effective tax rates can be calculated for a so-called "marginal" investment (where the post-tax rate

of return just equals the alternative market interest rate) or for a "infra-marginal" investment project (i.e. one that earns an extra-profit). This study has analysed both marginal and infra-marginal (average) effective company tax indicators. These reflect different hypotheses related to the underlying methodology, as well as to the domestic or international localisation of the investment, the profitability of the investment or of the firm considered, and the size and behaviour of the companies. The computations have been supplemented by "sensitivity analysis" which tests the impact of different hypotheses on the results.

(18) The broad range of data computed does not intend to present "universally valid values" for the

effective tax burden in different countries, but rather to give indicators, or illustrate interrelations, in a series of relevant situations. In fact, effective tax rates in a particular Member State depend on the characteristics of the specific investment project concerned and the methodology applied.

(19) A number of general conclusions regarding both the differences in the effective tax burdens and the

identification of the most relevant tax drivers which influence these tax burdens, can nevertheless be formulated on the basis of the results. Therefore, explanations can be given on how Member States tax regimes create incentives to allocate resources. A striking feature of the quantitative analysis is that, across the range of different situations, the relevant conclusions and interpretations remain relatively constant.

A recent study by Baker and McKenzie conducted under different hypotheses concerning the economic context and the applied tax codes, shows that in the most similar economic situation to that considered in this study (pre-tax rate of return of 6% as against a post-tax rate of return of 5% considered in the Commission study), the range of variation is 32 points in the case of a marginal investment (from 4.9% to 36.8%). When the pre-tax rate of return is fixed at 10% (base case in the Baker and McKenzie computation), the range of variation is 23 points (from 6.8 to 30.1). This study also shows that the most tax efficient method of finance is debt and that the tax systems tend to favour investments in intangibles and machinery.

(21) Differences between the effective tax burden in the EU Member States may be important for two

reasons. First, differences in effective tax rates faced by companies located in different countries, but competing in the same market, may affect their international competitiveness: two different companies, competing in the same market, may face two different tax rates. Second, when multinational companies face only the tax rate of the country where the activity takes place then differences in the effective tax rates between countries could also affect the location choice of individual activities. This can occur either as a result of the provisions of international tax codes, for example when the repatriation of profits by way of dividend from a subsidiary to a parent results in no further taxation because the dividend is exempt, or as a result of tax planning. A multinational company may therefore face different tax rates, depending on where its activities are located. As indicated, this economic reasoning is based on pure tax considerations and cannot, on its own, explain the actual behaviour of companies.

(22) Clearly, the EU wide spread cannot be explained by one single feature of the national tax system.

However, the analysis of general regimes tends to show that leaving aside preferential tax regimes

• - 

  the different national nominal tax rates on profits (statutory tax rates, surcharges and local taxes) can explain many of the differences in effective corporate tax rates between countries. Although tax regimes are designed as more or less integrated systems (in general high tax rates on profits seem to correlate with a narrower taxable base and vice versa), tax rate differentials tend to outweigh the differences in the tax bases. The quantitative analysis also shows that the relative weight of rates in determining the effective tax burden of companies rises when the profitability of the investment rises and that, consequently, any compensatory effects of a lower tax base on effective tax rates tend to disappear when the profitability rises. The study conducted by Baker and McKenzie concluded that, in general, the composition of the tax base does not have a great impact on the effective tax burden and that the level of the tax rate is the truly important factor for the difference in the tax burden.

(24) The analysis of the effective tax burden of transnational investment also gives an indication of the

allocation effects of international taxation by capturing the extent to which the tax treatment of

Corporate Tax Rates 1999 - Statutory, and Effective Average at pre tax returns of 20% & 40%

50,00%

45,00%

40,00%

35,00%

30,00%

EATR @ 20%EATR @ 40% Statutory Rate

25,00%

20,00%

15,00%

10,00%

5,00%

0,00%

ddl

lanenark

striaainurgcee

IreFinlan

edUK

Swnm

DeAurlands

Sprtuga

thePoFraneec

GrlgiumItaly

any

Berm

Ge

Nexembo

Lu

transnational investments gives incentives to undertake transnational, as opposed to domestic, investment. The data show that, on average in the EU, outbound and inbound investment are more heavily taxed than otherwise identical domestic investments and, therefore, the additional components of the transnational system add somewhat to the effective tax rates on investment.

(25) But, to the extent that companies are free to choose the most tax-favoured form of finance, then the

international tax system works such that foreign multinationals operating in a host country are likely to face a lower effective tax burden than domestic companies. This seems to be true even when the treatment of multinationals is compared with the more favourable domestic treatment allowed for small and medium sized companies.

investments are, in fact, subsidised, whereas more profitable investments suffer an effective tax burden which is in the middle range of the ranking.

(28) When the domestic analysis is updated to take into account the 2001 tax regimes, the overall

picture is broadly unchanged in comparison to 1999. However, as a consequence of a pattern of generally declining statutory tax rates (albeit with relatively small reductions apart from Germany), more profitable investments benefited from reductions in effective tax rates in a number of countries. As a result, the range of differences in domestic effective tax rates in the case of a more profitable investment decreased from 30 to 26 percentage points.

(29) The German tax reform that entered into force at 1.1.2001 is a significant reform which implies a

substantial cut in the corporation tax rate and in income tax rates, partly financed by the broadening of the tax base, including the abolition of the split rate system and the imputation system. However, despite these changes the German tax reform has only minor effects on the relative position of Germany in the EU country ranking and both the overall national corporate tax rate and the effective tax burden remains among the highest in the EU.

(30) Simulating the impact of a hypothetical harmonisation of particular features of taxation systems in

isolation on effective tax rates shows that:

Introducing a common statutory tax rate in the EU would have a significant impact by decreasing the dispersion - both between parent companies and between subsidiaries - of marginal and average effective tax rates across the EU countries. To the extent that taxation matters such a scenario would be likely to go some way in reducing locational inefficiencies within the EU.

By comparison, no other scenario would have such an impact. For example, introducing a common tax base or a system consisting in applying the definition of the home country tax base to the EU- wide profits of a multinational tends to increase the dispersion in effective tax rates if overall nominal tax rates are kept constant.

Moreover, two remarks have to be made concerning these results for a common tax base. First, the methodologies applied do not permit to take into consideration all the elements of the tax bases. However, the "Tax Analyser model", whose results are similar to those arising from the simulations of hypothetical investment, does consider a more significant number of elements of the tax bases. Second, benefits which would arise under either a common consolidated tax base or a home country tax base approach such as loss consolidation and simplified transfer pricing cannot be modelled using the methodologies used in this report.

to compensate for the removal of these financial intermediaries by making greater use of differences in general tax rates and structuring their investments to take advantage of lower rates.

Tax obstacles to cross-border economic activities in the Internal Market

(32) The Council mandate also asks for a "highlighting [of the ] remaining tax obstacles to cross-border

economic activity in the Internal Market" and calls for the identification of "the main tax provisions which may hamper cross-border economic activity in the Single Market". For this purpose the present study focuses on additional tax or compliance burdens which companies incur as a result of doing business in more than one Member State and which therefore represent a barrier to cross- border trade, establishment and investment.

(33) The underlying cause of those additional tax and compliance burdens is the existence within the

Internal Market of 15 separate tax systems. First, the fact that each Member State is a separate tax jurisdiction has a number of consequences. In particular:

· companies are obliged to allocate profits to each jurisdiction on arm's length basis by

separate accounting, i.e. on a transaction by transaction basis;

· Member States are reluctant to allow relief for losses incurred by associated companies

whose profits fall outside the scope of their taxing rights;

· cross-border reorganisations entailing a loss of taxing rights for a Member State are liable to

give rise to capital gains taxation and other charges;

· double taxation may occur as a result of conflicting taxing rights.

(34) Moreover, each Member State has its own sets of rules, in particular laws and conventions on

financial accounting, rules for determining taxable profit, arrangements for collection and administration of tax and its own network of tax treaties. The need to comply with a multiplicity of different rules entails a considerable compliance cost and represents in itself a significant barrier to cross-border economic activity. The costs and risks associated with complying with more than one system may in particular discourage small and medium-sized enterprises from engaging in cross- border activity.

Although going some way to resolving the obstacles to cross-border activity they do not provide a solution which keeps pace with the growing integration in the Internal Market.

(37) A basic concern of companies operating within the Internal Market is the removal of tax obstacles

to income flows between associated companies. The Parent-Subsidiary Directive abolishes withholding taxes on payments of dividends between associated companies of different Member States. However, its effectiveness is reduced by the fact that it does not cover all companies subject to corporation tax and applies solely to direct holdings of 25% or more.

(38) There is the further problem that - independent of the directive- certain systems of company

taxation have an in-built bias in favour of domestic investment. For example, under imputation systems applied in a number of Member States a tax credit is granted to resident (individual or corporate) shareholders for the tax paid on company level; that credit is usually not available to non-resident shareholders and is not normally granted in respect of foreign dividends. There is evidence to suggest that such systems form a serious obstacle to cross-border mergers within theEU and can have an influence on related business decisions (e.g. location of corporate seat).

(39) Payments of interest and royalties between associated companies of different Member States are

often still subject to withholding taxes that effectively create situations of double taxation. The Commission has already presented a proposal for a directive on this subject [COM(1998)67], and it is expected that this proposal will be adopted in the context of the "tax package".

(40) In addition to obstacles to income flows, corporate restructuring can also be affected by one-off

costs more directly linked to the restructuring operation itself. The tax-cost induced by cross-border mergers, acquisitions and internal reorganisations in the form of capital gains tax and various transfer taxes is often prohibitively high and forces companies to choose economically sub-optimal structures. Such obstacles place existing EU companies at a disadvantage as non-EU companies as new entrants will generally be better placed to set up the most suitable structure.

(41) The merger directive provides for deferral of capital gains charges in a number of situations.

However, a number of problems remain:

· First, not all situations are covered. Like the Parent-Subsidiary Directive, it does not include

avoidance, in some cases significantly limiting the scope of the Directive and leaving situations of double taxation unrelieved.

(42) The study identifies particular difficulties in relation to cross-border loss-compensation which,

from a business perspective, constitute one of the most important obstacles to cross-border economic activity. The current rules in Member States generally allow only for the offsetting of losses of foreign permanent establishments but not for those of subsidiaries belonging to the same group but located in different EU countries. If available, the loss compensation often takes place only at the level of the parent company or is deferred in comparison to domestic losses (which creates significant interest cost). The differences which exist in Member States' domestic loss compensation arrangements also impact on business decisions.

(43) The current loss compensation arrangements entail a risk of economic double taxation where losses

cannot be absorbed locally. This situation provides an incentive in favour of domestic investment and of investment in larger Member States.

(44) In the area of transfer pricing, the tax problems for cross-border economic activity in the Internal

Market have increased over the past years and are still growing. The problems consist essentially in high compliance costs and potential double taxation for intra-group transactions. A difficulty, according to business representatives, is that the transfer prices which are calculated for tax purposes often no longer serve any underlying commercial rationale in the Internal Market. There is in particular an increasing practice among larger companies to adopt, in EU intra-group trade, standard "euro" transfer prices for intermediate products, regardless of the production facility from which the goods are purchased within the group.

(45) There is also a tendency among Member States, fearing manipulation of transfer prices, to impose

increasingly onerous transfer pricing documentation requirements. Moreover, the application of the various methods for determining the "correct" (i.e. "arm's length") transfer price for a determined intra-group transaction is becoming increasingly complex and costly. New technologies and business structures (which imply, inter alia , more emphasis on intangibles) cause growing difficulties to identify the comparable uncontrolled transactions often required for establishing the arm's length price. In addition, there are substantial divergences in the detailed application of transfer pricing methods between Member States. The same holds for their implementation of the relevant OECD guidelines. EU businesses therefore face uncertainty as to whether their transfer prices will be accepted by the tax administrations upon a subsequent audit. The study indicates that the combined effect of these difficulties for companies can be a significant increase in compliance cost for international activities.

procedure, is rarely used and that certain of its provisions may act as a deterrent for taxpayers to make use of it.

(47) In short, the study concludes that, while there is evidence for aggressive transfer pricing by

companies, there are equally genuine concerns for companies which are making a bona fide attempt to comply with the complex and often conflicting transfer pricing rules of different countries. Such concerns are becoming the most important international tax issue for companies.

(48) The study also identifies the area of double taxation conventions as a potential source of obstacles

and distortions for cross-border economic activities within the EU. Although the intra-EU network of double taxation treaties is largely complete, there nevertheless remain some gaps. Most treaties within the EU follow the OECD Model but there are significant differences in the terms of the various treaties and their interpretation. There are also instances of divergent application of treaties by the treaty partners, leading to double taxation or non-taxation. Business representatives also refer to the increasing complexity of treaty provisions as a source of compliance cost and uncertainty. What is more, the study shows that tax treaty provisions based on the OECD Model, in particular non-discrimination articles, are not adequate to ensure compliance with the EU law principle of equal treatment. Moreover, the lack of co-ordination in the treaty practice of Member States in relation to third countries, for example regarding limitation of treaty benefits, is liable to give rise to distortions and partitioning of the Internal Market.

(49) The study also notes that certain areas of taxation which do not form part of company taxation may

nevertheless entail significant obstacles to cross-border economic activity in the EU. This notably relates to the taxation of fringe benefits and stock options, of supplementary pensions as well asVAT. It is important to note that together with the company tax obstacles these difficulties have a cumulative effect for the companies concerned. As regards VAT, this is particularly true for small and medium-sized enterprises for which the nature of the various tax obstacles to cross-border economic activity is generally identical but which suffer from disproportionately - and sometimes prohibitively - high compliance cost for dealing with them.

Remedies to the tax obstacles in the Internal Market

(50) There are essentially two approaches which could be envisaged for tackling the company tax

Moreover, ECJ rulings are confined to the particular case put to it and may therefore relate solely to individual aspects of a more general issue. The implementation of ECJ rulings is left to Member States, who often fail to draw the more general consequences which flow from them. There therefore seems to be scope for introducing a Community framework for exchanging views of the implications of significant ECJ rulings.

(53) One important example for the aforementioned principle is the problem of the bias in favour of

domestic investment in certain systems of company taxation, notably imputation systems, for which the case law of the Court has particular significance. Recent rulings, such as Safir, Verkooijen and

Saint-Gobain , suggest that tax systems which provide a disincentive to cross-border activity or investment may be contrary to the Treaty provisions on the fundamental freedoms. Such rulings raise important issues for the design of Member States' tax systems for which more guidance onEU level would be desirable.

Targeted remedial measures

(54) The various problems relating to the divergence of application of (both the existing and future) EU

Taxation Directives across Member States could be tackled via a regular exchange of best-practices and/or some form of peer review. This could also give the opportunity to develop a more common understanding of important concepts in EU company taxation, notably tax avoidance. Ensuring a more uniform application of EU tax law is an important step in order to reduce compliance costs and increase the efficiency of EU company taxation. At the same time, the need for litigation would

be reduced.

(55) The shortcomings identified in the Merger Directive and the Parent-Subsidiary Directive suggest

the need for amendment of those directives. The Commission has already presented proposals for amendment of the directives suggesting, in essence, that their scope be extended to cover other entities subject to company taxation [COM(93)293]. In addition to this and with a view to clarifying the scope of certain important provisions in the directives, notably those concerning avoidance and abuse, further amendments to the Directive and/or more detailed guidance on how those provisions should be implemented could help.

(56) As regards the merger directive, the study also identifies certain other areas where further

adopt the proposal and has ceased discussion of it. A review of the proposal conducted as part of the present study suggests that a number of technical amendments could be made to the proposal. For example, it could be envisaged calculating losses according to the rules of the State of the parent company rather than that of the subsidiary as under the proposal.

(59) Alternatively, a similar result from the company's perspective could be achieved by devising a

scheme similar to the Danish system of 'joint taxation'. In essence under the Danish arrangements a group of companies with a Danish parent company is taxed as if it were organised as a branch structure so that Denmark taxes the consolidated results of the group. The advantage of this approach over the Commission proposal lies in the greater symmetry between the taxation of profits and the offset of losses.

(60) There are a variety of measures available that would help remedy the various transfer pricing

problems. The practical application of the Arbitration Convention could certainly be improved and its provisions made subject to interpretation by the Court. Moreover, Member States could be encouraged to introduce or expand bilateral or multilateral Advance Price Agreement programmes; such instruments, although costly, are an effective means of dealing with the uncertainty relating to transfer pricing. Subject to safeguards to prevent aggressive tax planning, a framework for prior agreement or consultation before tax administrations enforce transfer pricing adjustments could also be considered.

(61) More generally, the compliance costs and the uncertainty could be reduced by better co-ordination

between Member States of documentation requirements and of the application of the various methods, for example by developing best practices. Such co-ordination could take place in the context of an EU working group and should build upon and complement the OECD activities in this field. It would be possible to develop that process further in order also to address the concerns of business. The establishment by the Commission of a Joint Forum on transfer pricing comprising representatives of tax authorities and business might allow the currently conflicting perspectives of the two sides to be reconciled. While on the one hand tax administrations view transfer pricing as a common vehicle for tax avoidance or evasion by companies and as a source of harmful tax competition between Member States, business on the other hand considers that tax authorities are imposing disproportionate compliance costs. The study finds that both sides have legitimate concerns to which it is necessary to seek a balanced solution through a dialogue on EU level. A more uniform approach by EU Member States would also contribute to a stronger position in relation to third countries.

exceptions to this basic approach which could be usefully addressed mainly at Member State level. For instance, the administrative tax formalities, bookkeeping requirements etc. for small- and medium-sized enterprises should be less demanding than for bigger companies, also in cross-border situations. Moreover, the difficulties with the cross-border offsetting of losses hit small- and medium-sized enterprises particularly hard and therefore seem to deserve a specific remedy.

Comprehensive approaches on EU company taxation

(64) The study also examines more general remedial measures aimed at minimising or removing the

obstacles in a more comprehensive manner and analyses a number of comprehensive approaches that have been presented to the Commission. All aim to address the various tax obstacles by providing multinational companies with a common consolidated tax base for their EU-wide

activities:

· Under the mutual recognition approach of "Home State Taxation" the tax base would be

computed in accordance with the tax code of the company's home state (i.e. where the headquarter is based), thus building on the existing tax systems and the related experience and knowledge. This approach is conceived as an optional scheme for companies in Member States with a sufficiently similar tax base.

· Another possibility would be to devise completely new harmonised EU rules for the

determination of a single tax base on European level. This again would be an optional scheme for companies existing as a parallel system alongside present national rules. Generally known as "Common (Consolidated) Base Taxation", this approach is advocated in particular by some business representatives.

· A further model suggested in some literature would be a "European Corporate Income Tax".

This, although originally conceived as a compulsory scheme for large multinationals, could also be an optional scheme operating alongside national rules. Under this model the tax could be levied at the European level and a part or all of the revenue could go directly to the EU.

· Finally, the more `traditional' approach would be to harmonise national rules on company

taxation by devising a single EU company tax base and system as a replacement for existing national systems.

would be reduced, many situations of double-taxation would be avoided and many discriminatory situations and restrictions would be removed.

(67) By definition, an essential element of all the solutions is that there should be group consolidation

on an EU-wide basis. At present not all Member States apply that principle even at the domestic level and only two at the international level. Under all approaches (with the possible exception of the European Corporate Income Tax) Member States would retain the right to set company tax rates.

(68) To a varying extent, all comprehensive approaches could potentially be designed such that not all

Member States would have to participate. In this context, it is important to note that the Treaty of Nice extended the possibility for enhanced co-operation by a group of Member States where agreement by all 15 is not possible. This may be particularly appropriate for Home State Taxation, which presupposes the participation solely of Member States with a fairly close tax base. However, a group of Member States could equally take advantage of this mechanism in order to introduce any

of the other approaches.

(69) A further key element of all the comprehensive approaches is a mechanism for allocating the

common consolidated tax base to the various Member States. For this purpose the USA and Canada use a formula apportionment system which allocates the tax base according to a key composed of factors such as payroll, property and/or sales. Another solution available to the EU would be to apportion the tax base according to the (adjusted) value-added tax base of the companies involved. Under all of these Member States would be allocated a specific share of the overall tax base according to apportionment keys and apply their national tax rate to that share.

(70) All the above models would meet the concerns inasmuch as they remove the need to comply with

up to 15 different tax systems, largely eliminate the transfer pricing problems arising from separate accounting and effectively provide for cross-border loss compensation. They would also provide a tax solution for the European Company. An appraisal of the various models should take account of their respective characteristics.

(71) An important point to note is that Home State Taxation does not require Member States to agree on

a new common EU base because it is based on the principle of mutual recognition by Member States of each other's tax codes. The other approaches all entail agreement on an entirely new tax code.

(73) In addition the solutions based on a parallel rather than a single compulsory system raise a number

of technical issues requiring further study. Among the main issues are those relating to restructuring, foreign income and double taxation treaties, and minority interests.

· First, as regards restructuring, since under Home State Taxation a company's tax base is

determined in accordance with the rules of its parent's state, each time the ownership of a company changes and its shares are sold the method by which it computes its tax base could change. This equates in current terminology to a potential change of residence and is potentially very costly. For example a Belgian subsidiary sold by its German Home State Taxation parent to a French parent could find its tax base changing from German to French, or if France were not participating in Home State Taxation, back to a Belgian base. In contrast, as under Common (Consolidated) Base Taxation there would only be one tax base such a sale within the Common (Consolidated) Base area would not involve such a change, and even if a company were sold to a new parent from a non participating state treatment under the Common (Consolidated) Base system could perhaps be maintained.

· Second, the treatment of foreign income under Home State Taxation, Common

(Consolidated) Base Taxation or European Corporate Income Tax is complicated by the current situation of bilateral double taxation agreements, the co-existence of exemption and credit relief tax systems and the need for a system of allocation. For example, a subsidiary in a state which operates the credit system, with a 3

rd country branch may be entitled under its

DTA to a credit for foreign tax paid by the branch. This could give rise to a claim under theDTA for the foreign tax credit even though the foreign income had been exempted under the Home State Taxation rules.

· Third, minority shareholders might find themselves receiving dividends under a taxation

system which is incompatible with their existing local personal tax system. For example a minority shareholder might receive dividends paid under a Common (Consolidated) Base Taxation or European Corporate Income Tax imputation system whereas previously dividends had been paid under the local classical system. This can only be avoided if the payment of dividends by subsidiaries to minority shareholders remains subject to the local tax code which is the approach envisaged under Home State Taxation. This would imply additional record keeping.

reliable means for ensuring that tax audits will continue to be made in an appropriate way and that none of the remedies under consideration results in illegitimate and/or illegal tax evasion.

(76) In short, the report concludes that there are potentially significant benefits to be derived from

providing, via a genuinely comprehensive solution, companies with a common consolidated tax base for the EU-wide activities. However, its findings are based mainly on the current stage of development of the research and further work would be necessary to implement any of the comprehensive approaches. Any solution going in this direction must obviously also take into account the competition rules laid down in the EC Treaty, in particular those concerning State Aids. Moreover, as already noted, the results of the quantitative analysis suggests that that the overall national tax rate is an important factor in determining the effective tax rate, and it is clear that a single or common base without further adaptations in practice would almost 'mechanically' accentuate this.

LIST OF TABLES

Table 1:Cost of capital and Effective Marginal Tax Rate

• - 

  average across all 15 EU Member States
• - 

  only corporation taxes

Table 2:Effective Average Tax Rates - average across all the 15 EU Member States

• - 

  only corporation taxes

Table 3:Cost of Capital and Effective Marginal Tax Rate

• - 

  average across all the 15 EU Member States
• - 

  top-personal tax rate, qualified shareholder

Table 4:Cost of Capital - maximum and minimum across the EU

• - 

  only corporation taxes

Table 5:Effective Average Tax Rate - maximum and minimum across the EU

• - 

  only corporation taxes

Table 6:Cost of Capital - maximum and minimum across the EU

• - 

  top-personal tax rate, qualified shareholder

Table 7:Cost of Capital and EMTR by country

• - 

  by asset, source of finance and overall
• - 

  only corporation taxes

Table 12:Germany Cost of Capital before and after the reform

• - 

  only domestic investment
• - 

  top-personal tax rate, qualified shareholder

Table 13:Cost of capital - average across all the 15 EU Member States

• - 

  only corporation taxes

Table 14:Effective Average Tax Rate - average across all the 15 EU Member States

• - 

  only corporation taxes

Table 15:Ranking of Member States by Average Cost of Capital

• - 

  only corporation taxes

Table 16:Ranking of Member States by Average EATR

• - 

  only corporation taxes

Table 17:Cost of Capital when subsidiary is financed by retained earnings

• - 

  only corporation taxes

Table 18:Cost of Capital when subsidiary is financed by new equity

• - 

  only corporation taxes

Table 19:Cost of Capital when subsidiary is financed by debt

• - 

  only corporation taxes

Table 20:EATR when the subsidiary is financed by retained earnings

• - 

  only corporation taxes

Table 21:EATR when subsidiary is financed by new equity

• - 

  only corporation taxes

Table 27:Average Cost of Capital for Germany and EU average

• - 

  domestic, average inbound and outbound

Table 28:Average EATR for Germany and EU average

• - 

  domestic, average inbound and outbound

Table 29:Summary of Simulation Results: basis results for simulations of domestic elements of corporation tax

• - 

  cost of capital and EATR

Table 30:Summary of Simulation Results: basic results for simulations of international elements

of corporation tax

• - 

  cost of capital and EATR

Table 31:Summary of Simulation Results: Interaction of Corporate and Personal Taxes

• - 

  cost of capital and EATR

Table 32:Tax Optimisation in the case of Germany; distributions by parent out of the domestic earnings

• - 

  cost of capital and EATR
• - 

  most tax efficient way, BBC and DFC

Table 33:Tax Optimisation in the case of Germany: distributions by parent out of foreign earnings

• - 

  cost of capital and EATR
• - 

  most tax efficient way, BBC and DFC

Table 34:Tax Optimisation in the case of Germany: comparison of domestic and outbound investment

• - 

  cost of capital and EATR
• - 

  most tax efficient way, BBC and DFC

Table 40:Cost of Capital and EMTR by countries, tax codes of 2001

• - 

  by assets, source of finance and overall
• - 

  only corporate taxes

Table 41:Effective Average Tax Rate by countries, tax codes of 2001

• - 

  by assets, source of finance and overall
• - 

  only corporate taxes

Table 42:Cost of Capital for Domestic Investment

• - 

  Ruding Report (1992) and Commission Study (2001)

Table 43:EMTR for domestic investment

• - 

  Baker & McKenzie report (1999) and Commission Study (2001)

Table 44:EMTR for Domestic Investment: Baker and McKenzie Report (2001) and Commission Study (2001)

• - 

  only corporate taxes

Table 45:Ruding Report (1992): Cost of capital for transnational investment

Table 46:Average Cost of Capital by Country

• - 

  Ruding Report (1992) and Commission study (2001)
• - 

  domestic, average inbound and outbound

Table 47: Mutual agreement procedures (MAP)

Table 48: The EU Arbitration Convention

Table 49: Total Mutual Agreement Procedures (including EU Arbitration Convention) of Member States

TABLES CONTAINED IN THE BOXES "TAX ANALYSER"

Table A:Effective Average Tax Rate across 5 EU Member States and the USA

• - 

  only corporation taxes

Table B:Effective Average Tax Rate across 5 EU Member States and the USA

• - 

  corporate and personal taxes

Table C:Effective Tax Rate in Germany before and after the reform

• - 

  only corporation taxes

Table D:German Tax Reform - increases and decreases attributed to different changes in taxation

• - 

  only corporation taxes

Table E:Effective Average Tax Rate in Germany before and after the Reform

• - 

  corporate and personal taxes

Table F:Results for simulations of reforming elements of the corporation tax base

• - 

  effective average tax rates

Table G:Results for simulations of reforming tax rates and local taxes

• - 

  effective average tax rates

Table H:Results for corporation tax system reform simulations

• - 

  effective average tax rate

LIST OF BOXES

Box 1 Tax Analyser:Effective Average Tax Rates (corporation level) across 5 EU Member States and the USA

Box 2 Tax Analyser:Effective Average Tax Rates (overall level: corporation and shareholders) across 5 EU Member States and the USA

Box 3 Tax Analyser:The effects of the German tax reform

Box 4 Tax Analyser:Testing the importance of the assumptions

Box 5 Tax Analyser:Impact of hypothetical tax reforms in the EU

Box 1:The globalisation of businesses

Box 2:Corporate income tax in the EU (as % of GDP)

Box 3:Properties of the measure of effective average tax rate used in the computation

Box 4:Tax provisions taken into account in the models

Box 5:The role of personal taxation

Box 6:Links between business taxation and companies' location decisions

Box 7:Description of the major tax changes in Germany

Box 8:Definition of simulations

Box 17: The imputation system and cross-border mergers: examples

Box 18:The Verkooijen judgement

Box 19: The main provisions of the Merger Directive

Box 20: Example of problems owing to the failure to update the list of companies contained in the Directive

Box 21: Example of cross-border restructuring not covered by the Directive

Box 22: Example on the taxation of exchanges of shares before the disposal of shares received in exchange

Box 23: Example of double taxation in the case of transfers of assets

Box 24:Example of the cost of a cross-border restructuring operation

Box 25: Example on restructuring operations and dividend taxation

Box 26: Loss-compensation on the domestic level

Box 27: Profit consolidation on the domestic level

Box 28: Treatment of losses of permanent establishments (foreign branches) - credit method

Box 29:Treatment of losses of permanent establishments (foreign branches) - deduction/reintegration method

Box 30: Treatment of losses of permanent establishments (foreign branches) - exemption method

Box 31: The definition of "losses"

Box 39: The development of transfer pricing documentation rules in national legislation and practice

Box 40: The Mutual Agreement Procedure (MAP) According to Article 25 of theOECD Model Tax Convention

Box 41: The EU Arbitration Convention comparison with the Mutual Agreement Provisions in Double Tax Treaties

Box 42: Articles of the EC Treaty which impose obligations concerning non- discrimination and the fundamental freedoms of the Internal Market

Box 43: The "Saint-Gobain" triangular case - far-reaching impact on tax treaties

Box 44: Tax-related labour compliance costs

Box 45: Example on stock options

Box 46: Difficulties for small and medium-sized enterprises to live up to the requirements of the present EU system of VAT

Box 47:How do multinational companies deal with differing requirements for financial accounts in 15 Member States and beyond?

Box 48:The tax treatment of leasing contracts

Box 49:The influence of accounting rules on business decisions: mergers and acquisitions

Box 50:Judgement of the Court of 17 July 1997 (Case C-28/95 Leur-Bloem)

Box 51:Article 8 of the Merger Directive

Box 59:Comparison of comprehensive approaches

Box 60:Example for reorganisations under Home State Taxation

Box 61:Example concerning the applicability of national double taxation treaties

Box 62:Example on the determination of tax credits

Box 63:Structures for analysis under Home State Taxation

Box 64:Home State Taxation and Double Taxation Agreements

Box 65:Are the obstacles removed by the comprehensive approaches?

Box 66:Example for potential differences

Box 67:USA Use of Formula Apportionment

Box 68:Canada - Use of Formula Apportionment

Box 69: Example on revenue allocation via formula apportionment versus separate accounting

LIST OF FIGURES

Figure 1:Effective Average Tax Rate and Profitability in Belgium

Figure 2:Statutory Profit Tax Rates and Effective Average Tax Rates by Member States

Figure 3:German perspective of international financial arrangements

Figure 4:UK perspectives of international financial arrangements

Figure A:Comparison of overall corporation tax bases of 5 EU Member States and the USA

Figure B:Effective Average Tax Rate across 5 EU Member States and the USA

• - 

  variation of tangible fixed assets to total balance sheet ratio

Figure C:Effective Average Tax Rate across 5 EU Member States and the USA

• - 

  variation of equity to total capital ratio

Figure D:Effective Average Tax Rate across 5 EU Member States and the USA

• - 

  variation of rate of distribution

Figure E:Effective Average Tax Rate across 5 EU Member States and the USA

• - 

  variation of equity to total capital ratio

TABLE OF CONTENTS

PART I:THE NEED FOR A STUDY OF COMPANY TAXATION IN THEEUROPEAN COMMUNITY

1.BACKGROUND TO THE MANDATE .................................................................. 12

2.SOME HISTORY AND THE IMPACT OF THE RUDING REPORT ................... 13

2.1. Earlier Commission initiatives in the area of company taxation .................... 13

2.2. The work of the Ruding Committee 1990/92 ................................................. 14

2.3. The follow-up to the Ruding report ................................................................ 16

2.4. Lessons from Ruding for the present study..................................................... 16

3.IMPORTANT GENERAL DEVELOPMENTS SINCE THE RUDING REPORT . 17

3.1. Company taxation and "globalisation"............................................................ 17

3.2. The achievement of the Internal Market ......................................................... 19

3.3. The achievement of Economic and Monetary Union...................................... 21

3.4. EU company law developments...................................................................... 21

4.BASIC ECONOMIC CRITERIA FOR ANALYSING COMPANY TAXATION IN THEEUROPEAN COMMUNITY ................................................................................... 22

4.1. General principles for the design of company tax systems ............................. 23

4.2. The economic welfare effects of company taxation systems .......................... 25

5.THE STRUCTURE OF THIS STUDY .................................................................... 27

PART II:QUALITATIVE AND QUANTITATIVE ANALYSIS OF COMPANY TAXSYSTEMS IN THE EU

A. ANALYSIS OF THE COMPANY TAX LAW

1.INTRODUCTION..................................................................................................... 28

1.1. Why a Qualitative Analysis?........................................................................... 28

1.2. What can such a qualitative analysis reveal? .................................................. 29

1.3. Findings........................................................................................................... 30

1.3.1.Tax Rates........................................................................................... 30

1.3.2.Accounting Rules .............................................................................. 30

1.3.3.Depreciation ...................................................................................... 31

1.3.4.Provisions.......................................................................................... 31

1.3.5.Losses................................................................................................ 32

1.3.6.Capital Gains..................................................................................... 33

1.3.7.Mergers and Acquisitions ................................................................. 33

1.3.8.Group Relief (`Consolidation')......................................................... 33

1.3.9.Inter Company Dividends .............................................................. 34

1.3.10. Inventories......................................................................................... 34

1.3.11. Expenses............................................................................................ 35

1.4. Conclusions..................................................................................................... 35

1.5. "Member State Tables" ................................................................................... 36

B. QUANTITATIVE ANALYSIS OF THE EFFECTIVE LEVELS OF COMPANY TAXATION

IN MEMBER STATES

4.1. The influence of domestic tax regimes on the organisation of companies' investment by

assets and sources of finance........................................................................... 96

4.1.1.Relevant economic measures: cost of capital, EMTR and EATR averaged across the EU..................................................................................... 96

4.1.2.The introduction of personal taxation ............................................... 99

4.2. Differences across the EU............................................................................. 101

4.2.1.Relevant economic measures: range of the cost of capital and EATR values across the EU................................................................................... 105

4.2.2.The introduction of personal taxation ............................................. 107

4.3. The position of the EU Member States......................................................... 108

4.3.1.Relevant economic measures: cost of capital, EMTR and EATR by Member States ............................................................................................... 108

4.3.2.The introduction of personal taxation ............................................. 117

4.4. The impact of the German tax reform........................................................... 121

4.4.1.Relevant economic measures: cost of capital and EATR before and after the reform.............................................................................................. 122

4.4.2.The introduction of personal taxation ............................................. 125

4.5. Neutrality and distortion effects: concluding remarks from the domestic analysis

....................................................................................................................... 128

5.TESTING THE ASSUMPTIONS OF THE MODEL............................................. 130

5.1. Sensitivity of the average EU cost of capital and EATR to the changes in the economic

model or level of taxes .................................................................................. 130

5.2. Impact of the sensitivity analysis on the relative position of Member States 136

6.THE TAXATION OF TRANSNATIONAL INVESTMENTS .............................. 143

7.1. Purpose of the simulations ............................................................................ 172

7.2. Scenarios involving domestic elements of corporation tax........................... 174

7.2.1.The Base Case ................................................................................. 174

7.2.2.Approximation or harmonisation of tax rates ................................. 176

7.2.3.Harmonisation of capital allowances .............................................. 178

7.3. Scenarios involving international elements of corporation tax..................... 179

7.3.1.Abolition of withholding taxes on interest...................................... 181

7.3.2.Harmonising the treatment of foreign dividends............................. 181

7.3.3.Taxation according to the parent country rules ............................... 182

7.4. Scenarios involving the relationship between corporate and personal taxes 184

7.5. Conclusions................................................................................................... 187

8.SOME EFFECTS OF TAX OPTIMISATION BY MEANS OF FINANCIALINTERMEDIARIES ON THE EFFECTIVE TAX RATES ON TRANSNATIONALINVESTMENTS BY GERMAN AND UK COMPANIES.................................... 197

8.1. Introductory remarks ..................................................................................... 197

8.2. The German parent's approach for optimising international financial arrangements

(1999) ..................................................................................................................

....................................................................................................................... 199

8.2.1.The legal framework of the analysis ............................................... 199

8.2.2.Relevant economic measures: cost of capital and EATR of a German parent and its EU subsidiaries........................................................................... 202

8.3. The UK parent's approach for optimising international financial arrangements (1999)

....................................................................................................................... 207

8.3.1.The legal framework of the analysis ............................................... 207

• 11. 

  COMPARISON OF THE RESULTS WITH THOSE OF THE RUDING AND BAKER

&MCKENZIE REPORTS ...................................................................................... 224

11.1. Comparison of the methodology and the assumptions with those of the Ruding and

Baker and McKenzie reports......................................................................... 224

11.2. Comparisons of the results in the domestic case........................................... 227

11.3. Comparisons of the results in the transnational case .................................... 235

11.4. Comparison of the results of the simulations of hypothetical tax reforms ... 240

11.5. Results from other studies............................................................................. 241

• 12. 

  CONCLUSIONS..................................................................................................... 241

PART III : COMPANY TAX OBSTACLES TO CROSS-BORDER ECONOMIC ACTIVITY IN

THE INTERNAL MARKET

1.INTRODUCTION................................................................................................... 247

2.DIVIDEND TAXATION........................................................................................ 249

2.1. Double taxation of profits and dividends distributed to corporate and individual

shareholders .................................................................................................. 249

2.2. Dividends currently not covered by the Parent-Subsidiary Directive ........... 251

2.2.1.Different methods for taxing dividend payments............................ 251

2.2.2.Obstacles relating to the imputation system.................................... 252

2.2.3.Obstacles relating to modified classical systems or shareholder relief systems

......................................................................................................... 254

2.3. Dividends covered by the Parent-Subsidiary Directive ................................ 254

2.3.1.Assessment of the functioning of the directive in practice ............. 254

2.3.2.Problems related to the implementation of the directive by Member States

......................................................................................................... 256

3.4. Conclusion .................................................................................................... 265

4.CROSS-BORDER LOSS-COMPENSATION ....................................................... 266

4.1. The tax treatment of domestic losses and foreign losses generated by permanent

establishments and subsidiaries .................................................................... 266

4.1.1.Domestic losses............................................................................... 266

4.1.2.Losses in Permanent Establishments .............................................. 268

4.1.3.Losses in subsidiaries...................................................................... 270

4.1.4.The computation of losses............................................................... 270

4.1.5.Consolidation of profits and losses ................................................. 271

4.2. Problems created by the absence of cross-border loss-compensation........... 272

4.2.1.Differing loss-compensation arrangements as factor for localisation decisions

......................................................................................................... 272

4.2.2.The cost of the absence of cross-border loss-compensation ........... 276

4.3. Conclusion .................................................................................................... 279

5.TRANSFER PRICING ........................................................................................... 279

5.1. The increasing importance of transfer pricing as an international company tax problem

and in the Internal Market ............................................................................. 279

5.2. Basic concepts of transfer pricing and the Internal Market........................... 280

5.2.1.The technicalities of transfer pricing and the OECD Guidelines.... 280

5.2.2.Fundamental tax issues of transfer pricing in the Internal Market.. 283

5.3. High compliance costs in relation to transfer pricing ................................... 287

5.3.1.The lack of comparables as reason for difficulties in the application of the arm's length principle................................................................................ 287

6.DOUBLE TAXATION CONVENTIONS.............................................................. 308

6.1. The requirement to avoid double taxation in the Internal Market................. 308

6.2. The incomplete treaty network within the EU and its insufficient scope ..... 309

6.3. Double taxation cases unresolved by double taxation treaties...................... 310

6.4. Conclusion .................................................................................................... 313

7.TAX-RELATED LABOUR COSTS ...................................................................... 313

7.1. Tax-related labour costs as a tax obstacle in the Internal Market ................. 313

7.2. Costs for occupational pension arrangements in cross-border situations ..... 314

7.3. Employee Stock Option Plans....................................................................... 317

7.4. Conclusion .................................................................................................... 321

8.SMALL AND MEDIUM-SIZED ENTERPRISES ................................................ 321

8.1. The particular situation for small and medium-sized enterprises ................. 322

8.2. Company tax obstacles and partnerships ...................................................... 324

8.3. The cumulative effects of VAT difficulties for small and medium sized enterprises

....................................................................................................................... 326

8.4. Conclusion .................................................................................................... 329

PART IV : REMEDIES TO THE COMPANY TAX OBSTACLES IN THE INTERNAL

MARKET

IV.A THE FRAMEWORK FOR POSSIBLE REMEDIES

1.

THE CASE FOR STUDYING BOTH TARGETED AND COMPREHENSIVE REMEDIALMEASURES ........................................................................................................... 330

• 2. 

  THE ROLE OF THE EUROPEAN COURT OF JUSTICE IN REMEDYING TAX

OBSTACLES IN THE INTERNAL MARKET ..................................................... 331

5.1. The classical system vs. the imputation system ............................................ 350

5.2. Remedial measures concerning the Parent-Subsidiary Directive.................. 351

5.2.1.Desirable changes to the Directive.................................................. 351

5.2.2.Closer monitoring of the implementation of the Directive ............. 353

6.REMEDIAL MEASURES TO THE TAX OBSTACLES HAMPERING CROSS-BORDERBUSINESS RESTRUCTURING OPERATIONS.................................................. 353

6.1. Relaunching the work on the adoption of Community legislation on company law

....................................................................................................................... 353

6.2. Remedial measures concerning the Merger Directive .................................. 354

6.2.1.Desirable changes to the Directive.................................................. 354

6.2.2.Closer monitoring of the implementation of the Directive ............. 354

6.3. Remedial measures concerning structuring operations not covered by the Merger

Directive........................................................................................................ 356

6.3.1."Freezing" the tax liability on the cross-border transfer of assets... 356

6.3.2.Implications of the reorganisation of companies for the Capital Duty Directive

......................................................................................................... 357

7.REMEDIAL MEASURES CATERING FOR CROSS-BORDER LOSS-COMPENSATION ................................................................................................................................. 357

7.1. The Commission proposal for a Directive (1991)......................................... 358

7.2. Fundamental issues for cross-border loss-compensation schemes ............... 359

7.2.1.Ownership threshold and indirect ownership.................................. 359

7.2.2.Recapture of losses.......................................................................... 360

7.2.3.Horizontal vs. vertical offsetting of losses...................................... 362

7.2.4.Which rules apply in computing the losses? ................................... 362

8.2.4.Co-operation between tax administrations...................................... 374

8.3. Avoiding and removing double taxation on transfer prices .......................... 375

8.3.1.Introducing mechanisms to prevent double taxation into the Arbitration Convention ...................................................................................... 376

8.3.2.Improving the dispute settlement procedures of the Arbitration Convention

......................................................................................................... 377

8.3.3.Advance Pricing Agreements.......................................................... 379

8.4. Conclusion .................................................................................................... 380

9.CO-ORDINATION OF DOUBLE TAXATION TREATIES ................................ 381

9.1. The need for co-ordinated EU double taxation treaties ................................ 381

9.2. Possible instruments...................................................................................... 382

9.2.1.Multilateral Convention .................................................................. 382

9.2.2.EC Model Treaty............................................................................. 384

9.2.3.Work on specific EU concepts........................................................ 384

9.2.4.Tax treaties with third countries...................................................... 386

9.3. Conclusion .................................................................................................... 386

• 10. 

  REMEDIAL MEASURES FOR ADDRESSING TAX-RELATED LABOUR COSTS

................................................................................................................................. 387

10.1. Pensions ..............................................................................................................

....................................................................................................................... 387

10.2. Stock options................................................................................................. 387

• 11. 

  SPECIFIC REMEDIAL MEASURES IN FAVOUR OF SMALL AND MEDIUM-SIZED

ENTERPRISES....................................................................................................... 389

• 14. 

  COMPARING THE DISTINGUISHING FEATURES OF CONCEPTUALLY DIFFERENT

METHODS ............................................................................................................. 401

14.1. Comparing a Home State `base', a Common (Consolidated) Base, and a Harmonised

Base.....................................................................................................................

....................................................................................................................... 401

14.2. Existing level of comparability ..................................................................... 405

14.3. Distinguishing between mutual recognition and harmonisation................... 405

14.4. Mutual Recognition as preparation for Harmonisation?............................... 405

• 15. 

  A PRELIMINARY EVALUATION AND ASSESSMENT ARE THE EXISTING

OBSTACLES REMOVED? ................................................................................... 406

15.1. Compliance Costs ......................................................................................... 407

15.2. Group Taxation ............................................................................................. 408

15.2.1. Groups, mergers and acquisitions ................................................... 408

15.2.2. Cross Border Loss Compensation................................................... 409

15.2.3. Dividend Taxation........................................................................... 410

15.3. Transfer Pricing Issues .................................................................................. 411

15.4. Tax related labour costs ............................................................................... 411

15.5. Double Tax Agreements ............................................................................... 412

15.6. Small and medium-sized enterprises............................................................. 413

15.7. The taxation of Partnerships ......................................................................... 414

15.8. Value Added Tax .......................................................................................... 414

15.9. Potential `new' issues and related technical issues: Foreign Income and Double Taxation

Agreements ................................................................................................... 414

15.10. Conclusion ................................................................................................... 421

16.7. Uncertainty about tax yield ........................................................................... 427

16.8. Accounting .................................................................................................... 428

16.9. The European Company Statute.................................................................... 429

• 17. 

  REVENUE ALLOCATION : THE DIFFERENT METHODS.............................. 430

17.1. General Background separate accounting and formula apportionment ..... 430

17.2. Allocation on the micro level formula apportionment............................... 432

17.2.1. The examples of USA and Canada ................................................. 432

17.2.2. Contrast to EU situation.................................................................. 433

17.2.3. Complications ................................................................................. 433

17.2.4. Despite complications in practice it works ..................................... 434

17.3. Allocation on the Micro Level `value added' ............................................ 437

17.4. Allocation on the Macro Level ..................................................................... 437

• 18. 

  ECONOMIC EFFECTS AND THE RESULTS OF THE QUANTITATIVE ANALYSIS

SIMULATIONS...................................................................................................... 438

18.1. Base costs...................................................................................................... 438

18.2. Redistribution of tax revenues between Member States ............................... 439

18.3. Quantitative Analysis Simulations relevance for Comprehensive Approaches440

• 19. 

  CONCLUSIONS..................................................................................................... 442

19.1. Equity..................................................................................................................

....................................................................................................................... 444

19.2. Efficiency ...................................................................................................... 444

19.3. Effectiveness ................................................................................................. 445

PART I:

THE NEED FOR A STUDY OF COMPANY TAXATION IN THE EUROPEAN

COMMUNITY

BACKGROUND TO THE MANDATE

The tasks given by the mandate to the Commission essentially ask for illuminating differences in the effective level of corporate taxation and, at the same time, identifying the main tax provisions that may hamper cross-border economic activity in the Single Market. The analysis should take into account, inter alia , the results of the report of the Ruding Committee (1992). The tax policy issues involved in reducing tax induced distortions should be highlighted and possible remedial measures examined. In doing so, the analysis should take into account the respective spheres of competence of the Member States and the Community.

The general background to the mandate for this study and the reasons why the Council requested a comprehensive study on company taxation in the EU can be found in the discussions on tax policy at the Vienna European Council of 11 and 12 December 1998. At Vienna, the Heads of government, in endorsing the ECOFIN Council's call for a study by the Commission on company taxation, concluded that

"Cooperation in the tax policy area is not aiming at uniform tax rates and is not inconsistent with fair tax competition but is called to reduce the continuing distortions in the single market, to prevent excessive losses of tax revenue or to get tax structures to develop in a more employment-friendly way"

3 . The mandate

for the study subsequently agreed by the Council explicitly refers to this common denominator between Member States. The statement condenses the current challenges for EU company tax systems: to achieve an efficient allocation of resources in an undistorted Internal Market, to ensure an equitable distribution of tax revenues among Member States and to guarantee the technical feasibility of taxing mobile tax factors.

It thus also creates a link between the study and the general debate on tax competition and the efforts to curb harmful tax competition in the European Community, as well as the employment effects of taxation.

In March 2000, the European Council in Lisbon placed the mandate in a new perspective: "The Union has today set itself a new strategic goal for the next decade: to become the most competitive and dynamic knowledge-based economy in the world, capable of sustainable economic growth with more and better jobs and greater social cohesion"

4 . This overall objective adds strong emphasis on the need to achieve

capital (economic activities and investment) is increasingly sensitive to taxation. Firms and individuals benefit from the freedom to move their capital to locations where the highest after-tax returns can be obtained and their investment decisions are thus more responsive to differences in effective tax rates between countries than without the Internal Market. At the same time, however, tax obstacles may still hamper the exercise of this freedom. It is therefore logical for the mandate to call for the analysis of these two - different but related - factors jeopardising allocational efficiency in the Internal Market. At the same time, it should not be overlooked that important general international developments have also significantly changed the view on international tax problems since the Ruding report was published.

Against this background, this part of the study briefly considers earlier initiatives to harmonise company taxation in the EU. It then, in accordance with the mandate, takes a look on the outcome of the work of the Ruding Committee (1990/92) and continues with an assessment of important developments which together form the framework for the subsequent analysis. Finally, some criteria for analysing company taxation in the European Community and the underlying economic considerations are explained. These criteria are used later in this study to evaluate possible solutions to the problems highlighted by the study.

SOME HISTORY AND THE IMPACT OF THE RUDING REPORT

Earlier Commission initiatives in the area of company taxation

Since the early years of the Community various committees and experts have put forward proposals for harmonising crucial elements of the corporate tax regimes of Members States. On the basis of the Treaty of Rome, the perspective was right from the start the objective to create within the Community conditions similar to a true Single Market. The first initiatives in the area of corporate income tax were thus marked by proposals for radical reform by establishing uniform rules for the core problem of the corporate income tax, e.g. the double economic taxation of companies and their shareholders. In 1962, the Neumark Committee developed concrete suggestions for the harmonisation of the company tax systems in the Community in the form of an imputation system with a split rate for retained and distributed profits

• 5. 

  At

that time, harmonisation was seen as the appropriate "soft" approach as opposed to uniform rules. The Tempel report of 1970 suggested the introduction of a classical dividend taxation system

• 6. 

  Both reports

thus identified, among other things, the tax treatment of cross-border dividend payments, unless harmonised, as a major problem within an internal market.

approximation of the rules to determine the taxable profits of enterprises. It was suggested this measure would produce greater transparency by the abolition of special incentive measures inside the tax base. The stability of the rules would also make it easier for enterprises to plan their activities in future. However, the proposals would still leave the necessary flexibility to Member States. This draft was never officially presented due to the reluctance of most Member States to support them.

These initiatives were not all successful for various reasons one of the most important of which is no doubt the unanimity requirement. Recognising the marked lack of success in progressing the above initiatives, in its communication of 1990

7 the Commission focussed on a different approach based on three ideas: direct

tax measures should be geared to the completion of the Internal Market; they should be consistent with the principle of subsidiarity and all initiatives should be defined through a consultative process with the Member States.

On that basis, and following Commission proposals, three measures - two directives and one convention - were adopted in July 1990. The Merger Directive

8 is designed to defer taxation of capital gains resulting

from certain categories of business re-organisations, in order to create within the Community conditions similar to those of an internal market. The Parent-Subsidiary Directive

9 deals principally with the

elimination of double taxation on distributed profits between a subsidiary and a parent company of another Member State. Both directives apply since 1 January 1992. They are considered in more detail below. The principal objective of the Arbitration convention

10 is to establish a procedure to resolve transfer pricing

disputes giving rise to double taxation. The convention entered into force on 1 January 1995 but its application is currently suspended as its prolongation beyond 2000 still awaits ratification in several Member States.

Two further proposals were made, both in January 1991. The first proposal aimed to abolish withholding taxes levied on cross-border interest and royalty payments between companies of different Member States. After almost four years of negotiations in the Council, no rapid progress seemed possible on this proposal and the Commission decided to withdraw it so as to be able to carry out a comprehensive review of it (November 1994). The other, the imputation of foreign losses proposal

11 is designed to allow an enterprise

to offset against its results the losses incurred by its foreign subsidiaries and permanent establishments. The proposal was discussed in the Council in 1992, but not since then. It is considered in detail below.

The work of the Ruding Committee 1990/92

burden of business taxes among Member States led to major distortions affecting the functioning of the Internal Market. The mandate was based upon three main questions:

• 1) 

  Do differences in taxation cause distortions in the functioning of the Internal Market?
• 2) 

  If such distortions arise, can they be eliminated through the interplay of market forces and tax competition or is Community action required?
• 3) 

  In the event that Community action is deemed to be necessary, what specific measures should be taken?

Unlike the Council mandate given for the present study, the mandate given to the Ruding-Committee did not explicitly call for analysing the tax obstacles to cross-border economic activity in the Internal Market.

The Ruding-Committee produced its report on 18 March 199212. Its main findings were that tax differences

can affect the location of investment and cause distortion of competition (the average cost of capital in every Member State was lowest for purely domestic investments); and that some convergence had happened in the past but the main distortions could not be reduced solely through market forces or through independent action of Member States. The Committee issued recommendations that fell essentially into two categories: (i) on the elimination of double taxation and (ii) on corporation tax (rate, base, system).

Among other things, the Ruding recommendations were very favourable about the three measures agreed in 1990 and recommended the further extension of the two directives. The recommendations also welcomed the proposals for directives made in 1991. A detailed list of the recommendations of the Ruding report and the follow-up is presented in Annex 1.

The underlying approach of the recommendations appears to be one of a "soft" tax harmonisation designed to establish a level playing field for free and fair competition between Member States, by setting minimal standards for European tax legislation. The Community would thus not impose uniform rules, but only the basic standards which the Member States should observe in designing their tax system. This requires for example minimal standards with respect to the basic tax rate, and maximal standards with respect to what could be allowed for systems and rates of depreciation, provisions, and the treatment of stock in trade. These standards would only determine the limits beyond which Member States could not compete with their tax systems. Within these boundaries Member States would remain free to determine their own tax systems. In some cases such as depreciation of goodwill, harmonisation would mean that all Member States accept the same rule i.e. either a common positive or a common negative answer to the question of depreciation. This softer approach to harmonisation would still leave Member States with considerable room for manoeuvre.

The follow-up to the Ruding report

The Commission indicated in its response to the report of June 199213 that priority should be given to the

elimination of double taxation on cross-border income flows. A more qualified assessment was given of the second part of recommendations, as some of these seemed to go beyond what was strictly necessary at Community level. It was suggested that the proposed measures could have the effect of reducing the tax base, which might in turn involve an increase in tax rates.

The Council conclusions on company taxation of November 199214 introduced a number of criteria that

should be taken into account in deciding whether action was appropriate at Community level. The need to eliminate double taxation was however recognised. At the same time, the need to ensure effective single taxation was stressed.

On, among other things, the basis of the Ruding recommendations, in July 1993 the Commission published two proposals to amend the two Directives of 1990

15 which were designed to extend the scope of these

directives and notably to include more legal forms of enterprises. Both proposals received a favourable opinion of the European Social and Economic Committee and the European Parliament and are pending in the Council. However, so far no unanimous agreement could be reached in the Council of Ministers.

Since the mid-nineties, given the limited success of the earlier initiatives, a more comprehensive approach to tax policy has been reflected in EU tax policy discussions. In the area of direct taxation, the "traditional" harmonisation approach was complemented by the notion of tax co-ordination. At the informal ECOFIN meeting at Verona in April 1996, the Commission, contrasting the need for progress on tax matters in theEU with the limited number of actual decisions adopted in this area thus far, proposed a new and comprehensive view of taxation policy. This approach resulted in the 1997 tax package

16 to eliminate

harmful tax competition within the EU which to date forms the most important ongoing EU initiative in the area of direct taxation. As noted above, it is in this context that the Council asked for a comprehensive study on company taxation in the EU to be undertaken by the Commission.

Lessons from Ruding for the present study

Generally, the basic problems raised and most of the issues considered by the Ruding-Committee are still relevant. Any current analysis of EU company tax problems can therefore usefully take into account the work presented in the Ruding report. After ten years, however, the analysis needs to be updated in many respects. First and foremost, the Internal Market and, for most Member States, also Economic and Monetary Union is now a well-established reality whereas it was only a prospect for the Ruding Committee. In combination with the relative lack of progress on company tax issues on EU level this means that the existing problems highlighted by Ruding have now become even more acute. The "tax package" of 1997 has shifted attention from distortion of market competition through basic structural elements of the tax system, to distortions caused by specific privileged tax regimes. At the same time the notions of "legitimate protection" of tax revenue and its equitable distribution Member States were introduced. Finally in fields closely related to taxation such as financial accounting law and company law

new developments took place like the increasing influence of international accounting standards on tax accounting and the agreement on the European Company Statute.

Moreover, the economic framework and business strategies have significantly changed since 1992. Technological developments and more open and deeply-integrated markets impact on the behaviour of companies and it is necessary to look into possible repercussions this may have on the taxation of these companies, especially in cross-border situations.

In essence the basic analysis of the Ruding report still remains valid today. Because of deeper integration, because also of new developments in tax competition with more emphasis on specific tax regimes and the equitable distribution of tax revenue among Member States, the pressing need for tax co-ordination has become much clearer. However since the Ruding report very little has been achieved in the field of specific regulation, and in that respect its impact has been disappointing.

IMPORTANT GENERAL DEVELOPMENTS SINCE THE RUDING REPORT

Company taxation and "globalisation"

Although it sounds like a hackneyed phrase it is nevertheless true that the globalisation process has significantly gained momentum since the Ruding Committee produced its report. It has profoundly changed the international economic landscape and, subsequently, created new challenges for national company tax systems. Globalisation means, among other things, more integration of international markets due to new technological possibilities and the gradual reduction of market access barriers. This development is, as such, independent of the Internal Market, but not surprisingly its effects are particularly strong for countries that are already integrated in one market in which the liberalisation process is relatively advanced (e.g. telecommunications, energy, public procurement, financial services). The same holds for the companies that are based and operating in that integrated market. The result is increasing competition, both between market operators and between Member States. Tax competition is an increasingly important aspect

of the latter.

Generally, the basic elements of the tax systems of most countries were established when economies were relatively closed, capital movements limited and information technologies less developed than today. Inasmuch as the tax systems of some countries do not yet reflect recent economic and technological developments there is scope for companies to exploit loopholes and for other countries to try to attract business from those countries. Hence, the opening of EU economies within the Internal Market and of that Internal Market towards the rest of the world makes a case for collective action, in particular on the co- ordination of EU company tax systems.

purchases in 1999 18, representing more than two thirds of the value of all world-wide mergers and

acquisitions.

This development impacts on the way in which companies and tax administrations confront the taxation of cross-border mergers and acquisitions. When the Ruding report was written, these trends were already marked but were far from being as strong as they are today.

Box 1

The globalisation of businesses 19

The following figures illustrate that companies increasingly operate, in various facets, on a multinational scale. Tax administrations however broadly continue operating on a national scale.

· The number of multinational enterprises has increased from some 7,000 parent firms in 15 developed (EU and non-EU)

countries at the end of the 1960s to some 40,000 at the end of the 1990s. There are now approximately 63,000 parent firms and 690,000 foreign affiliates operating world-wide.

· Accordingly, international production, trade and investment have increased significantly. Sales of foreign affiliates world-

wide accounted for an estimated $13.6 trillion in 1999, compared to about $2.5 trillion in 1980, a figure twice as high as that of global exports. Multinational enterprises now account for about one-tenth of global GDP, compared to one-twentieth in 1982.

· This corresponds to a broad increase in foreign direct investment (FDI). The ratio of world FDI inflows ($865 billion in

1999) to gross domestic capital formation is now 14 %, compared to 2 % twenty years ago. In the same period, the ratio of world FDI stock to world GDP increased from 5 % to 16 %.

· At the same time, both the number and the value of mergers and acquisitions have increased significantly. The value of all

mergers and acquisitions (cross-border and domestic) as a share of world GDP has risen from 0.3 % in 1980 to 8 % in 1999 while the value of completed cross-border mergers and acquisitions rose from less than $100 billion in 1987 to $720 billion in 1999. The total number of all mergers and acquisitions world-wide has grown at 42% annually between 1980 and 1999.

Looking towards the future, the structural technological changes driving the globalisation process are creating new challenges for taxation and may introduce "tax termites"

20 into national tax systems. The

increasing use of electronic commerce could become a fundamental problem for the correct taxation at a national level of company profits

• 21. 

  Electronic transactions leave far fewer identifiable traces than "real"

transactions and many traditionally "physical" products are becoming virtual (e.g. software, music, films or educational services). This makes it increasingly difficult to identify the economic operators, the territory from which a transaction is made, etc.

Some commentators suggest that the taxation of corporate profits already today a relatively minor source of state revenue could eventually vanish as it will no longer be enforceable. Others argue that there is ultimately no economic case for taxing company profits as only individuals eventually bear taxes. However, others see economic justifications in taxing companies that consume public goods and stresses the practical link in levying both a corporate tax and a personal income tax. This study does not attempt to rehearse those arguments further; it is written on the clear assumption that company taxes will continue to be levied in the EU for the foreseeable future.

In this context it is however worth noting that, generally, corporate income tax has been fairly stable source of revenues for European governments in the past 10 years, after a period of growth in the 1970s and 1980s. Thus, at this point in time, there seems to be little empirical evidence of a "race to the bottom". Corporate income tax as a percentage of GDP varies considerably between the Member States. The following box gives some information in this respect.

Box 2

Corporate income tax in the EU (as % of GDP)

197019801990199619971998199920002001* 2002*

Belgium2,42,22,43,13,53,63,53,53,53,4

Denmark1,11,52,63,43,73,63,63,53,53,5

Germany1,71,81,81,71,91,92,02,11,91,9

Greece-0,51,72,22,42,93,23,33,33,2

Spain-1,23,12,12,12,12,12,12,12,1

France2,22,12,41,92,22,72,92,92,82,8

Ireland1,31,52,23,63,73,73,73,53,43,3

Italy3,02,43,74,24,33,94,14,13,93,9

Luxem- bourg5,97,66,66,98,38,38,28,28,07,6

tax package and notably the Code of Conduct for business taxation is addressing the issue of harmful tax competition. But in an increasingly integrated Single Market in which harmful forms of tax competition are being removed, the competition effects of the general features of EU company tax systems become significantly more important. The Internal Market thus accentuates general tax competition between Member States and it needs to be assessed which welfare effects this increased competition has. The analysis of part II of this study can be appraised in this context.

Moreover, the Internal Market has affected the perspective of EU companies: they now increasingly change their focus from the national state towards the Union as a coherent economic zone. This is evidence of the success of the Internal Market. However, it makes all the more urgent to address the remaining tax obstacles that prevent EU companies from exploiting its full benefits and that still bias companies towards national rather than multinational economic activity. EU businesses have to deal with 15 company tax systems and tax administrations in one market. This creates efficiency losses and unnecessary compliance costs. These run against the potentially high positive welfare effects relating to the opening of national markets via Internal Market integration and puts EU businesses at a relative competitive disadvantage (compared to third country operators). These issues are considered in part III.

More specifically, when EU multinational companies define the EU as their home market they generally wish to re-align their business structures accordingly by creating pan-European business units instead of country-based organisations. Small and certainly medium-sized companies can also face this problem. Although operating at a smaller scale and in fewer Member States, the basic idea of a market for goods and services extending beyond the domestic market (be it via e-commerce or distance selling) translates into practical business decisions and subsequent tax considerations for small and medium-sized enterprises. The Internal Market thus concretely determines the way EU companies carry out their business within the Community.

The creation of pan-European business structures can essentially be achieved by three means (which are not mutually exclusive): (i) cross-border intra-group restructuring and expansion within the EU by way of acquisitions or joint ventures, (ii) fully-fledged mergers and (iii) establishment of foreign branches. Within the EU, this trend is clearly driven by the Internal Market. However, in a broader perspective it is reinforced because businesses tend to concentrate their activities more and more on core activities (on EU level or beyond), thus disposing of non-related business units and in turn purchasing related ones from other companies. The result of this process of business re-alignment is that the number of cross-border mergers and acquisitions and of intra-group transactions cross-border can be expected to rise even further.

relating to transfer pricing take on a new dimension. While this issue is now dominating large parts of the current discussions on international taxation it was hardly mentioned in the Ruding report.

As regards the re-alignment of other functions such as marketing, R&D and group financing, these generally will be centralised either at the head office or in designated countries throughout Europe. Consequently the costs of these functions have to be allocated through some sort of cost sharing mechanism. Such cost sharing arrangements, sometimes involving a large number of units, are becoming more and more complicated.

In short, the process of creating pan-European business structures is today at the root of many specific cross-border tax problems that did not have the same importance when the Ruding report was produced. These problems are considered in detail in part III.

The achievement of Economic and Monetary Union

The above trends are reinforced by the introduction of Economic and Monetary Union. For the euro-zone countries the question of tax competition is even more important now that monetary and exchange rate policy are no longer nationally available policy tools.

22

Moreover, the transparency achieved by the single currency intrinsically generates a tendency of price convergence (certainly for tangible goods) within the euro-zone. Consequently, it becomes logical for multinational enterprises to set intra-group transfer prices EU-wide as a single harmonised price in euro per product or product group, regardless from which production facility the goods are purchased.

This concept, which is often referred to as "euro pricing", can already be widely witnessed in business practice. Transfer pricing is a traditional management tool and there is good reason to believe that euro pricing is now also used similarly, its advantage being intra-group disputes about price levels and optimum efficiency in the structure should disappear. In fact this concept treats the various factories in Europe as production lines that happen to be based in different Member States but all belonging to the same single European manufacturing unit. It is evident that such new tendencies impact on many features of international company taxation, notably in the area of transfer pricing.

Economic and Monetary Union thus increases the integration achieved by the Internal Market even further and. It also raises a further question as to which tax problems hamper the completion of an integrated EU capital market. This somewhat separate problem is not specifically mentioned in the mandate for this study. Moreover, unlike most other goods and services markets, the EU capital market currently still suffers from relatively fundamental non-tax barriers. The Commission has recently put forward a number of measures to remove the barriers to the Internal Market for financial services, and good progress is being made here in collaboration with Member States

European Union24. After the adoption of the appropriate legislative acts EU companies and specific other

legal persons governed by the law of Member States will be able, as from 2004, to merge, create a holding company or form a joint subsidiary under the legal form of a European Company. Moreover, any public limited-liability company with a registered office and headquarter within the Community will be able to transform itself into a European Company without going into liquidation, provided it has a subsidiary or a branch in a Member State other than that of its registered office. This supplements the existing possibilities for co-operation between firms established in a number of Member States under the European Economic Interest Grouping (EEIG)

• 25. 

  Proposals for a European Company Statute have been on the Internal Market

agenda since 1970. The agreement thus constitutes an important break-through.

The basic idea of the European Company Statute is to provide companies in the EU with an additional company law option as how to organise their activities at EU level. The existing national systems remain unchanged. As from 2004 EU corporations will thus be able to carry out their business free from the obstacles arising from the disparity and the limited application of national company laws. The registration of the new European Company, its formation and personality are governed by the domestic laws of Member States. The same holds for the subsidiary rules. It is however possible to transfer the registered office to another Member State without winding up the old or having to create a new legal person. The regulation deals with a significant number of legal issues encountered by the European Company. Those issues that are not covered are in principle subject to domestic laws.

As regards taxation, an earlier draft of the European Company Statute regulation26 included provisions on

loss-compensation within the European Company (parent - subsidiary and parent - permanent establishment). However, these provisions were dropped in 1991 in order to facilitate agreement on the statute.

Ultimately, the tax regime applicable to the European Company is that of the Member State in which the parent-company is headquartered and, according to the general tax rules of the Member States, where the related subsidiary or permanent establishment is based. This is coherent inasmuch the whole idea of the European Company Statute is to remove certain company law obstacles but at the same time to keep companies anchored in the legal system of a specific Member State (and not in, say, an EU register or similar). Nevertheless, it is imperative to examine the existing body of EU law on direct taxation in order to identify necessary adaptations and, to consider the necessity of an appropriate EU tax regime. These issues are, among other things, addressed in parts III and IV.

Against this background, this section briefly presents some criteria for the assessment of company tax systems. It then offers some basic considerations of how these criteria can be used in the context of this study for assessing the welfare effects of different effective levels of taxation and persisting tax problems in the Internal Market. They also serve as assessment criteria for the possible solutions to those problemss described in parts III and IV of the study.

General principles for the design of company tax systems

It is common ground between economists and tax experts that an "ideal" company tax system has to be equitable, efficient, simple, transparent, effective and provide certainty. These inter-related general criteria can usefully serve as basis for the analysis of company taxation in the EU to be carried out in this study.

Equity

The requirement of equity has two dimensions. "Vertical equity" refers to the re-distributive feature of a tax system, i.e. to its capacity to operate a distribution of the tax burden among taxpayers according to their contributive capacity ("ability-to-pay-principle"). "Horizontal equity" holds that taxpayers who are in the same economic circumstances should receive an equivalent tax treatment. The concrete perception of these concepts is strongly related to societal values such as solidarity and fairness. Vertical and horizontal equity therefore strongly condition the political acceptability of a tax system. In the context of international company taxation, equity mostly relates to the fair allocation of the tax base between states in which international companies operate.

Inter-country equity traditionally involves three main principles: source-country entitlement, non- discrimination and reciprocity. Under the "principle of source-country entitlement" the source country has the prior right to tax profits earned within its jurisdiction. This principle can be justified for efficiency reasons and it can help to achieve some redistribution of resources across countries, since the proportion of foreign-owned businesses is generally higher in relatively poor countries than in richer ones. It is also

sometimes justified as a quid pro quo for the provision of public infrastructure and services in the source country. The "principle of non discrimination" implies that countries agree, usually on a bilateral basis, not to discriminate against foreign firms and shareholders in their tax laws. This principle is strongly linked to horizontal equity, since the same tax treatment is applied to similar companies independently of nationality considerations. The "principle of reciprocity" can, for instance, be illustrated by the requirement of equality of the rates applied to any withholding tax levied on interest, dividends and royalties by states involved in a tax treaty. Reciprocity applies to any tax arrangement which leads to similar effective tax burdens on

foreign-owned investments. This is particularly relevant when states have strongly differing tax rules and practices.

harmful for the environment. There are also other instances when national governments will try to reduce existing (non-tax) incentives through the use of the tax system. A good company tax system should avoid distortions with regard to location, etc., unless these are deliberately decided (e.g. in economic "free zones" which are designed to boost economic development).

The two main concepts for considering the international economic benefits of efficiency and neutrality are "capital export neutrality" and "capital import neutrality". Both concepts are considered in detail in the analysis of the effective tax rates in part II .

Under "capital export neutrality" a tax system does not affect the decision by any specific company as to in which country to invest. Resident investors in a given country have no incentive to invest at home rather than abroad, or vice versa. The domestic/foreign composition of the investment income does not influence the world-wide tax thereon. Other things being equal, capital mobility would then tend to equalise the required pre-tax rates of return on investment across Member States, thereby eliminating differences in the cost of capital, and thus distortions in the demand for capital in the EU. Capital export neutrality could be achieved if income were taxed only in the investors' country of residence and if there were no discrimination between domestic and foreign-source income in the capital-exporting country. This could be achieved if all countries applied the "world-wide" or "residence" principle, that is, levied taxes on the income accruing to their residents regardless of the source of that income.

A tax system achieves "capital import neutrality" when all investors, both domestic and from foreign countries, investing in any one national economy face the same after-tax rate of return on similar investments. This implies that the cost of capital and the tax rate for any inbound investment must not depend on the home country, that is the country of residence of the investor. In fact, the application of the residence principle can lead to cases where a domestic company investing in a given country is placed at a competitive disadvantage compared to a similar foreign company investing in the same country - because the tax rates applied in their home countries are different. Therefore, in order to avoid distortions of competition and to achieve capital import neutrality, income should be taxed according to the "source" or "territorial" principle. According to this principle, a government should tax all income originating within its jurisdiction at the same rate, regardless of the origin of the beneficiary of the income.

Inefficiencies do not only arise due to different tax treatments of cross-border investments. There may also be distortions in the decisions for the types of investment, as tax treatments applied to the assets used by companies or to the sources of financing of investment may vary considerably within and across countries.

offset as a consequence of an increase in investment and in the international mobility of tax bases, which in turn directly and indirectly generate increased tax revenues.

Simplicity, certainty and transparency

The requirement of a "simple" tax system is relatively straightforward. It implies the minimisation of the costs linked to the operation of the tax system. These costs are "compliance costs" for the taxpayers and "administrative costs" incurred by the administration to enforce the law. Administrative and compliance costs are intrinsic to any tax system: governments have to raise revenues and taxpayers have to comply with tax rules. However, one might wonder which amount of cost is proportionate for meeting these objectives. Generally, these costs are higher for international transactions involving more than one tax administration than for purely domestic operations. For instance, even the mere co-existence of two simple but conflicting principles source or residence taxation in principle creates cases of double taxation or unintentional double exemption that can only be overcome by appropriate usually complex and costly - international agreements. The criterion of simplicity is thus linked to efficiency and effectiveness. Simple tax systems do not only mean relatively low costs; they usually do not provide intentional preferential tax regimes or unintentional tax arbitrage or tax avoidance opportunities. They may, however, also imply a loss

of equity.

The requirement of simplicity also requires that the rules according to which taxes are levied are certain and clear to the taxpayer. Certainty relates to the stability of a tax system and of tax practices in a country. The uncertainty resulting from frequent changes in tax legislation and its interpretation has, as such, a negative or delaying impact on investment decisions. Simplicity and certainty are generally linked to the criterion of transparency of the laws, regulations and administrative procedures of a tax system. Transparency usually supports equity. For instance, it can help to avoid the replacement of direct State aid by tax incentives offered by administrations on a discretionary basis. Moreover, the transparency of a tax system is generally important for ensuring accountability of the policy-makers.

The economic welfare effects of company taxation systems

Broadly, one can say that if a company tax system meets some or all of the above criteria it contributes to more economic welfare. However, for the purpose of the study, it is necessary to clarify how these general criteria work within the EU in the context of the Internal Market and which trade-offs may exist. So far, it has been shown that the study is necessary to deal with two basic company tax issues of relevance in the context of European integration: (i) company tax obstacles to the Internal Market and (ii) differences in effective tax company tax rates or tax competition in general.

different investment opportunities. To what extent taxation has a negative impact on investment decisions, depends on to what extent taxes offset or reinforce other distortions in the economy. There are two different aspects that need to be considered for approaching this question and for evaluating the effects of different levels of taxation. First, the impact of taxation on economic efficiency depends on the economic environment in which the tax is imposed. As pointed out above, taxes are sometimes used to correct or mitigate a market failure and/or offset specific externalities. This increases economic efficiency. On the other hand taxes may be used to compensate for physical or locational disadvantages. This use of taxes often decreases over all economic efficiency, although it may increase local or regional social welfare. The second aspect is what could be called the "transmission mechanism", i.e. to what extent incentives lead to changes in actual behaviour. To the extent that economic agents are not affected by the imposed taxes, there will be no negative effect stemming from the level of taxation.

It is essentially an empirical question to what extent investment decisions are affected by taxes, as predicted by economic theory. Consequently, the associated welfare implications are also to a large extent an empirical question. Despite these difficulties in assessing the precise effect on investment decisions and welfare levels for ordinary citizens, it is still possible to draw valuable conclusions. If there are really significant differences in the effective level of taxation, then one can certainly argue that there would have to be a very complex structure of externalities to justify such variation in tax structure and tax rates. For example, on pure economic grounds it is hard to imagine that debt financing should be favoured above equity financing, since the negative macro effects to the economy from excess debt levels are substantial, particularly in times of financial turbulence.

It is also hard to imagine that it would be desirable that similar investments face markedly different effective levels of taxation purely because of their country location. Even if one cannot on economic grounds rule out the need for different levels of taxation in various countries, given their natural resources and skill levels etc, one must question the size of the differences and their dispersion. To the extent that there are no convincing economic justifications for these variations in levels of taxation across countries, types of investments and modes of financing, it can be concluded that, overall, the tax systems distort investment allocations. Decreasing these distortions would hence enhance economic efficiency and growth in the Union. It would contribute to a better allocation of resources in the Internal Market that is based on real economic factors rather than tax considerations.

Furthermore, as mentioned above, the tax differences also entail large administrative costs and they foster tax planning behaviour with further costs to the business community and society at large. With reduced tax distortions, these efforts would be put to economically better use.

harmful tax practices. Moreover, the European Commission has stepped up its efforts in the control of fiscal State Aids under the EC Treaty.

For the purpose of this study it is not necessary to review the history and results of these initiatives in detail. However, it is most important to note that as a result of these efforts many preferential tax regimes for companies are being changed, abolished or phased out. It follows from this development that the

general features of the company tax systems of Member States will become more important for economic decisions than today. Therefore, differences between Member States in (general) effective company tax rates also become more important in comparison to a situation where the recourse to preferential regimes is possible. This development provides a common thread for the present study. EU enlargement will compound the underlying trends.

The welfare implications of tax competition in general are manifold, as tax competition affects the tax structure, the tax burden and, ultimately, the financing and the provision of public goods. Tax competition may affect to different extents the various existing tax bases, thereby inducing differentiation or approximations of effective tax rates, and the corresponding increase or diminution of a number of tax distortions. At the same time, it may induce a change in the overall tax burden, in the form, for instance, of a downward pressure (a "cap") on the overall tax level. Its effect on welfare will then depend on a number of factors, such as the State expenditures and revenues structures, the overall public finance position of the State, etc.

To sum up, taxation ultimately involves a political choice and a trade-off between some costs in terms of efficiency and other goals, such as redistribution or reduction of market failures and funding of public goods and services, being pursued through taxation. The same applies in the European context.

THE STRUCTURE OF THIS STUDY

Part II of this study is devoted to the detailed analysis of the company tax systems in the European Community. This includes a comparative analysis of the qualitative features of the company tax systems in Member States and the detailed determination and calculation of the effective rates of company taxation in Member States under various scenarios.

PART II:

QUALITATIVE AND QUANTITATIVE ANALYSIS OF COMPANY

TAX SYSTEMS IN THE EU

A. ANALYSIS OF THE COMPANY TAX LAW

1. INTRODUCTION

Why a Qualitative Analysis?

Effective Tax Rates are principally the product of the nominal tax rate and the rules governing the computation of the tax base. In some countries financial accounting standards may influence to a certain extent the rules governing the tax base and specific tax incentives over and above the general computational rules, may also have an impact. Where there are large differences between the nominal and the effective tax rate a comparison of the structural elements which make up the tax base between countries can assist in the identification of the causes.

The Quantitative Analysis in Part II B includes the results of two economic models based on a `forward- looking' concept. One is based on a hypothetical simple manufacturing investment with a well-defined but limited number of computational rules and covers all the Member States. The other is based on a hypothetical model firm in the manufacturing sector and uses more computational rules, but does not cover all the Member States. (Part II B explains the models and the results in detail.) Through sensitivities and simulations the models identify and quantify the effect of certain structural elements of the tax base, including the nominal tax rate, on the effective tax rate.

The Qualitative Analysis complements this to enable a comparison between more of the structural elements. To the extent that the relationship between nominal rates and effective rates is relatively constant (as is illustrated by the quantitative analysis) it should be possible to identify similarities and differences between Member States' approach to company taxation by comparing a number of the major structural features of each Member State's tax legislation.

The ten structural elements were selected as the most important `common' elements. Specific measures such as those identified by the Primarolo Group

28 as providing for a significantly lower effective level of

taxation than those levels which generally apply in a particular Member State and regarded as potentially harmful were excluded as they are the subject of separate initiatives.

What can such a qualitative analysis reveal?

In the tables the basic structural elements of a tax system are identified and for each Member State the approach adopted by that Member State is briefly explained. It is therefore possible to classify the different approaches for each structural element and identify groups of Member States who follow a similar approach, i.e. apply the same or similar treatment to particular elements. Groups or individual Member States who apply a different treatment, outside this `average' can also be identified.

As mentioned above the Quantitative Analysis models necessarily summarise the structural elements of each Member State tax system and this Qualitative Analysis includes a number of important elements which could not be included in the models, such as certain rules governing capital gains, the treatment of losses including carry back and carry forward rules, and domestic consolidation.

In addition to facilitating a comparison of the different treatments by Member States of the basic structural elements the analysis also permits comparisons within the Member States to understand how certain features interact. For example some Member States do not tax certain capital gains nor give relief for capital losses whereas others tax and permit loss relief.

The analysis does not attempt to identify which approach to a specific element is the `best'. Each Member State has established its rules with the aim of constructing a coherent tax system to meet its particular needs. Without quantifying the effects of the specific measures it is not possible to say, for example that one particular method of calculating depreciation is better than another. The Quantitative Analysis, in particular in some of the simulations, is where the effect of particular tax measures can be quantified. However, the Qualitative Analysis does reveal the range and complexity of methods in use for example in depreciation, where the overall aim of each Member State is likely to be broadly similar and might raise the question of why such a diversity is necessary.

The tables do not seek to explain every aspect of each Member State's tax system. As explained above only the basic elements have been included, and within each sub-category only a summary description of the principle rules has been presented. In many instances there are exceptions and slight amendments to the main rules applicable in certain circumstances but unless these were considered material they have been excluded.

A common or harmonised base implies, at the very least, that the basic structural elements of the Member States' existing tax systems be aligned, or in the event of the existing national systems remaining alongside the new system in parallel, that a new tax code be drafted. The tabulation of the basic structural elements of Member States' current systems illustrates the number of features which would need to be harmonised and the identification of groupings of Member States who already have structural similarities.

Findings

Tax Rates

The tax rate is the most visible element determining the effective tax rate in any corporate income tax system. Rates vary considerably from one Member State to another and in addition to the main statutory or `headline' rate details concerning reduced rates, surcharges, minimum rates and special rates are also given where applicable. Two Member States, Germany and Italy, at the 1999 reference date had a `dual rate' system distinguishing between retained and distributed profits. A number of Member States assist small and medium enterprises by having a specific reduced rate for companies whose profits are below a certain threshold or via a system of progressive rates but there is no standard definition. Certain industries are also sometimes subject to rates other than the main statutory rate. Certain types of income, such as capital gains are also subject to different rates.

With the exception of Ireland these variations from the statutory rate, although important to the companies subject to them, are not considered to be material to an overall comparison of EU rates, as the majority of enterprises competing across the EU will be subject to main statutory rate, adjusted in some cases for `temporary' surcharges. Ireland is a special case because although the main rate is 40% there is a special rate of 10% applicable to certain companies including those in the manufacturing sector and for the purposes of this study it is more appropriate to consider the rate of 10% as the `main' rate.

29

The range of statutory rates is substantial. Ireland's rate of 10% is the lowest followed by a group of Member States with rates around 30% (Sweden 28%, Finland 29%, Luxembourg 30%, UK 30%, Denmark 32%). At the upper end there is a group with rates around 40% (Belgium 39%, Italy 37% + 4.25%, Greece 40%). Germany's rate was in this grouping but as from 2001 the new statutory rate is 25% plus `Trade Tax' typically at 12 to13%. The extent to which these differences in statutory rates are reflected in effective rates is covered in the Quantitative Analysis in Part II B.

Effects and the results of the Quantitative Analysis in Part IV C in relation to the Comprehensive Approaches. A more detailed comparison would require not only the tax treatment of particular transactions in all the Member States to be compared, but also the normal accounting treatment of the same transactions to be compared.

Depreciation

The significance of the rules on tax depreciation as regards the effective tax rate varies depending on the capital asset intensity of particular sectors of activity. When the rate of depreciation permitted for tax purposes exceeds the true economic rate of depreciation there is effectively an tax incentive for investment.

In many Member States this appears to be the case.

The initial depreciable base for capital assets is uniform across the EU in that it essentially equates to the cost. The only exception is Greece where an element of revaluation is permitted in certain cases. The most common methods in use are the reducing balance and straight line or a combination of the two. There are also sundry other methods in certain circumstances. However, it is in the rates where the greatest variations are seen with some Member States also providing a number of rates depending on the type of asset involved.

Given the range of methods, rates and asset categories it is difficult to compare the depreciation rules without carrying out a series of computations. When choice is also involved any comparison becomes even more complex. In contrast some Member States

30 have a relatively simple system, which when combined

with group relief and an unlimited loss carry forward can give the same amount of flexibility as the more complex approach where the amount and timing of depreciation `claims' is an important factor.

It is more straightforward to compare which assets are not eligible for depreciation. Member States take quite different approaches to the depreciation of intangibles. Concerning patents and trademarks a number of different rates and methods are applied but the biggest contrast concerns the treatment of goodwill. Twelve Member States permit some form of depreciation but four Member States (France, Ireland, Portugal and UK ) do not permit any depreciation. In principle land is treated as a non depreciating asset across the EU. Some Member States do however permit in certain circumstances a deductible provision for a permanent loss in value of some non-depreciable assets. The rules for the accounting for tax depreciation in general follow the overall rules governing financial and tax accounting in the sense that in most Member States depreciation is only deductible to the extent that it is provided for in the financial accounts, whereas in Ireland, the Netherlands and UK this is not a requirement.

some sectors there are special rules reflecting the relative importance of debts to the business activity, such as in banking and other finance activities.

As regards provisions for future expenses in the most part these are deductible although in several Member States provisions for repairs are not, possibly to avoid this category being used as a type of tax free reserve.

The major differences appear in the category of Pension reserves which could be considered to be in the second category of provisions as outlined above, a type of tax free reserve. However, this particular category of reserve is heavily dependent on the method by which pensions in general are provided for in individual Member States.

One approach would be the example of the UK where provisions for pensions could generally be described as non deductible, but where the general system is that a company does not have a liability to pay pensions, but agrees to make contributions to a separate pension fund which will eventually pay the pensions. In this case the UK company would have no reason to make a provision for pensions, it would be required to make a payment to the pension fund to the extent required by the fund, and the payment would be deductible. An example of the contrasting approach would be the situation in Germany where the liability remains with the employing company and the pension fund assets are not separated from the company's. Companies have the option to fund pensions by provisions (discounted at 6%) to what could be considered a tax free reserve. However, given the summarised nature of the tables it is difficult to comment in any detail on the detailed implications in each Member State.

The main effect of these different approaches to pension provisions is that in some Member States31

companies have the opportunity to obtain a tax deduction on the basis of a provision and can therefore retain the cash within the company, in others deductions are received only when the cash has been paid out. Whereas the actual tax situation of two similarly prudent companies who provide for the future pensions of their employees might be similar, one company might be able to retain the cash in the business, the other

not.

Losses

The possibility of carrying losses forwards for relief against future profits is particularly important for new business start ups if they are unable to utilise the losses before they time expire. Seven Member States still retain time limits and these vary between 10 years (2 Member States) and 5 years (4 Member States). The possibility of carrying losses backwards against previous profits also varies across the EU with nine Member States not permitting carry back at all and the remainder permitting it for between 1 year (3 Member States) and 3 years (2 Member States).

Loss carry forward provisions also illustrate the parallelism between structural elements and specific tax incentives. The latter are not considered here but it is interesting to note that an unlimited loss carry forward (structural) could be considered equivalent to an extended tax holiday to new businesses (incentive), although in fact the structural element could prove more beneficial in cases where losses are incurred during a tax holiday but time expire before they can be offset against subsequent profits.

Capital Gains

The treatment of capital gains can be divided into two main areas: tangible assets and intangible assets such as shares. In general the rules relating to tangible assets have moved closer together. All Member States observe the realisation principle; although in certain circumstances some Member States permit unrealised permanent losses in value to be recognised. With the exception of Ireland, France and Greece the applicable rate

32 of tax is the same as for trading income. Only Denmark, France and Italy have no

provision for rollover relief and only Ireland (related to the separate rate of tax) and the UK (possibly for historic reasons) treat capital losses differently from trading losses. Only Denmark, Spain (for immovable property), Ireland and the UK make allowances for inflation when computing capital gains. Overall the treatment of capital gains and losses is broadly similar between Member States and the rate is largely determined by the normal statutory rate.

With respect to intangibles, specifically shares, there is greater contrast in treatment. A growing number of Member States exempt gains (and losses) on the sale of shares, or apply a reduced rate of tax to gains (and losses) whereas some tax (and relieve) at the normal rate.

Mergers and Acquisitions

With respect to mergers and acquisitions the situation is relatively similar across the EU. All Member States permit, under certain conditions, some sort of deferral of gains on mergers. The deferral can be achieved by two means: a full deferral until subsequent realisation (similar in effect to rollover relief), or a deferral by means of an instalment plan for tax payments due on any gains arising as a result of the merger. The deferral also provides for the transfer of the existing tax base of the `old' company although there are generally strict rules concerning the transfer of losses.

This area of taxation is very technical and other than identifying the broad similarities it is not possible to provide a detailed comparison between Member States on the basis of the summarised information presented in the tables. However, it is worth mentioning that some Member States have extended the deferral rules provided for in the Merger Directive

spread, and corporate structure of an international enterprise its overall effective tax rate can be significantly effected by the range of rules across the EU.

Inter Company Dividends

The main distinction between Member States concerns whether they operate a credit system or an exemption system. In general the exemption system is considered more advantageous to the tax payer. The `generosity' of a tax system can also be measured by considering to what extent the relief granted in a Member State is more generous than the minimum required by the `Parent Subsidiary' Directive

• 34. 

  The

majority of Member States (Belgium, Denmark, Finland, France, Germany, Ireland, Luxembourg, The Netherlands, and the UK) are more generous than required by the Directive and make no distinction between dividends from Member States, and from outside the EU. Spain provides more generous relief than required under the Directive, but makes a distinction between EU and non EU dividends. Austria provides the minimum relief in accordance with the Directive but extends this to non EU dividends and finally Greece, Italy, Portugal and Sweden apply only the Directive.

This illustrates that the Directive is essentially operating as a minimum standard, which a majority of Member States are prepared to go beyond. To the extent that the Directive sought to provide for the same treatment across the EU it has not achieved this aim.

Inventories

In periods of low inflation the rules relating to stock valuation are unlikely to have a decisive impact on effective tax rates. However, even when inflation is low individual sectors may be subject to large variations in stock valuations when particular commodity prices vary, such as crude oil which over very short periods has fluctuated from over US $30 per barrel, down to US $10 and back to US $ 30. The distinction between Member States who permit the LIFO method of valuation, and those who do not can therefore be an important factor. The majority of Member States do accept this basis for stock valuation but Finland, France, Sweden and the UK do not.

In certain industries the extent to which certain administrative and overhead expenses are included in stock valuations may have a impact on the final stock valuation and hence the taxable profit and effective tax rate for a given period but the summarised tables to do not permit a detailed comparison between policies in different Member States. Such considerations would only be applicable in certain industries.

Expenses

The basic rules concerning expenses are broadly similar. Although differences tend to be permanent rather than simply timing, and therefore potentially more significant, where there are differences they tend to be very specific and in most cases concern relatively low value items such as entertaining and travelling. However, without going into a great deal of detail it is not possible to make a general distinction between Member States.

The treatment of interest, specifically thin capitalisation is an exception. Several Member States do not have formal rules on thin capitalisation and this could lead to substantially different effects on the effective tax rates in Member States when comparing those who do have protective rules and those who do not.

Conclusions

On the basis of the tables it is not possible to draw general conclusions concerning the quantitative impact of the variations in the structural elements. However, it is clear that there are substantial qualitative differences in certain areas. Within each category one can identify `clusters' of Member States who approach specific elements in a similar manner. However, it is difficult to identify a group of Member States who consistently, across all the structural elements, form a coherent grouping on the basis of their current tax systems or a group or individual Member State who are consistently outside the `norms'. In a number of cases the changes which would be required to bring Member States closer together would not appear to be major and in a number of the categories one could question the `need' for the detailed differences.

It has been suggested that the tax treatment of structural elements of the tax base is compensatory and in terms of the overall effective tax rate a high nominal tax rate is indicative of a `narrow' tax base, and a low nominal tax rate is indicative of a `broad' tax base. Analysing the individual structural elements reveals that there are differences and these may themselves be compensatory, for example within a Member State restrictive rules on the depreciation of goodwill might be `compensated' for by generous tax depreciation of tangible assets. `Compensation' could therefore exist at two levels at the rate/base level, and within the different elements of the calculation of the base. The qualitative analysis helps in identifying the potential compensatory factors, and in identifying the different clusters of Member States. For example with respect to the statutory rate `clusters' around 30% and around 40% can be identified and at the `extremes' the tables show that the highest rate is currently four times the lowest rate. For a single market this is a wide range. However, without measuring the relative financial effect of each of the structural elements any statement concerning the degree to which one measure compensated for another would be subjective. It is for this reason that the Quantitative Analysis is necessary in order to place objective values on the structural elements and test the hypothesis that the treatment of the existing structural elements is compensatory and to quantify the respective impact of these on the effective tax rate.

s

tf

' s

rf

mp anie

s s on a nep any end o

d omc h )

on

onänkte

on cox ce

co

s c

hr and theg lei

v e r

in em pute

f the

r annumten

r annumit Ber annum con os ver

pen og i nning

l l s chaf pe pe

tur

U ROm e isiso

bem ögen

U RO

U ROten m

U ROm p ar

rian g e see inco at the ye ars v er

a bls e

ts taxetr i eb

Au st

34 %n.a.n.a.3 .750 EA k tie

1.875 EG e s e l l s chaf

H a f t ung

5.625 Ew i

th annual

37.5000 Etaxw o r t

h co

asthe( B

ee d

w

t h osl o w

n thea

x lae al

d o

of t follow

less e s ar

m puteely

con t rul osuntspo

l,

c pur

m e ise ra

ss accoo r tax

h

e releva

dn s f

e inco busine

a blductio

Finlan

29 %n.a.n.a.n .a.n.a.taxb a s i s

of tw h i c h , i n gen

o f thede

s s

"s

ine

n the

d oacticeo v i de

l y pr

s o und bus

m putes s pr

icalf "

coine

e o

m e

is sp e c if

s o und bus l a wincipl

e incof "i s ei

c pr

a bli s oe s s tax

h e rw bas

fo rss t

n co

o v i de

e binding pr

l a wo ductio

ns arr pr

e s s then o

i tio

va l u atio

s e s unl

poi s e

m me r c ial

purt o r

ic acquis

cotaxoth e rw

his

s t

unts.n co

acco

o ductio

r pr

n o

r d e d

in thei tio

e co

h e n r

y wt o r

ic acquis

o nlhis

s t

n co

37

o ductio

r prlue)

ncen o

udei tioi d u a

l va

res

"

is pr

e low

t o r ic acquisacticeot b

s

a tee s

n reet

eens

:

up to, ar

)

• : 

  2,5 %e r agn tabl

e d

in the

av

na

ve bpre c iatio

m p aniechas ye ar

annualr d e d in

e

n

t

:

t e s h deu r a ncepre c iatio

), purf the re cor y and canno

pre c iatio ras s co

i dualiner insquipma

n des e

ts fi nancial

des

• : 

  2%u s t bea ndato

a

te m

i nef busG e r me f e r e ncef the

st a n d a rd

d, indive db ank oe

at 50 % on

rn munts

n isd

ne

riaht lghs os (d ingr y and ee r%i x e

d asf o

l o

w l i f e (s s i ble f

ductioss acco

t r a igv e l o ped ingr buil

e f ul pos

• : 

  12,5 nd hal

pre c iabl

pre c iatio

depre c iatios t po

po

Au st

• - 

  sa l t h oudeare ald ing

buil4 %builo t hem achine

usascarm o v a bl

s e codedetaxbusinedebe

cesn

o ngh eo u sno r

s e s ,.r a

ne s o ura l u e isfee

a

rir

e

:si .en f

n vi c lia y bpre c iatio

fe of tr ve

t e

s ad inge r thanpo

n ( e s and la l r

itioom dei c li

e

s fot r

a

tivtrial

• : 

  20%e

n

t:

25 %e a r s

r d e d inpre c iatio

o

w purr d e d unden take

so be

pre c iatioiblont e

e a r s momi nisity

deo r natur

acquiso s e anyon re co de

n and builquipmf theses is lo r taxpre c iatio

re co

e c iatios e s

poe

d f

a ncee sd f

t h r

ee ye e n 0% and ae p r e c i a t i o n rau s t

be) de be

toe pr

ses can alpo

fo r intang

e thoh i c h t h

e eca b

le ecatu t o r y rat

and adm and indus su c c e s s i

ve y

tructio purl o w

bali nen munts. I al (t ax

)

is. Dpo pur

nditurn mh ich theeeda y chotw

e p r e c i a t i o n ra

e ber obp t sts e tsm d

ntialnsarch activ

r y and e

n

al pur

dht le de r

mm ds

• : 

  4%e

r

c

ial

s

• : 

  7%e se

lancliningh

e px cef asx i mu

t r a

ig ex pe

ple t

io of w

a

ll or of w no t exc

nsp ay

r c e n tagx i mut , ee s i ded ingm m

d ingo r

r: over 5

ductioss accountingadingo r tax

h ich isf e r e n ce he

d fachine

i c l es de additio difunting

f e r r e

d f

Fin

-

de- ster m

• - 

  deass e ts

s md o ese x pea taxpemaover tassekind ot h e maf o r rbuilf o r cobuilf o r l i g h

t cousef o r mvehtaxbusinef o r accothat wthethes p e c ialaccode

andfs

t ' sd

s e t'

na nceitte

l i f e o as 20 %

of be

balf assea r i e t y inr mm e n t a lly

s orone f ulf the 33 % r d e d inN to

38

f e r r al

pre c iatio

a sir e a

t vn is pen vi

on usi s

o 3 % 4 %

e n

t 10 % - re co

deO P E

des )cliningn

i

th g. g . e)

bass

:

2 % -

r

s

:

25 % -e

d ( a nced ingf depre c iatio

n t h e

b we ntsndingu s t be

s e tspre c iatio

t s (e

pen ther

a

l

:s

• : 

  1,5 % -quipm

d ing

i ne bal builn o dessen munts, nol o w

e )d dee s t

mn al

ht li ne

f o r as usa i n

ad de n oe ne

d ing builm pute

r y and e co

a tes e

ts builtrial:

30 %ductioss acco

t r a igclining

m b inatioht l

e p r e c i a t i o n oe (l e r a tee rty inv

agndlcul aspre c iatio

r s o n al depre c iatio

rifie d ! )

achineucks

,

en

ible ra t

es-

e a r s i

nw n to

a r k

et t h e

t i mr ite

5 yh ortn d or

pre c iatio

e s lan do

a i r me lowp e is' es

r i e n cea b l

em me nt

e

cos arn at thee

ct w

o r intang dea y bt l y

b l e g a l and scollow

e

d fu g h in thelue ofr itte

wf t h

e f

a n en bei dea y e rs

acts andx pen toaim

a dir

l o wunts thew i

ll m

a y belue i

e toe fs s er e au l cla l u atio

s at r a tio

ali a t i o n over 1tho

,

al m vae r mctivineubtfr n by

n ise p r

ec acco goode ps t .

t havo unt, a w t o t a

x p

b jes i oni nis

o untse p t i o n t h

e vas e ts

asa r k etn co novenn bus

a s gon gin o

l y or ovid mo r dou g h

a v

Atho- c o n

ce co mmon

rias dollyd o

pre c iatiot

sw i

ll: d

am

m me r c ial

c q u i redn exca i r mitio

lue he raciali f i c pnali s

e)

e d, alo i ng

ecatioo v i s i o n s fs more

Au st

deassegoode qualcofor aA s a

f i nancialt h e fvaacquisclaimr t ain in am

cegenif base

e s peSp( Noth e rw

prall o we r g

l o woff i

t

e d no

t h ex cee s ,

a y bes e tsa nds ,.fe

r a l l y

antea bla s

e of

i thh ent e

the and the t h i s orf f

n l y beu ard inga b l

e ie nee s .l o w

n we a r se s noa l u e mi x e

d asc e ofo s e d o

0 ye don vf fs e s iftantial

ar, such as lo r

g or in

o r builllowith gt al

incipl

f 1e p t t h a t

ws can opod

en di sp

pt fr

e a w

pre no

f usi tio

pre c iatio

o d oa l u e

o pur subs

evi

x ceh ippingtry f

s a

e p t i o n s i n t h

e c

,

exc and tei tie

t s are

dee rie es i onr d ance

untingr s .

acquise arcur

o r taxa l u e isu c es

wn indusr ovio v i s i o n s arr e excade

di nem pd emr i o d o

f f

f pr

od asse

ch

ht le pe

s , thee dcre a s e

in v

ductiblc r a f t and se a tr

x i muw i

ll i

e arns

dec

t tor t ain sen o

tructio

i f i c pe in accod acco

cre a s e

in va y er

nsy arpter a l pr

Finlan

s t r a igmagoodpro b abl

3 ye x pethesubjeand cew r itte

det a xpw h e n su

N o

t de

f o r aircoS p ectheacceG e ne

b u t t h erf i nancial

ffr d

n og e d

feln o

serve

readm p anysk

char riy acco

w i

ll i

t y ) ari th alo v i s i o n

s. '

, t h e ree sp co

eiumme

39

o bi m the

e m

d pare wpre c iatio

r a l practice

p u l s o rysts t

of te by

d i n g goodd s , de

a thirinciploda n y for sae ned that the

s s pr

n c lue thoo m

e

s

e

r

v

e

:

in anticipatio gr o s s pr

o mpo v i deine

(ir o

mn mo t c

n

r ex p e n d i t u ren t co

rre pe riductibl

d fe ,

in prs nt u recuic and ge pr

a bl

i z a tio annual

p e c if

chasew i

ll i

pre c iabl

pre c iatioa i n ful e

s rely over a

e s e r v e de toran c e c

l o ws o und bus

th s

a nde ali th `

2 )3 )

d , (o r

ar

t u re1 ) in

e (ion, (n f

for funs pe n s i o n

m pute

ductiblr an

e pet u re

l y coe ductio of 6% pe

serve

e dei al a ra te

re arn undeus

ol

d ag

a r e

of fur e s t r

m e n ts

b l i g a tio pay

t a b l e

sh 20 % pl

n

s as actuar

b u t i o n s t o a

ion payf

an oe n

t tod by

nse or e e mh e ra

c o unt at an inte

C o n t ri

pecasagfor to b l i g a tior e ducedisannum

t

e noe l y

f f e ctiv

t u reions, ar

s t s e

funs

co

a b

le the

r ob

m plo y e e pe

e until

s for p of e

sionductibl

m e n ts

de

p r ovipaytaxo ccur

c ted

(1 )

ionm e n tsion

ns7n d

nss ubje thet u ree ifo v i dec t u a lly

pee (e s e r v e

ar a pe prh a n 4e nt

ion pay

o n t r a c t a40

ided byr e

is an

pay for fu

ductible n t s a

m e ise r te m

e tir

u tsion rns

ee c

ductiblns pe then toserve

e dec he

an o a pe thea intaine re are e s olda r l y rp l

oy pa y m

toe

nt s

r emt u re

d

nsr a l l y de toe ifr m

l f and ifb l i g a tio

e nen s )ns

e

d om e n ts

e

mm p l oy

u

t an ecte

e tirh om fu

i butioe

g itsee o

ex pe

ntrnditioi butioductibl

a nagb u t i o n s t o a

ntrm p any

nsion pay

a r l y

• r) 

  for ei tho

(2s we i n t h ei

coe mr e v o cablo n t ri) for w

f

e oe ast 7nte n t

tims si thin 1n-fivate

we n av e r

ine bee s i deishm pr

s e

ts fo r

at l

o l l o

e

at thes tabl in a no sa l e os by

ea bly busn

i l a r aso l d hasm p anyh ich re d

in a rn

t ep aniept

co n d i t i o n of

i m us

s e

t s coo r wane and them

x e m

ee tax

n

in time or dinar

eler r e ctio

n coef on asr me h o l ding

f thes e

t fe d, ish arp any in coe e

ductibla bln, asi a

l ru ra te

e nt in sns ar

n: thet o asaim

cl or pe

f

a sm

ns

riaitatioails s i

bl ga ins ar

izatio

m e

p ecl atioe s t mnt co

sp e c ial

a rnditios , the

m p any

sa l e o

ticipatio

Au st

N o

n de

n o l i

mN o t av

N o t po

capitalr e alincoNo sn o infnoroll over relir e invyecof i x e

d assey e arr e l i e f iscother e s i deparf o undatio

ds

e arfm e

e

h e)

o v i der 6

t ore ori s esh e re nt

t a

x fret i t l

e tr eme s t m

i thin 2 ya l

of unde

r

e lose l i e f ' timo r

s s inco

ine unde

e s e r v e , pru s i n e s ss ps for ot

e inv

d, wi s p osl o wns

e

at thee s e r v e fl o w

f r

aces s e t s ae e beptioG r o

up r

a bly busn

for b (w h i c h en

u s i n es

m e

nt rm e

nt re e be

a

ge or losn o

a s

e of dx e

ms

ace re pl

r

8a `

n cise s , s; ee tax

or dinarace

ise s ; s

r r e c tioserve

h a r esh e ba m

e pls e

t ise d i f t h

e ae m ye ar

e

unde re ple ms '

n co ra telie f i n t h e form of as

e tnditios

n t ren coe ar

d as or ia bl

s s

pr ga ins'

l o

w ga ins arn, asn:

s s prp t i onl atior re an d s

t o ua s

e of do y

lan fr e e ri s es

ly allow

a gedine 10 taxail

apitalisatioptio

ine sp e c iala y er, upo

Fin

a taxthat theond a mbus` Cup toN o t av

s e e becapitalr e ale x cebus`exemn o infnoroll overep l a c e me

p r emt a xpa n d i n cass e ts

w i thin tw

es

be of

d

l o ngi ble are of

s .

t angr toi m

s s bee arm e nt

re duce

s e

t (e d

o of

4 y

u mpt in caset t h

e tinei n s on to isace41

e s e r v e , as asx ces e ts

re pairs e tse pl

ens eb l

e ay busnpt r

for gae asi thin 4 y

m e

nt r bea x i m

a xa

toibld ww as

aced that thei ctio

)r

e t or dinarr r e ctioserve

acen

ne ex ce

e plndee ) isithin a mn

in tims

ibl re str

s

ee 7i n s

an co ra telie f i n t h e form of ae pl

n t rer intangr rn of tax

d, we arn, as

c ials ( fr e e

r intei tatio

ace 3 yl atior ree o

o unt o

l i mizatio

m e

inf sp e c ialible d oe c iatio

am it isp too spee r g era p i t a

l gap l a c e mee prserve

c t

ubje

sh a r ess , s

ofn

mp anie

o r t r a n s f ernt coticipatio

f par

p t i o n fn-r e s i de

n s o

nditiop t i o n

l d

in no

T a x exem

heto co

exem

s

lat i on

e gu

i a

l r

p ec

n o s

42

t r a n s f

er of

i n s on

l d ing

ho

p t i o n for ga

tantial

x em

df f

h e

n i n ad a r

ofi te

y wi zeda l eni mt o

o nlo s al sequal

o p e ri n

g ci sp or

a l of

e

d (n l y bee v e

n e

y l o s s pr re c ogn

ed

h e

d ye ar

aim toh enr ec

clm tm p anya y o

, mm e

in s

in the

or dinarr d ingn d wo r p

t fd froc

k co

e asrre

n can bed accoi n

g aenm p anyr incoting

t arn

n cu cot he

riak eeps s e s s

ms i

ductiblductiom pute

a r s )e s

in a stoityt o s

nsb l e

ainsr t iop o s itiossis

Au st

dedecob ookt a

x ayel o ssesharl i abilagpodisPoYe

e

se

ar

sh

ar he ld

r g e r se ,,h ee asl e

e arn ru

the or

as of the

eent

sm p anyf e r e elet e

ly of t sa m

s e tsr s hav

ga insi th Me

trans

l o s s

y i n t h e

o mp0 % ther ant ae n e ral

a y , upo

e r , g

y l o s sy l o s s

if ha

ve b wm e s tic

doe r o r cod that the

f thee h o l deh a r ess mh i s g

fi x e

d as

t capitalh a r esr d anceo v i dei s i o n s isp ays t

transfs

o of the

t h

e sa r d i s

c t h a n 5

r itie tax

g a rd

or dinar

or dinar

l d asainsfin a n c i a l asse

uding off div

n t s/

e ase ase agf theard, prl d ing

g i nningm p any forwr s

autho the

de n heme incl

e

in accoe

-ns

s oo r wd ho and its shar be 50 %

a rryh e n morel d

e n

t orgen by

i o n

as re

t orya blede so

ductiblductible be

ductible t y p e i f t h

e sn v e

st l o sser i e d

f the

m b ine

m p anyeedr b e d coi t e d we s ar

tre a tm

w e v e r , taxicatioe n sat

Finlan

dedehavi n vendes a mas iA v ail

D i r e ctiv

transactio

thecarcocos i nceexcabsot h e loss cforfeshartaxt h a t for mehoappld i sp

to ry

ae

e nor g e r si f y fd bi th

e s u l t ingrre sa mi ng

i t u t i o

nn wter is

af

f ae s ari th Met qualn cus itsf its the hav

43

wm e s t ic

nos il d bym p anyn t i n st

y l o s

sn oh ar

domem p anyt atus

izatio of s

o r l o s s e s rc o n tinuet 70 % o hem p e n s a tio

pt fr d ancee as be a co

or dinar

re all d ingn.uding l o s s e s doi

o

n

• : 

  l o sse

t at l tod by i n v e sto r

co that cof e r e

nt s

e as hox ce

incli ng

n thee , ee in accoe

-nse n sath ich disf ani f y f by

d

a dif

w no

ntinuea de

tantial l i quidatioa blo l l o wmps s if

h o l d e rsu s t aine

atu s o

t qual

ductiblm p anyinee s co

are no ob taine

e ductiblo s s e s ov aili r e ctivansactioh e fo f its m

e

s s i bl

N o

t po

d

e n s e

o v i des e s thea ining

prn tailpoe m

nse ei r r

• n) 

  pur

actioa s e i n t h

e s

i r e ctiv

ansa

x b

n tre r r e

d at the

r g e r Dh e t

a tio Me (d e p r e ciatioansf

e

tr va l u e

o r tax

o r g a nists ara nce

Ref o r

in thet r a n s f

er of of tthat fassebal

e

m p any

e t a

x free

d co44

r t o

b sh o u l d take

e

rge acquirlu

m p any

f the va

h e me cots oook

i ng

e

r for t assee b

a m

acquir

o s se ast

l e d

f

at lt i o n i n

i n a t i o ny cal

o f

it and l) o

e nt

b

o

rdidiartr ag

n d t a

xa par

ubs2

)

a prver e a

to l l i

ng

i

p

of suungs

nshn c o mntr

e

nt and s) , and (en

ll i

of

co

lat

i

onchaft g reemabführ

ria

nso l idatio ree e n parn

g asn

g of a

) atwr gans r gebnis e

ar hands

Au st

Co(1be( O p ooli(

E 5 yPoolithe

e tax

f

90 %e ntsu r a nce and

n oe mk e

, ins

y mam p any

r o u pe quirb e r s that arp a n y

ticipatios s rns

e mo r coo m

n, but ginei nancial

e parf ft itutior o u p mar m pay

a y

ee c

ns

sh are busn o

ptio

do l idatioion insf t h e

g ot

he the

toe by

u mo t h

e p

lannsi butiof f e ctivx cense r

s o

ntrmbm e n ts

ductiblb l

e t

Fin

n o cocom i nim

and ew i

th e

and pemepaydet a xa

ree

m e

) onm e

f thee ntr e

o p e r ative re g i m" inn and

ns inco

par fu l li o n sp a

m p anys i tioo r e ig

r e e l

y inco, ...

45

) , coante

t 99% oe tax

nditiopo

fo r fc t tonce

sactoui thin the

ic co gu ar

e

as sa m imr p o r ated in f

of the unit: can be

d co coe ntse s i de

i def r t h

e grs e tsa l u e w

publm p anyi scall t r an

u tualf

at l thed the

f thee m

dive s , subjey o

hands

nn e fs, als ofo k v

b er

V ' s (d co

itec

t tountr

"s tandarn o

re quire shar

n

t Nl ims and msr s hip o

ubje co

n

al sa f e g

u

ar

n

als

• : 

  capitaln in the,

as of l

o

ssee

md: fi x e

d as

r

o mde

' s (t iene

ciem p anie

owe s sr tol l e ctioe r abl

co in them p anysfeeen

r e g arf e r r e

d at bop

o n s o l i datioe s ideansfa x a tioans

n s

s + of 25

nds sh a r es

s

nditionds

i de

m p anie of EU

coi de

se h o l dingi n s on

o r dive n ts

i a, noo r divh arm p

anie

n

• : 

  snt co

a n i e sn fishm

m e s

tic cou strm p anien f

a p i t a l ga

o mpptio

dos tablptio

n t es in Ant conditio of c

n-r e s i de

sti c

c ex e m

ex e m

meive d byane

r e s i dee d , cot i on

r

mm p anie

n-iv

l d

in no

Dof u l l taxr e cepeconof u l l taxr e ce%n o t a xahe

i th

h en

e but wm p any

r

25 % in

a blm e tax

p t , w

ent co

w e r o

e

tax incos

e ctiv

ar

e exeme s i de po

a n i e sive d arr p o r atem p anie

y dir and ro t ing

o

mp re ceo r cont coidiare c e i v ede a ty

c k

sti c cnds

e d

it fs ubs a tr sto

i den-r e s i de

e

nt-d e n d s rr e

is 10 % in v

ns

d o medivf u l l

cr nopard i vitheo wcapital

e s)

n ife nties ,s

:t s ,2 5 %

s in

ptioe capital shar

e ntsu s t ber ofie s t m

n applm p anie

46

x e mh ar

f sc k m po r t

h in the

rec t i v e (>

unt and ifs

:e mptio

e h o l dingnt co

sn tity

r e n

t stoe quirr t f o l i o

inv se

t f

Di

x e m

nt ea l t a

x onn e sh ar

e 100 % er > 5% om p anier rn s as

d ton-r e s i dee

mp anies oint acco

u r t hea t e i s ,n-po

r e s i

de na t i onb s i d i a ry

ductibll d

as curnt co

equal in jon- t h

e r a nonditio

t icipatioe l a teductibl

e s

tic cot her e s i de, but 2 f not t o al d as co

ren t

-Su

pare s r

nsnt and not de

l d ingqual

o me noo n-h a t everC Pa

e no

u s tem e

e d

y arr

aimats,

n paid

c e

rn fo r

one t h o d

m thee inco, o

cl bos ,

e ifa bli ng

e )r tingitiense s ,

r a tio orm p any

co

n g

-ce re t h ode curmn l y be

o h ibit thex peunes ory

s, spo facil

e mf the

s lowe l e c t e d myo s t mductibli

th tax, s

d wi nge inco can ol y prn e

a g

e cse dectea

blnse

• : 

  car of ru p e rvi

icalh

e sard o

co s t

or goi

, t h e sistentl

nse s arnne tax

h i c h e ver iIFO av er

nsduction ablfe s e n tatio fo r m

of te bo

ria, Le d coe asoachts, huntingp e c if

s, ...n oe pr

f rf anyb ersa tiv

lue w

FO usx pey cor e

d in acquir

e ctlr t ain de ru l e s s

ductioe mi nistr

Au st

A c q u i s i t i on

vaFIbew e i g h t ed

fun g i b

le good

all e

dir( i ncurm a intaining

cew h e n r

l u x u r y yantiquetaxde50% o50 % ot o madm

st,ffr

n costs andier s oe o

a l u e

at ye arf the

l cos e s o

pom e ,i

th a

t

c.e s

e

s v as s e t s i s

fo r

o unt oe s to

t oryneo r al purn w suppl e

a i d t ons

t salunting

amns

dor n i n

g ornee f undabl

e

d fa intainingpt incoctioe rx pe

e

ct acquisitior nex pe

accon aln vene isr mm e

ses

l o wo r the

, o i n eannel e phos e ari o n p

e n

t e

f dirst o sa mpo

e alr e d fi ngs s inco- e x e mg e s paid to, te

e

st o of the

o p o r tio co s t

of i

the purine tax in coicityn thee n sat

r t ainm

n

s are cur

de s incur,

s i n c u r r edisingw he

e ) ,mp

s ,n char

l e c trs (n te

lan l o

wm e nt co

a pr ov e r h e a

d e

acee d, if

l a st dayuntingnse bus

a ble n s ess art iectioh o l d e rs,f e

rge r ,r , e

nalnnei tiee r ablv e co

ssi

are

Fin

at ther e pltheaddinge s s e ntiala c q u i s i t i onall o w

accoFIFO

D e ductio

e x pee a r n ingtaxexpm a intaining

l o ssemepecow a te

f aciltransf

e x cesh50 % o

doe

ss

nse al

h i c h e verm e nt

cte

wt o c kacelnne

e pl als s and ar

e n

x pes, m

47

lue,

udeine "s o und

e n s ess )

r such as.o t he

et vai n i t i a l su t ro s e l y co

tor t ain e

itho

a rkt h ode s inclr clf a bus exp. fi ne

e .gf ceactef , cl

nfo r m"

s t ofe ( t h ert mn d t h e "r wnsy o

,rk mey coo o d stuf

oc, ax pee ctl

actice 75 % o

i x e

d char

o stFOw ith o

e enduct o

the

a t i o n eit co s t

or ae r , o

, FI " ( )e s dir co

nss s pra u s t i v

e liductibl

a mi f t s, f

ductiblr s ifinee r dei ng

r t ain ga

lu) a)at c

l o w)ba s e st

IFOserve

i

th then exhn l y up to

a s

u l e s ,n

n ro u r t ha s

t w

en

i z a tiot i

ve Ce

d upo

s bas

i n i s t ral iner r a n g em

d mn a

u ide

ic thin capital t h

e Ad g

ghisher i a t e"

h e r t h

e loa

sp e c if

p p r op

noa l t h oue s tablw h et

"a

a y

as

ed

thinr e s t paidnt m

aeem

ra tio

of inten-r e s i de

i s de nt.

ity l o

an has i .

e.e stm

e quityi

th a no inv

bt-u l e s ,

n rductibiln wh e n theh a r a c t er

icit de capital

izatiod wt c

niea n en

ex pl

w e v e r , de

dequity

nocapitalhoo n l o

an takebep e r man e

r

-n

r t aine r sio

of inte

. fo r

nv

i tye .g co

s (48

u l e s, but censei cial

n rductibiltif

dex pe ar.

izatiobt)

n ther e

st ed to

de

n s

o intee l a te

s

r into

nse

thin capitali tatiom p anyquity

om e

u i t ye r

twf the l o wg e

d by

f inco "n o n

-

edn thee

ss tax

c

t toa l u e o)t be

n op a n i e s

d o

s:t v

ye ar

r t ios e i n t h

e eqn l y

to deem

e subje canno co

m

is: ne

tax poies

os and is charu

mm puteunts, unl

ise

s art e t a xeb asn c r eaa te

e r lis t ed

cor w

ora)P ) ( appli n i m

m p anie

mss accot he

m p anieo rpE GG (t o the

t h

e i

IRAn" in thef f e ctivG (n ew

e a r

s)

ra tem e is

s o

"

coo a

R P

E fr om

eP Ee y

u m

ian coren t cIRP

I

R t h rem ee incof busineo v i de

a bl pr

I t a lyI t ald i ffe37 % (4,25 % (" p r o ductio19 % Id e r i vedcapitalthan 27%)7 %first

n.a.Minimo p e r atingr e f e r e n c

e tincon.a.taxbasis ol a w

E P a

e ntg e d atingnd

added by

e s t mc haractur

e r than Ia nufuntingr p o r ation isace

e pl

co

l o we invo f e s s i o n al

i s t r i b u t i o n s ofss is suro r m

f prsn accoe toicant

x cepre c iatio

s "

)

o f

it ist r i butabls t h

e d e

• : 

  10% fd upo si g n ifo f

it and r

eed se r v ice

m e l i able pr

a nce

m p anya tec t to( e . g . dea bl

l o w

dh e r e pr

h e r e dis

m e and 50% oe excr 15% .e re nts

n tivi nancialu g h base

m e , inco subje

t m tax al

is

Irelan28%: 25% w100.000Y e s , w

incoi n c o mclo s e d co

20%, on.a.I nceand fA l tho

incotaxadjusback to"

capital

t

e noe

)

ns ga ins

h ich arx changn thee

ss tax

wd o49

nsr p o r atios ? ? ? capital

st o c

k e

nsm puteunts, unl

nt coise

p anieied to

cor w

r p o r atiot hem

Ae s i dess accot he

nt con rm e is

ted cos o

e s i dee incof busineo v i de

ted in theo r no

a bl pr

p e n quo

s

te

a blned

arnts

n coailfi tioe . is

s e tsl e r a te

usefr d ingo r

ee

a

bl t h

e l e s s

is av asy me quen iss s e t a

o ductioo unt oo p e r tya n b

r i o d oi x e

d bytifu m),

ubsn d t h e r e i sh

e a in the

of acquisn l y

in the

e pre fa r y acco

r prdal ams . I f the

e

d oe put intoibl, i t c pe

on

n or

m ye ar

e arn ars e

t and thea y jusn in a taxa

x

i

m

meciation e

in sn takem a

hand, such accel o wy ar as

i tioe tho

m noo

y tangi lliu t. e mn .n takex i mus e of tal use

r m

nd- ale l y

in thecre e and v

f theductiblpre c iatio

s e

r u no

l ine then

in thecon ispre c iatio

dee or i e d o de les the

ht-icex t twr i o d thef theL 1 m

n tirnsive us

nee

d se

e pes t

od en.

coitiof der ialpre c iatio

pre c iatio

l

e

s

s

than ther e duced depr ise s s def t h e ma

t o r

ic acquis tws

o natur care d to

ity dee d ( t r a igpre c iatio

pre c iatio

a bls t h a n I Tducte

I t a

lyhis-

s- Accelerated depreciation

up todeand theacquirdetaxI f thelesdeacquisRatem i niste

to ther e inter i o d isf e r e nces , unla l f o

r oof of a

m p ar

activMor a pid deI f thepea llowdify e art h a n hn o pcosecto r .

r y

y e arse s i n

achined ingr t ainh e

capitale v e

n-e hiclo r thet h err

eia o

builr ve d

f se v e nth

one a r

s t

d byv e r

a strialo to

fo r capitals ared and

ant ands us

r y and cex y

aceant and ms oo ckl e r a

te of 100% in ther e d .

sis 4%

fo r m

a ncew pld inga r r i ed

a terst

re plu s e

d acce r incur

achinee s 20%d ing

h e fi

n

is fo r pld ing

fo r indusa ncel o w

n

ne 15% and in the buil

l inel ine baln, ale

o a t r a d

e c

s Ho ise hicl

r vtrial

d

e c iatioa n c esht-r t ain builht-r y and builo

md at theh ich it is

ant and mi n

gs for t ra te

10%

t r a igndituru stn can be

r i o d .t r a ige ducingpre c iate in w iso to

IrelanD e pr

a llow- sand cepe- s4%-

r I n additioe x pem achine

p u r p o s e s oft h e CS h annodey e arF o r plb u i ldannualy e arF o r mF o r indus

..e

a yi t hr sa ny

d

f then.e

d to

e a r s . tax l a w.

n mt s wd ari ne

s t oe quirr ysse

o r

ed in mitte

e r yr mu

t bye thoht l

e ro u r yp e c ialt o mc t

or of 350

n co

e v a l u atioy

fn

s

ar fr o m

a s

pre c iatioachiney 1993.d a

xeo r e than Ds e

t is pe

t m ass , m

t

in a ve

se st r a ig

a nce

de

o ductioter r

isingc

t to

s ar the

pre c iatioa nuar for fi thee d .r

ea gr ante

of no

e are n

t ise suln.

a

te- bal

y

to by t h

e fa

r prm p anie

as s e t s ever

subje

dea nce J

ant and mst

ice yi r s

t usa y re a s e

in the

e r

n

o va l u e af

ga

in ar

l ine bal prr fe a tm

i n c e n t i v

e a incrpre c iatio

cliningi p li ed

itior itsi xedn isicablu lt

1992, coht-o r plter 1

e p r e c i a t i one d oe tr

a

x-h ich mf def deh ich appl m

lue fining

t r a ige cle d fe d af

use dchase

chass wtantials os o

s e w

c quiss e

t oh e capitalh e apple t h od

t

srdn

ds.d

tief nor i o d .i zeds ane

d t och andr fi s cal

r e

d oi v e

r d e d ind parr i o d orrew is

s e arh ichogn

servea xa y beu r p o see ntio

e ono r

ei theo r thev e r fpre c iatee pe

re

co thir i n cuw hoe incur dea bl

s w

m e fe s or e rec for ty

mw p ms a

l o w

r o mv e r

a peo sts . Rey art ates ah e

r re

i zedr e ady als k s fromr e

d fs uch as

e ara y bea y beize s ( u s t bee d fd o inco sh ar

mach tax

l y ss i onognu g h thea n y lam e s in

l a wh

e ri

n muntse ars. C and knos t s m

thekso r e

h ich thequal

icalr ovis. No oto mps e

al an d t h

e ons incur

nser prc he

pre c iatentsv e r

3 y

r o

m w

d in em ar

nth fr

e rec thon, theover ts s

ductioss accoi l l acquir de 10 y

e oe n

t co

d fingr ade sp e c if

an d p

u r p o ses av e n tho

n

to t o ch a n ge,x pet he

f ets) .

a v i ng

deh ane pate

on e tei z e d for c se ctio

ductiblv e l o pmducte dur

pre c iates .

T l a ws i ons e s , es i on

e rvesa

x ppount or i e s and on sarke

r m

upo

taxbusineI t a lyG o o d w

m a y be

l e ss tacquirdededey e ardey e arup toL a n d

T ax

resfor tp r ovipurrec ogn

I n additioin thisp r ovif o r e i g n excaccol o ttecos upe

v e r

f f oe a r

s).

n o7 y

d inh e

ne

us

r ittew and

ductem e w

a p i t a ls .

e w

arm i t e d t o 1

w ho de

inco ri g h

t toe ar

ntsfe (li

n knoch is

n thefie s for c

e ov e r

7 y

e s e ar trading

e

ow i ll

de

d patesid u a

l li qu a lis o

oode

ndituric rd

ntifnditur

m putingw a rea nce

n.a.IrelanP u r c hast h e i

r reE x pe

s c iecoi n c u rreE x pe

softall o w

L a n d , G

D e ductibl

a y

w i ll,s in

of

s t s

m ye ar

r d e d in51

f e r r al

re co

dee n

t co

f f

in the

e d

u s t bed.

l o wv e l o pmn or e d .

n munts, nor ittee

n als s e t s , i n c l u d i n g good

pre c iate we incur

y ar

ductioss acco dech and de

all y beductibl

depre c iatio

s e arr ma y beh ich the

a n do n de

se

e d byem i

t

s

:

e aches ar

dg

t hei n g liaden rv e r e d

f trt coo l v e n cy

e noo v i s i o n s aro v i sio

ar pr no

serve , l o sse

r ins

btsr a l pr 0.5% o the rebto

de.

e nei t h i n t h

e followtalh de

d the, Ge up toe s until tont that it ise n e ral

o uge d.

a b l

e w ex te

thrl o w

u r a nceivablf the

al

P r o v i deinsa llowD e ductibl

r e ce5% oT o the

w i t h i n t h e g

arising

also

e

a bl

l o w

e al

o v i s i o n s ar

ic pr

S p e c if

ithin

f of

e w

a bl5 %

52

l o wf

0.5% oi n

g 3

e ala x oeed

m i

t

s

• : 

  a mt exc

e s

o v i s i o n s ari n g lii t h ou

pre

up to wivabl

over, re ce

e n e r ale ductiblade

eidestsi ngs

n y

e.ns

o re ardi tym e n t as coe s isxedr d

ns re corcraft re pair

annual. Au d e de

ar a

u s t beuctioa yo untic

r v e fue fo l l o

wn to

nfo r mp l oye ttingg i nningr aiu ms t

as ye ar am

r e n

t ye s ms andnstr

h

e good publ

e see l f arf actual ex pe

lue of fi

the

actualss can beclical

copre c iabl

co the

i p

s o curnso r k s m

l o cation

a rh

e emh e s basis.n o to be

and wo unt accr in contrualh

e vax cea x i m

t s

in theo r

cy mf thef

de of thest be i n cl

ex pe

al re pair

ic we

al annualo s t

of tf e r to

d to am

ye ari n

g tr d ing at the

n y

e pars .i n g shu m

o r k e r s . Tductios e tsr v e fr mh

e c

w dea intaining% of t

. Aod, the 5% o

g i nningserve mu

m e

in the

o l v e

d in thef publ trans

m p e n s a tio

el o catee v e r a ncer y and con

an accrts accoe as

e arquale rip e r ate se

l se in the fi scal

s governnanceing re g i s t e r o

beo r non oductibla x i m

e y

theb l

e pi e

s on bee incos s

in actuals

inv- de% of t

e to

h e ms 5u t co

t atutoi duale

d o

s i onr a l , the and mp t o 5

e asse

a bld in e

a xaant up a ra inte

n in the at then t h e re

x cee d as fnance

p e r atio a tax

r s o nneductiblingl o we nee paire d u

iblpre c iabl

ducteductioss ia bl

s e tsm p anies i d

e ii l l havitho

y

w w

dy

I t a lyD e ductibl

A mo unts al

pededurw i

th s

p r ovio f indivis alI

n gf o r ra llowtango f deo f taxdefive t

C o mp

m a y se

and mdes h o wase x cein taxA n y e

claim

m a inte

Coand om a ke

r e s e r v e . Ts e t athebo

ns

i butioe

ntri th

we s .

m e s ar

c he

w e v e r , cor d anceincipl

pr

e , honsion s

in accounting

ee if

dductiblo v e

d ped acco

IrelanN o n de

to apprductibl

deD e ductibl

standar

53

e

e

ductibl

e ductiblo n de

of the

r i v e

d in the

g i nning

be

the

o r l o s s e s de.

n fr o mity

s f

ye a r s

itatioe ar

s s activ

t o 5ine

UpN o l i m

f i r s

t 3 ybus

ade

e tr

on l y

ard sa m

rw

f the

ee d fo

m e o

n

in tim be carrie inco

itatio

s mayt fu tur

ains

N o l i m

L o sse

ag

54

s

e ar

5 y

p to

t

e ass.

s.e ar

r o m.ye

o r

at le ari l a r

d to ap p l i e s

oda m

n of 27%

r i v e

d fd fi f y ingi teo re ne ri

h e

s si m

v e r

5 ye ast 3 ye

f or dinar

l i me bee t h od

a

bl l a s t

3 annual mi n

g p

isation ra te dew ne

a y asa-EU

e ale ad on iso r

at lailo ld

e w

f rr r e c tiop e c

ial ga inss s o

iner t ain qual sprl d fh ich havI F O

op tioso av

t

sm e s t

ic. T intr

eee os ;

sf ce.ts he wh e h

u

ch in the sa m

sseo r doies to

a bla bln coa ten is als e

ts s;

t h

e L

n i n

g td a

ailaill atiod

r capitalf

a busns

n can be ash eet

exea l fns.

utr

ies tos and o

sa l e o

e articipatio sp r e adingn assep tio

n c

e se t e rmid in thee n

t appl

abl

I t a lyN o

t av

N o

t ave d

at tim

T ax

N o inf

S t andarapplthe3 yparT a x a tio

T h is

g a ins oT h

is o

f i nancialclass i f i e d as

b a lafor dT ax

T r e a te

l o

ss.A s for fi

T a x ne

tre a tm

transactio

p

ou ar icet edf.f f

si th

s

ot o

o f its

a gry i nge r Pa le ofr

os w

e i n vese are tax

pr

s

of t h

e sade in the

d and sen

dl. Ca l .

r d anceutre s ar

b ern applum

ns

p t i f re u t ra

e misatiot e d by Co

fromi thin 3 yt tra insar

t g ne

f shar

me aln the wainso r

w ga ins.

vedainse tax

eenf

r adjusa

x exems e tsf o r et sstruc t i o n s aa x n

r

e tns in accoe

s o

bef f agr i e d f

e t wsse

r e l and.e oe d oefe ri

r

e t

s e

t isi n s di x e

d ase are capital

se

t o se

t ag

or card

a-cons a

x chang

db l

e b in If asi e r basr

1 yet beu turxeansactio

ye arn l y bee arsti c rem a t i one 90/434 ar

e d

at timx . as s e t s ar f

t hee , oe y

ainst fa l gar a l l y e

r d e r tr

IrelanO ne

A v a i la

trading

T ax

C o s

t o

m u l t ipl

I nde20%R o

ll-over reli

C a p i t a

l ga

f i xedin osalT a x a bl

CannoCan osamagA s for fi

Domea mboD i r e ctiv

G e ne

offt ed

f.

y tot s

s oa i n se .gre

ne s (30%)en

sa l e o

e i n vese arns (e cton l y if

ntiv

pt in cases applh arain s t g

s .f

a s

x ceisatioa tef strial

nts

ll elem the

d rr o mp t i f r

i thin 2 y

n ee alny l o s sf

f agaren o fr e e o

ince

f

r sa l e of indus pate

wt oansactio tax l a r g e

e d sh

i ctioe o

r r e ctioe ducee d for a

a l e or i v e

d fa x exems e tse sei butioa dep e c ial

e tims ;

r der

e t or dinarf l i sti l a r tr

ntr mr sating

a

bl re strn co

a

te on the s su

ch as

p let

o me l i e f

e

as on l y be oa y be

ailc iall atiod

r ga ins

and the ri g h ts

o t

c ga insi x e

d as

• e) 

  mi f y undee d

at cr

rge r s

e d

at the as s e t s ar f

t heeabls can

m t h e sal

r g e r

s and simi s i o n and coitya im

y quals ( o

t avo speaxo info p e r tya b

le no l l

• - 

  o v e r raxe ductibl

al.

n e u tr

56

s).

m p anie

t bet

en

e r g e re r g eri t hpt

ora a no

of them p any

o rpu t

wA P )

r i

ty of the of m

nsa b i l i t i e s

e re s u l t ing t h

e ms andp a n i e s

ndsm e tax

I Rr o

m ex e m

U

E

o s s e s cannoa jop a n y i sityach co

n case for ci deG ) b

f ei tation d li

s arr omibute co m

P

E incoc

t tos

:ive d fe tax

ar

n, l mo

mm p any

gime

s

• : 

  divI R

ubje re ce

itio theh

e c activi

m attr

e

• d. 

  Is os s e t s a

coe rged

taxp reb l

e (r p o r atet s

m p anie

ndsm p any

ard if

t s

of tf l o sser t ain lm p anie

theh

e m

ce t h

e a

rge r .e s u lt i n

g f

l them p aniea xa

f acquiso r wgh is change r od bys e so r coe no

arnt coi dec e i v e d from a

d co

e oc t top a n y rs of t

b ovee c i a

l groupo ar

e t

r i e d fi n

g rif e r r e d and theansflues of

e r g e d co

mh e meo mv e r alt i on

m p any

trubjee vam te cs o

liga pure s t

ic coe d

it fndsf div

ider e s i def i l i ate

afvid e n d s re

I t a lyI n cascarvottrans

cotheis s

Tho f them a intaine

froThtakeobS

ee aNo sptaxD o m

r e c e i v edf u l l cr( d ivN o

n-

• - 

  60% oEU- Di

ard at

wu c he nto r

a l thes s

e e n 75%o r

w se t ne

utr ga insst t o t h e

p a n i e se d

it f

tws can be

t be neo mf par

on e

p t i o n .

beee

o sser i e d

f of the

o r i c cod a t i o n au l l crr s hip.

ades le taxi stp of cs

to up o

r a n s f ers caro f its fo r capitalf e r l o s s e s and ex

e

ms

• : 

  f

w ne

f

a trns armo n s olia ncea de

l o ssef e r cannoe prn. grouans mi e s .

a

n

i

e

s:m p anie

e r

om p anie

anss the t h e i

r ha x c trl o

w 25% o

fu turs at of a al

o mpnt co

dansfd co, the

r e d t o t h

e tise of tr.ansactio of td to

e s if

ciater

wt thee r takeo l idatiob ersitte gr o

up isidiar

ubs

dateu r p o sens

y s t eme mr m

ss capitalr . Asti c c

r e s i den tax

ainsm p any

h e r e trt heme

IrelanO n

a tr

assot r a n s f erO t he

theagcoWacquirt a x ps e ller.N o co

N o s

b u t mare pe

e x ceanoand 75% sDoN o

n-

f o r e ig

r ofn.e

frgi n g

ardrgee v e r ,s e tsm p any

e os e t oi n

g fromo r b ingptioive d

a bl

o wf as

f meo r

w a me riu

lt absx e

m re ceo r

e and taxe tax

s oves. Hh e t a x

r b e d co

them e .

n in casr i e d fgime

o ssei n s ress e tsu l l es

:ndsn l y f

e ctiv

t car ref asfie s for t va l u e o

absos

• : 

  fi deive d ar

v e r bye d

it o

rge r , l

e nop a n i e s gin ou a lie , thef them p aniey dir

re ce

e d it

re s t r i ctioa n d a rda p i t a l gar qn od a t i o n regi

m p anie

ns ar co m

h e strgee g i

ms o

takent co re

l

i

e

f

• : 

  divith cr taxidiarnds

o n s olir ale wubsi deu l l cr

p e c iali tio of me

r eek of c

mee ritie

e s

tic cor e s i dea

bll

d

ing

• : 

  divith f

m p anie

d e

r to Gt i

on re

• - 

  e v a l u ationtivi abil

m p anya x cn il ate

o so mo

n-e taxi thhoa r e

nt s

e a ty

pt.

i

th a 25%x e m

w

m p anye 95% e

d co

r s hip ar

ne

aff i l i ate

o w

58

stn

e

d co

lue

hteo ductio

prn c o m

e ign l y

in the

e t vae ob l

e i

a rke w

e r agr e d

in the

ductibls t a xa

n d

mm e .

de.

les

o s t

a and av incur

e ife t h a t gros inco

i ontal

toi a

l ru

er of c, F I F O

t h o dn c o mr e s t paid isortp ec

ar to

I t a lyL ow

L I F O

meD e ductibl

of iI n tep r opbeNo s

d ..ns

e tho trade

e s ,e s .

lue

st mnse s

f thef taxe s t r i ctions

e rs sx pe

et vae co

ex pe

s e s o

a rke r agpom e n

t oo miner e

o to

paye . S

re v e nue pur of bus

.

o s t

or m and avnditurityles

e , ifo r theo r l a tee n

t and m

dI F Os fi a

l ru

er of c, Fr e

d ftie ex pe

ductibil

r t ainmp ec

IrelanL ow

L I F O

D e ductibl

incurP e nalCapitalo n dee n teNo s

d .ef

e thoa il urm e n

t o

lues s

st mineo r f

s f pay59

et va

e

co bustie

a rk

e r agd tonalo r l a te

r f

n d m.

re l a

te and pen s o

les

o s t

a and av

I F Oe

if tax re tur

i a

l ru

e

r of c, Fity

r p o r atee sp ec

o wI F Oe ductibl fi l e tax

U RO

sns

a tios es tate

s

m p aniee al

e bankse sn titie

r g a niz

f

90.151 Eunds

t

om p anie coe

nt e

s and in r

acken coa nceo p e r ativo p e r

ativ and fe s t m

t sbon-pro f

it o

itie

insurans

e cur

i r s

t br

n

fE pr ofi

dro car

u tualion ple inv

in s

u r a l and cor 25% nonsl l e ctiv

e s t ing

S p a in35 %30 % oof SMn .a.n .a.40 % hy25 % m25 % r20 % and 25 % co10% o0 % pe1 % coinv

s

o f itsg e

a r n ingd pr)c har

d este m

sur

i ty

ye tainet r i bute

l i dar

ano r ro r dist e t a

x sy

l i t ra

Germ

40 % f30 % f(spn.a.5 ,5 % son .a.n.a.

60

e s and capital

h ar

unds on s

funds

ion f

nds

nsions

i dens

o r pen div in pe

e dn l ys

tt

l o

w by

e

d o

n thee

ss taxe alunts

m e is

r t aini s eda y ben co is

a w

d oe taxsed assea l u e .

s e s

ar accor i

a me e

Lt v

iteo ductioe crw ne

m puteunts, unl

isepoo s t incom e

ar au t h orf e r e n t

i sh enn , dif cr.r prg u l a ri

cor

w puri ng

ss accot her d e d

in thes basis but cef incon basly w

a y ern oy a l

• - 

  D ne

m e is

s oo r taxe coe and muals o ont r a tio

i nisa xpi tioa l u e ) , rer this

r

96 Ro

n s fasseizatio

e incof busineo v i deh e n rnditurn accr

d cl re

al admn and tim

sid u a l v

e d unde

a bl pr by t h

e t

ductiox pe

e

d oi tee p t i o n a lly taxtor i

c acquis

o r tis

taxbasis ol a wdeo n l y wall e

taxl i mo n theexctheall o catio

u s edhis(le

ss re

v a l u e undeam

fs t

n co

n thee

ss taxe o

d o

p r i nciplo ductio

m puteunts, unl

ise (r pr

cor wn o

ss accot hei tio

m e is

s oe n e

ss)

e incof busineo v i de

a bl prr itativ

t o r

ic acquis

taxbasis ol a wauthohis

s t

n thee

ss taxn co

d o61

o ductio

m puteunts, unl

r pr

coise

r wn o

m e isss accot hei tio

s o

e incof busineo v i de

prt o r

ic acquis

a x a bl

r

(n o t

o l s )

-i n gs,D and

a nceu i lde d

, R&ns unde

ball o w

e and toi ning

ot for b

d

is als o ciatio

cliningo r m

u r n iture thon fe as

t ur

des , fi g i t

s (n

i

ne an d t ools)

e

in m

ht ld ingh e

-d

pre c iatior i cul

m-of-trni t u re

changr t ain ag

S p a ins t r a igf o r builsufunof r e e deceA c

t 19/1995

sdn

e

d ingo re thoe d ist t o

r fici n e s ) and, ar

l o

w theghe d

in the

pre c iatio

o r buils e

t op e c ifg. m

d al to ye ar

ou tput m de50%)chas annual

f o r s

(n o t fs s ase thoa ncea s t h

e ri

nn alf the

ined s

-d ( as s e t s ,

e.

bald (20%-20%), pur

i nancial

n e

r hE (s e

ts f

a ncee thoe thoxede

in me thoa te

clining

mi c owpre c iatio

r additio SM

f the

yn

e bal a bus

e s )n mn m

dei ne

anible s of fiom ded o

L änder

e we

at 50 % on r

e

nt byi x e

d asf o

e f

rai g h t li

cliningtituting theht lon the

nso ductiople t io

on l y changl e r a teh e n

e s t mnd hal

pre c iabl

pre c iatio

Germ

• - 

  st- decointang- pro r dec a t egorithef r o ms t r a igt h

e ecclaim

accefor tinvm o v a bl

s e codede

e a r s

ant,e ar

, pllue or

62

s )e n

t)i r s

t y va

d ing (s hipss t h a n t h r

ee y

n in f les

w a re

(buila ncequipm

s s e t s

of lowe of

l ine bal

ht-r y and epre c iatio

f e t i

mD P soft

e d for a

t r a igclining

u l li

achine

8

ra teo ds

e rif f i

ce 25,i

e

d

a

r

e

:no r

m pd ings , o

e n

t and

18,20,25,d

w: 8% -u l t ipls ) , 2.5 (n at any buile ar

m 2.0 f

taket ex i mu

trial

68 yquipmr harns 14)

pre c iatio

bee s s than 5e ar

de8 y

m ram putel a tio

t ale toi e d by

a y ben d mai

ndus: 16% -n in l

n ( 100, e insa nces are e n

5-u l tipl

n mx i mut e

s ae

s

:

3 % -

8, coa ns

bala te

b e t

w ms e ts

m raa tious

s

:

2 % -r y : 8%,10%,12% -f f i ces ,

v r

o r e ) be

e c iatioo r tisa r e ho 8, o, carcliningi ned e p r e ciatio

ht lr me to

a x i mud ing

1.5 (s ) , 2.0 ( a r s os arnd hand as

D e pr

u p t o t h e maMo f amand wbuilm achine

too

l

s

• : 

  30% -

25% -trucks

f o r des t r a igbyy e aryer a tes e coS p a in

s,,

d ings , 2,5 %t he

e are n

t (d to10 %)

20 %)

buili necliningite

t r a i g h t

• - 

  l i n e0 %)r y (f f i ce

o r l i v i nght ls and 1,25 %f 3, or s ( ye r c ialf o r e 1925, 2 %quipm l i m

e d fa iga b l

e sm oachinepute25 %)

ertm ms bee e n 1925 31/03/85,ter 31/03/1985n

• : 

  de

fi r s

t 8 y

e are a r

• s) 

  is

6 y

6 y ra te20 %)m

r ops e usn): str

d ingtws af

o r thei ngr y and ellow

x

i

mus

• : 

  m, co20 % -

e n

t (t c ( thos bed ingm mo datio

i n i n

g 3a nce

ma ra te

b l

e p and coo datio

o r build ing (5 % fmai ne10 %), e

y

i t h aht l equipme (ucks

an

m ovatrial

m mo r

buil acco fo l l o wachine bales t h

e a

i m

a nceh e res , tr

I mindus( e x c l udingacco2,5 % ff o r buil4 % fl i v i ngbalf o r thefor tplant, mclining

det h r

ee trat e , w

s t r a igo f f i cef u r n iturcarGerm

e

n

t:

quipm

8 %63

of

5-

r y and e

u m

a x i machine0 %

f 3

s

• : 

  mant, mm o

ding, pl

x i mu

o rf,

d fa l

fa t es

t h en s .m o u n t

unte va l u e ot i m

onit oe d

fr eed i t o

ri actual

l i mp e is' es

acco the

s i on oft s i s t a xnditio

t for ee acquirt

r

o

ma

s

es: annualf t o t a l au m no

on the l e g a l and sco

bei dea y e rs

acts.

u

st ben f pr ovif asse

r t ain cor e sents/ 1 0 o

pre c iabla x i mo unt ift y , if

e toe f

n ms as

f

1 m

ductiou t i n ca l u e or ceitie amnding

t o t a

x pctiv

t h

er a) be det o

and patei minnualtald parpe

t havo unt, a w

a deh eetne t v

e undestric t i o n s a

e cur

e s ar toa l u e

e l aten de

novenb je

gin o

pre c iatio

iblm l

der ass s e t , ei

n c e sr mark and sm arks

w i

ll: af the

a unrs dollyd o

pre c iatio

ductible c i a

l repre c iatio

cre a s e

in v

r t ain in ame ra

any( e ithet h

e ab a ladeL e sse

deSpf undsI n tangT r adex i mu

maG ood

1/10 of r o mdedel a nd.claim

cegenif base

tht

t r a igw n

to ar e

e

e x c e p ta r k

et t h e

n d

or itsr yn de d

t s a

r d e d inn a si a t i o n lan do

a i r me lowe n s eso r tax

forses ar

e fe n e s s )sks are

gh

a ndatoe n l o dg

d o l i f e (r ittense haspo

be

re cor y and canno

e f

ul de p r eclue of

wt l y b

f t h

e fb l

e exp pur

a n enductible v e r s e dr itative n t a l

ri mo p y ri

ex

pe ye ar

a n d a t o ryf rn mn is has

u s t bea ndatopre c iate

i r usxeds )a y belue i

e oc h

of caim

m m vae r m

n munts,e de fi

e are pst.o r e s eear e munting or autho

n v i ror ea

n isdo r fh e n the

nev e r thei t h a

e s are p t i o n t h

e vas e

ts fi nancial

accor y and der eh e n actioa b le

asa r k etn cos f

ductioss accos t poibli s oi ll wf

15 ya s gone ws t h a t ap r i nciple ws for bh e r e

a cl pr ob

depre c iatio

po basw

r i o d on exca i r ms i ona ndatoo n

s fo

lue h

ductible

in thes i on

m me r c ials e s ( v i sis i on, w

ductiblntss very

taxbusinedebeI n tangl i negoodpeA s a

f i nancialt h e fvaacquisitioP r ovidecausProvicoals

o

mpo

purMassgeblichkeit ProdeProvipateor i

f

,o r

r ei s andv e nds s o

ines s ine fr e e

a ndt s t o

inens and

en

w i

ll) a basd the

udingducteu r a ncen

ductiblr i

al tax

i ne de bus

e and lr o v i s i o n s fnditioa kea y m64

s

sm a te

e e n taxineht l4 , 2 8 % (se (i nclo v i de

pr theo r bus

twd toe d fb ank, insm mo datio

e . Pe de

art im

d i n g goodt r a igm 1ntse ar

r (alth cara blnst r i

ct cos can mt u

re p

x i mu, accoaile no

l i

nk be and bus

nsns (in c lun a sw e v e r , can be

fi r s

t ye l i nkeise s us

se cto

t

c.)t av

b l i g a tio and sy arm p anieover fu

noe of ma and pate

e s are m

e etel s ) , he

e or o wr s o n s .

d the cos t o c

e c iatiot e

o- h o wiblt e r p r i s es , pr

h e r e isprpre c iatio

pre c iabl ra

a r s )nse s ) , ho

fi nancialr a l l y noanter nara nces i on

u r e

d pe

n o wf f i ceo k e r age neo v i de

ly,

urtd

ion

i t

hm e s

nso rt

undsh ich

a b l

e.c her pen l y fe no

a b l

e on ov e r duen y

d ws o ciate or

re s u l t cons. oion f

llowllow of theion sr s o n w

o nthsn t for ah a r gei th as

ns unde

e ,

i.eh ich arns

pem u m

r peee

r

e ar e as c wt r a tio.

pee ntsa xi pep l oy

s as a any on the

btsi nis

tom e

s wplo y e r

r n al

s i onnse mductibl

em

inteee ar em

s i onu l ifiess i n solvee b t

or inds

peo r de

e quire de

r ovi.n

fic adma r a n t ees

r oviubtf

applo r e than 12 mi bution sche the toi f i e d mr y

ns

ductibl

l pE ' s doe b t

or i) th e d

y gu

ntre e t

rn ar sp ecn peg

e of

i f i c pi ng misatio

bt ish e d (3m e

nt dei n gs

s , publ b

o l l e

d byi butiot des ai t h a

) Tu deedductio

tientrntre ii butio w

ntrr i es

S p a inS p ecG e n e ra

for SMa def o l l o w( 1 )

it is(2rea s on

fra( 4 ) p ayp r ocN o de

parc overedo n l y cow h ich m

l e g i s l atiocapitalcocoare no

t h ercova

m u s t

r iteed

w ar

a b l

e (tialo v i

de andd

at anns in

llowm e n tsi alal

s e s pro r mc o untei ctio

s i on

r

e a re str

r ovi

s

a of

a par

lua t i o n ) .ion paypo actuarn

a f

ns purd oan; dish

e p

s i

on fo r m

r t e d by

o r peo r taxion pl

yr ovie fuppons basens of 6%, no

s

e of t

an in thei n g reva

i f i c pa dee s pea te

mductibly arm putatior e

st rc t u a

l u

Germ

S p ecbeoff followP r o v i s i o n s fdethecobindingintet h

e a

t

e r a

ln

e .h e

y t h a nion

ns

a blo r gen. Tbts

ionu s t beer.

pe an or

l o wn actuale g i s l ations mp l oyr e f o r e no

(o t h er65

e ala n c

e f bankso r de

eese the

e s s them p any

funds

llowies tonsion l

e

t up pev e r e

d by co t h

e em ar

o v i s i o n fr than o

pet sp l

oy theunds

i t e d

a thee r s ) unle coa nce

o v i s i o n s aro n t r o

l ofion f

lim, ra the

s . too r i t s ema nage n

ts ar

i

c prs

a ru l e appl

e d insurund; thush

e cns

pe

cifs i onduct a pr

r d inge s fr ize d

ere itandings

plo y e r s cannor a l m

m mi tm

nsion f

t s i d

e tr n al

sp e c ial

a y decco

eso r

s aritie

n anyan f

activ pl tax

re pairp o r

• t. 

  I

d the the

i s h ing

tantialf f trans

o v i de

ubse oo v e

d bys

o r sh ipping so pr

appr ye ar

ee

e

in cassa bla bl

and se

it iss is

r itie 10 taxailail

ductiblr cas

re pair

P r o v i s i o n s fdeand airo t hetheauthoup toN o

t av

N o t av

a y

nance fr e e if

i ngi l l i o nh ich the

a inte fi nancial

e taxu

t in the

o r

tax m ar

fo l l o

w in w

E M 2 m

r y fo f itsr i e d o

tantiale d in the pr care D ye ar

is of the

thed

rre

m andatoubs

o r so r k cause ducen

in timr e d in 1999 and 2000 mior to

wnanceo nths

pr i n cue

s e s ) fh ich re e

m ye ar

itatioe rea bl

r i e d back up toe ar

poe pair wa inte

ms wail

car

P r o v i s i o n s (puro r ry e arthef i r s

t thrf i nancialn o l i ml o s s e s incurbef o r

1 yl o sseN o

t av

chaft'

66

r

gans

e` O

a bl

ailithin

t avse

n l y w

e ara bl

r a l l y

no 5 yaile o

e nep too t avv a il abl

y

ynt e

fo r

n yitio

r dinarfh ich thec o untinga tet i o n ra

e

on r. A

e

as o timy dis acquis

x a tio i n fla

1.023)

a blr i o d

in wr d e d

r d inari ndeo p e r ty

e tax pee co

ao

m e

at the

taxx trs e

d (o r 1996:e pr

e i n d e

x on

nld b

ra te

ga ins ar

s s incoe

in theize d and r

a s pasm o v a bl

1.851, fh ou

ineizatio

a ble ale , im

lue s

sp e c ial

S p a incapitalbusr e altaxare r

I n 1996 an ev a l u e w1984:timvano

y

r dinarf

e o

e

as o tim

a bln

e tax

m e

at ther r e c tio

yn ru l e

n co ra te

an ga ins ar

s s incoc iall atio

ineizatio

sp e c ial

Germ

capitalbusr e alN o spe

n o infno

e of

i m

s s

t t h

e tine67

b l

e ay bus ru l e s

n

a xa

r

e t or dinaruntingr r e ctio

i n s an, as accon co ra te

l atio

izatio

m e

p e c ial

a p i t a

l gao l i

nk too s

t so l l

sse af re n t

t

me

d ai t h i n 3

xe ona y o

wn vesductibl

e n t wh a r es de

l for fit m

rray l o s s e s , taking

e fen ves of

5 % byus tax

a

x d of rei a sa

le of s

l d ing co n d i t i o n

of rei

or dinare v io

l on ho

e asunt prs

e n t

of te ferrau mef on

co n d i t i on

a r

s accos i on

E l e m

onyet a x

dm i nim

over relid e ductiblintop r ovi

n off

e ofs)e ar

fr e ee a r s

t i onn orise

a 6 yn t a r y

f ao r e igw o y

e oa xan t e

rp a tax

c

t to sa l

e fr e e

n volus e tsm t

salpth

e tm t h e sal

o ductio

pr the: tax as

y l o s s e s

thee

nt in a f

ex e mstry e

e a tingubje

crs e tsr o mx i mu

s

s

e

t

s

• : 

  tsin g fro

r theior toa s

e of i

r o me s t ms on c ase f

n d foree s e r v e , sf thele. or dinar

inve tax

are n t

of ad ingf e r r e

d byr i o d pref ii t h i n ma

ga ins f

ga i n

s (ariu r a l a

dem e

nt r

pen oductibl

e d wi a

l rue as

i f y ingp ec

m p any

l a c e mi t alc u ltace

l d ingp o s itio

capitalqualcorepcapl a nd, buila g rim a y be

r e plhoroll over relidisr e s e r v e derei n v e st

No sD e ductibl

e nto r e .hy

ert

e s t mt

• c. 

  are me s ine r taxark

a nis

s eh ar Dr op

y68

y

roll overe invare

ertf

r ye a r s or

on ss that hasl y l o we nm

Dc

t tob l

e pert

f sh

r ope oe o

l

p fo r 3ubjem ovaainst capitalr op

ga insm p anie

tantiale d to

e ld co

subs

m p are n

s i mf f agb l

e p

reah e

sal hy to bet o

n )

ies

in cas

m th en

i n s ont appli nancialc

t to

a tios e s on sem ova

i m

p t

wn fubjen than co

e n sdee s s

it hastaxn l y be

C-i n s on

a i n

s froo e s noa p i t a

l losa y o

a s e

f

s ,e r

e r oo rde 90/434. i n c

s i on

v ei viansfss-b

, di r e

ctiv sh a r est i o n

abo D ofn i za

s e tse rgerse s and tru d e

s cror the

t r a n s f er

f shar i n cl

i x e

d ase oh at unde an d reorga

ns

o r fd e d for m

t s . Tp t i o n one rgers

S p a inA s f

Provie x changasseo p e r atioexemof m

r

t he

t i on

e r g e r s and on i za

y l o s s e s

f m

e ot e reorga

or dinar

ora

ye asn in caso rp

anptio

s

of c

Germ

D e ductibl

E x e m

form

f fl dse

i n s

e l dre 3te ar

m e .

se t oso he

acte ton c o m

t

• c. 

  ho te foainsne

s 5

h

ar inco of

de r g er

s er e ne s alo r

5 yntrr

l d bn l y bet ag co t h

e i

n sa p i t a

l gai ng

e r g e r ,t s anr ee m69

areard fductibl the

eeding

f she a r

s as soa y on sharr l o s s e s oe d ot thise

up to

l of cf m

ainsa bl

aile of asse (t a x f

e oe yares ( but noo r w

i nancial

on l y deo t excs and unde

act dur

n

sh ot

he bas

e ferra re captur

r o

actsf f ag av

s or s hip m ga ins oe arr i e d f

n

a f n

ga ins)

care s iso esntr ye ar

ntre

t o in cassfe

ard is for duding

nsr

an sh a r es

n t h e

sal t h a n t h reew nee co

n that no co of

s of oa y

be l o

ss oh

ar 5 tax

n se n t i t de n so r w

ductibli t a l l o sses o sam

ainst capitalr capital

y me d onditioe ber y

f allown incl

o n s, t

a r s .h a n ges

o ssee c e d ingu lesa tiopre c iatio

c . )

s

tf

d if

n colue ofarh e

end ot ae in

f t h e

a r do r w

r s hip, theitio ou

i n a

l vat y

of tter the

chang

forww ne

acquiso mr i e d fj o ria rry

e

d afd, at the the

25 % o

o t c

a r

ry car

e

in oh e maw ne

h ani d nior to

e e n thet be

r

twe r s o pr

an d t h

e ne ss t

of loss c bei p

of t has changccu

chas, anda n y dity

n

of

a chang

h a r ese s , canno

n e rshi d oe ar, l

puro mp

s s activ

t a t i on

f e r e nce shar) owe capitals de capital

) th e ciner s hip

nenso l idatio

limi

in casedifof t h

e s

the(1sharl o sse( 2 ) thet h e l o

ss yshar(3buso wS p a inCo

o r ee is

mre

ww

e s ars os in thei ce

e

d ifi t h nf ne

n

om p anye arm p any

)

l o w'

s sharntinue co

con l y wo r

5 yr

s

hip and vc

haft

o ductios s fh e n

a cot ne

m p any

ine

m p any intrh a b i l i t a t e t h e l o

ss-

s

s

and theO r gans

f

a co cou s i n e

ss mai bus a parn ( r y

• - 

  o v e r

is disale s s theine

rve s t o re

s thep er y

• - 

  o v e r w

d intoya tio

s i t

s b unl bus

scoe r tean

o

ss carartt s se

ntinue l o

ss carnva

a lthan 50 % os o l d and therest

ass e ts

assem a king

cosamenocoversGerm

G r o

up tax

s e

s of a and

r i e do r e

c . losth them p any

sn loss

et carn . md byt t h er y

f bon cases i n t h e; ife a

t bea tio car70

t i me

e r g ers or b e d coard. Ip p l y forns'

s capital

ow ne

i tio is, the

e arme

les a

int taxm p any

h a n t h os

r

ee mc t t r a d i n

g i

m e

y absoo r wly t h

e los canno

r jo

r i e d f on acquis co ye

ar ye ar

s so

s istri

a t a

x f

and them p any

e c i a

l ruf the taxh e

r t

oto ren

ior inco care r g ers undef thee

rs of

a tax

s)

a s

e of prr b ingt be

b l

e mr b e d coa r d . Spe s p e c

t toold of l o sse

c t e d (to l idatio

m p anie

ith r end oard

g i nningm p anie

nsrwstri

f 1u ch

a,dainst

nmt

wnso r.

rr i o d os t ff 3nde

e ntithinc

t to

i t h i nfp a n

ys fro

t beains

u

st be

y ,n o pea intain s ye ar

mh ich o parly t h e

o r e than 90 %u

m re quen alr i o d s

o ex te

o s s e s

w subjef f s e t ago m

, onn t l y wr i o d

o oe co l idatio

ied ag

r p o r atioi nim theh o u l d mo r pea y be

of the unit:

m e isrre

pea mns

con i n g l o sse

vo l untarm p any

p ain wy , mi scal and l

then l y be

comair i o d cannons

n is in Se ctlnt co

o r

a ma king an additioa s t s

f handsn e fo f itsm i n a t ed

t incoi e

s cu

an

ior to

n can o by t h

e s

h e n thes, red pe

ard and appl

r p o r atio

o l l i ngn tityr indirndens f mn and singn ls , and m

,

as od prr e elid ne

r n edn, w

o l idationtrpe

nt ey o deior to

dur ye arl y

n in thep

a taxmp

x p i reo r w

ouf co

l idationso l idated co

nse ctlr p o r atio pro l idatio

nsl d ingnso l idatiof i nite

a tioo l idateo l idates o

m p any

nsnsr p o r ate gr o up, pr

nsoi t s eao d e cor i e d f

coT h e co

r e s i dediro f thecoy e arcohoCof i nancialindetaxcocot h e grcocol o ssethecop r ofin additiop e rithecaraff i l i ate

.t

anen of

stm p any

e ntn

o l inga tio gr o u p

can beg reem

r

a con d par po as

a n y mum p any

c ann g am p any

of the unit: l o ng

o mpnt com p any

omiooli

g r atios pi scal gr o

up tax

d c cor s hip, o

e s i det nec on

inte handsn e fo s s e s .o l l e

d co the

a ro l l i ngntrf e r r e

d as

naln d

• - 

  losies

o n t rollentr,

a part

-a

n a n c i a l, ea tion in the, as o and lior to

f

a co trans

r a l l y be co

i dual fi pr ofi

o f itss or e

d prt ben appl

a tio

(1) th e c

g e ne( 2 ) theindiv(3) a

o r g a niz(4) a

T a x a tiom p any

coall pr

L o sse

incurcannotax

ntn do l l i ng

r e s t

in thes e s antr

co71

n-r e s i de

e

ct inten d los

t s

a of the

r indirith a noe d

r ofi

n wl o

w hands

ll p

e

ct o

m p any

a tio is al

r con g of an

in the

m p any

a tiom p any

5 %e r

s

s inr l o w

i n g

s >s t o

e d

it < 5%r t ain case

e h o l ding crsn co

a r e h old

h arf taxe d u c t i b l e

a n i e srd in ce

o r

sm p anie

o ductio

e thoa x d

o mpr pr

e d

it fnt co

r e d i t s for shn mn olue.r

e t

s a

sti c c crf

5 %, hal

itio va

s i on

mes s or e s i dea x cptio

e m

S p a inDof u l l taxe x ceN o

n-

F u ll t OExacquism a r k et

Provi

s

n

e d it,s

n cr fo r e

ig fr e er e a tie

n Ta r k e tlue i

m

putatio fr o m

ive d tax

a x a tio ora irn t

ndsr fs e i n va

ane

re cee T

a n i e se imi des t , oe c r ear m

divo ubl

c q u i s i t i on co pe

o mpa y be

ing t h

e d be

ye f undablpanys m

r t ain D

ansti c ce r of a

acturh ene d to

mer c o mr ce

m p anie

lue w

Germ

Dof u l l and rintecoundet h

e lowm a nuf

vapre s um

ae s ,l dh

t h efa y ben d

sa p i t a

l ofsm p any

c

t to sh ar ho

coark

a niseeni n a n c i a l (a

st od s ml y ) .

h

e c fr e e ande r tax

e t wt orye ar

72

m p anieubje

e

s of 21,12 % fr o me

nm D

taxe

d at them p anie

f theu s t be

s

:ndse b t h

e fr co

d of, oe tho

n vene d y

s

• : 

  co of t

5 %, ar ra

te tax coe s ml y l o

w Dc t to

nd isi dei nancial

n fe d toubjeh oosice

e taxi des har prp

of i

ive d by

ye a r )m p anie

tantialp ar

a n

c chang

m p anie

r i v e

d bym p anyt t h

e eni f f e r e n

t m

div 66 % ispt (o r divo r e ige n s

n )chase

des t h a n 2 co

f f e

ctiv of

32 %)

re ce

o r e than 25 % om 1nt copt fubsm

be

i ngf thex e m

a ininga

te a sa tioa y e r clue ae . Dc h grou

pure s can be

e s

tic condstax

i dei n

g lesd e

al rnds m

n i mux ces in a f

pay re mr e s i de, ec

t ton than co

r mu l l y e

l d ingdee s s

it hasC-a

x p va

e t, theactur

for ea

o mi v i del d inge fr mio

n-a r k eta nufincipl

i a l

m e rce d .

us

co m

a y be

i zedd s m

ogn

e tho

h

e recn m

a n y

of tv a l u atio

p

ou

s

o r grs

sha b le

lue f

e rir good

i la

a g

e vai m

av er

ex c e p t for p

lua t i o n

of s

L I F O

w e i g h t ed

va

d

hte

e ig

or w

73

I F O

n l y F

s e s

od

poe tho

pure m

e r ag

e d

l o w

e aludeies in

t y , t h ei th

w

1 appl t r eai ne

s that arses

nse s incls a

nsepo

x

pe of

3:e id

in l

h erine.

x

pe pure e,u t,

e s ,,

ns ra tioe tns

untingn tax paid oh erter m

sts and eductibl ex pe

quitye denditio

dends

n-r p o r atioi dei ngbt/ee r a

• l. 

  Ws ar

S p a inall co

f o r acconocodivfin e s,

gifts ,

g a metc .

a degenr a tiom a r k e t co

fs

o

rn oee si on

e d fes,ia.

ns

t ieorti thh iche a ser e s t

-

l o

w orr alf the

x per op of

bt wn w

op e r atio

ia..f etal

e

al t o t a

x-frenal

theo r debt oi x e

d inte

fts i n p to

n s arnse s that arr als ,...)i nteb e r o

e (e m

c t lyuse

r the1 fo r deivil e g e incr

fo r f

e

d byi ntei re a m

ex pee ( ga ins, pe

ductibl

ductiou e s t hoard, 80 % o

e a l s , gim e oa r i e s ,...) of 3:1 f paid pr1

ra tio 9:

f i ne

ye s causs

ductibllat i n

g d

r capitalo r gt l y des m ra tior e s t ism p any

anr a l denso r y boe incor e s t , 0,5: coquity

t des fe e s paid to

f f

e nex peu s i n es

ru l e s dem e o

nsen l y parr v isu s i n esa blquitye intebt/e l o ans to

n o v e r and salbt/el d ing de

aring

Germ

in gall e

t h

e btax( 1 ) noe x p e n s e s reincoe x pe( 2 ) o50 % os upefor bto tax

turdef i x e

d intev a r i abla hothebe

ofbte

de ar

e n

eo l l i ng to

a intainingnse s

ntr ra tio

e be

74

r e

d in

ductiblx pe coe l a ting

s rquityt y

and me dee n

t ebt/ed that thet hav

d par

nse s incur

i ngo v i deuld no

x pee curm e , ar

r t ainmnse s paid tom p anie

e

pr a thir

n teo an co

r a l e, sx pent cof

a 4/1 der o m

i ngs s incof es s oductibld f

e neiner e s t eo l l e

d l

n-r e s i dex ce

t dentr

o r

ne

a bls e ts

£300,000, re l i e f fl o w

untingr p o r atio in. tax

capitala nce as

e d ingm p any coe nts (e .g balf the

x cen accoe tot md bye o

l a wd back to "capital

ace

t eall co

clining

smd upo adjusd by natur

m e l i abl

i

th tax addeace re pl

i n g d o mo f

it nov e £300,000 and bec

t to wn ise pl")n

is or de

Ki

ne on the

edo r pra r g inal abo

u g h base

m e , inco subje

a n c esht l

isr d ancee c iatioa n c esnding

pre c iatio

pe

Un it

30 %20% fw i

th m

pro f its

£1,500,000.n.a.n.a.n.a.A l tho

incotaxaccodepro f

it and r

a llowD e pr

a llowS t r a igde

s e )fn

c i s epot be' s)

ye ar

e

nt inr alua t i o

nd oo s th e n

a tes ,r a ls e ts

o t exera

in pure e n 1% ocannoo r e thanusm e nts

n thee

ss tax

m e is

s t oa nce

re va

tw.culo r md inge d we neo r as

bal us Ge f75

o n

i r me v ioe p aye s t m

p a n i e sd o

n co

ber mh en

n o v e r ( pr pro ms ed

all y cald but ff f i ce

o r builininguld be

e d bys s i blg e r than 40.000$.

t h a t

d the

as taxy

cm puteunts, unl

iseo s t inco

cor wo ductiolue w

r me thocl po

au t h ori

t i t i esityf e r e n ce' s tur

and theundse

nt and invert

ss accot hes basis

r pr no m (n o t f

e s and o ded s cog n is.

e coa xesn ist big

no

en difr p o r atee mr opm e is

s oe and mual

n

o ne t vally

n isl ines e tse hicl the

e tho

theus ye ar

T E 100,000 o

y coion f

b l

e pht-

itio legae r ve c iatio

a l u e is

a s

e ofs s activ 10%e nse incof busineo v i denditurn accr

ine toe v iol a r i s edeen

prE 300,000)e c iatioi x e

d ase ) , by

m ova

a bl prx pee d o st r a ig

w fs e ngt h o d .ied and r

t i o n

of Tpr

t ifrecl ec quisegue pri f f e r e n

t mh o s e v

76

de s .e s.

r p r i s ee a rht

n tee thos

mude.e hicl

r vict r a igr itingincipl anye .g

n

i

nee

s

)

• : 

  25%

a nceo to Fi r s t Y

ntifd in prc

t toa r y

(e

n o t

in eht le hicll o wc ieudentr ar no

is

s (o r m0 %

n se n

t 100%

s 4% s

o r

a 4% annualw hich inclr

vw : 25% wunting

, subje co

y

incl ther t aining

r t aining

d inge f st r a ig

r y (o tow n alo r 4e o

w ho

ibl dov e l o pmding

n c e .

i n g d o m buill

ig £3,000 fg i b

le fnditur

w i llr r e ctli th accoo r taxn te

we f

ru l e s too r

e en te

Kr iting

and knon

s coptable

d ase )

edtriale

e on the

achine

e n

t and md toal Buil

) ara nce wr

e eli

a n c

e ex pe

ch and det ur

l ant, ms t r i cteE s ants al l o wa

r d anceic tax

a l l o wductibl

Un it

• - 

  I ndusz o neall o w
• - 

  Pe quipmannualReSMA llow

Capitalr e s e arA g r i cul

li n

en.a.P a ted o

wnL a n d , good

P r o v i sioaccoare acce

s p e c ifpro v i s i o n s f

disde

ef of

e

dn

r d ingn,i e dc

t tof thet

icabls ou mf f asm u m

nog n is

u l tipl subjeal.

e ar

r d e d inntse coa x e s .

applf yi nimn oa xi

i s

l e g i s l atio77

a r y accos e

t.pre c iatio ms ) and 1,5

s e tsr m

r os , 25% of T

f as de be

re co

be a mr itte bas

wl inen d ) pateu f f e r an

a l u e

rn oe if

and ve oe to l e s s than 5e ar

6 yo r as

a nces).

d fe than nod at thec

t toe la &y s

l a ws

ar l i f e isu s t

be numa y bei a t e d over mht-d ing

bale e n

5-e arhteubje t o barks

n in vi r e ctioductibl

ra te

e f uln munts,t r a igf buil

edme s s the

b e t we igive us

pre c iate

n thee , se s me p r ec

nsn se oader a l D

i x e

d byr and nature wns

deeeme unle duction l y de

s fcliningl inei f us

ductioss accoe

d od us or ds o

s di l l , tre re ne

Gr a l o

e ctoht-

1,5 (s ) , 2 ( o r e than 6 ys ar

de be

basctes . ex pe

e ari n cas

e arD

lue i

pre c iabl

thee neies

so r e intea

nd (o o d w

f

y oe .da teade

isingntl

a ble tr

ns arndea b l

e.

pel o

w fundee par on l y

ade

ard sa m tr

nsllowt al tor s

erw

f thes if

b l i g a tiot inder

e ae noe n tse ar

ex iss a

e actioa y mi

th undee

ductible d fom e o

3 y

h e r e os i one . Palt w

d to

e wv e n tsu tur

t e' s fi n g d o mr ovio v i s i o n s are deo n den

in tim be carrie inco

nde

Kductibl

pasn tityedi f i c pr a l prm e s ar

r a l l y Nitatio

s mayt fu turx te

, es .

D e ductibl

f r o man eUn it

S p ecG e ne

N o

n de

s c herules.

ainse ar

G e ne

N o l i m

L o sse

ag1 ycease

. g . ri s k

sm e n t

e inr it ish ichith

er t ain

e ee n tala blv e r

24 w

bt iss , o in w and ce

l o wn s arbts

ic78

e s u lt of deo r o

d

up by to

ductibl

as re n tse .

n v i r o nme alo l v e n t, pay

• s. 

  A inso c e e d ing ov e r dueductioo nthsl d f

duct de

r publ banks

s

e n s es, etries .

r is deo nthso nths

e he de

s or backey toa y m

stancebtourt pr

o nthsi ngivable to

t ien s o

o v i s i o n s

de su its

e induso v i s i o n s arm p aniee . Pductible

se

c um dee ces s i bl

d pare de

ic prb l

e exp l a w

activi

c pru l

if on co

fo l l o w

d -o r r

t pot r a tios. co

ru l e s appl

ductiblductible ara bl

ndingx trcifnds

r t ain cirubtfpee theitteo r > 6 < 12 mo r > 12 < 18 mo r > 18 < 24 m

r mi nisantiea ncem e s

ar 6 yail

t avo r e than 6 mo nths no

s o ciateo

n deo

n dep to

y

eenn.

e t wr s hip.f isl atio

or dinar

r e d

bw nee os e tsf inf

e as

a

bl tim

of as

unt o

r a n s f er

m mo

n o

ice

e tax

arm e

at then

pr accos .

s

in coi tio take ra te

e d i f t r a d

e t ga insn.

s s incod to

m p anieineisatiop e c ial

A llow

coCapitalbusr e alT h e acquis

adjusteN o s

of

nducty

i x e d

u l e s nor coe fe d t o

n rs ae o or dinarf

e oibl

j u st79

pt in casee ie asf tangs adn.

o l idatiox ceh era bl tim

o

nsns ee ticel atio

coh ere in natur

e

tax ye ars i

arm e

at the

n prr 2f inf.

thei ctio

n d ws s

a insn.itio ra te

c

t to re str

s , a

icant changine gs s incoe l d fount o

c ialineisatio

t s

h accop e c ial

e r g erh e acquiso s

 

terf ff fff

fo rfr

s afe t oe t o

ee onse o

e ase s ther s

rgee oa jo

t s one aro v i depet bef the

ty.y

ssei t h i n

t prx changnditio ex chang oft bth cas capitalt ric t

d aard and s

ard and s

f therec t i v e .

such timed our o ms

sd the

changso a m

n bords me

cannonduct o

xee n t w

t mo r wo r we s nor Dim e r g e r s "

ar

h an er t ain coed or by

on the

i s p ospt frioue n te

ga

f o r e and 3 y

n ves doa rrie s . Ir e

is alr co

ber i e d f

ga insr i e d f

ga insi s i o n s oe rge

o s t "o ug ce

e

cm p any

x e mple m

s ref o r e them p any

thee o

car car l a we et t o vat im

y md thrc

t toising

f e r r e d until ar

e d

coh ar

a y bf s

i n g d o mlie f for fi

e ar

y t h e Mntl

Kr re of rei1 yubjee

de then or s aru b j ec nor e d bef a coard if natur

eds

(em p any sh a r ess mn s .rec t i v e a

r Dii s i o n s .o r w

e

ar sa l e )abln l y ben l y be

ainst capitalainst capital cored b

e que

nsm p l i she ga ins ar

e s . Se s are w

vis i on

t r i butioe h o l deK has

s incurr s hip oe

in the

i n s t a

x snditioe rgener i e d f

Un it

R o

ll ove

c o n d i t i on4 ytheT ax

C an o

agCan oagU K

m e r g e r s and div

c oveCoaccos h arcapitalshart h

e nDil i quidatingdisshargacoT h e U

Mand divL o sse

o wcarchang

r

i las ss e

s a

al d asi mc

t toe if

los

t s on heubjeo cro

e t y p e i n of thessibl

ssei thin can

r i o d

to or as

an d s

to

d a pes e tsa l , sp l i e

s ta mh o u l d be

xee

nt w

i nancey l o s s e sy l o s s ifs i on

asi viutrrec t i v e

r a n s f ern l y poi t s

e s t mf Fnd the, d neso ap

t h

e sm p any

is od ( alh e Di

, t h

e t

t e r oe taxn

• s. 

  of t

lie f for fie

inv or dinar

or dinar

f

r ex te

i nanciale rgersg i mee r g err b e d com p any gr ante

r ren oe ase asns ar

nditio reera t i o n s of

w o rkE .

m

abso co is

e d toe a r .ee nt, f.

opmeu l i ng

i t

e a l i n

gf ar ed

s ds e s t os ofeen ga in

p t i o n .nds

s i

on up oi e s

a der o u pb erre t wu

ll c

or.

x e mi de

r

ovi midiar

ubs of gs t a

x lose m

ma r d s or e d

b capitall l ei t h f

pee

d.

llow fus

• : 

  div

u t w

i f i c p tyaimeenw n wr a n s f er

r a n s f er

p ecn as such.

gr o up is clef" ae t w, dor o up, no

r

e ta

n

i

e

s:m p anie

b l

e b

o sr 75% s

i n g d o m on theing

p relir e d bards t o t h

e t

s in a go mpa xa

nt co

r

e nl

idatio: A

ns

upws s e t s ar i s

es ar

e t

.e r g e r s K

nso

ere aedr 51% ondinge n

t bea y s ).e as asti c c

t h mpee "grouh erm p anie

mer e s i de

trade

ThwiUn it

N o conditio

Coe i thedetre a tm

Thb e t r a n s f erg r o

up (sid e w

Wcoor los

DoN o

n-

r e c e i v ed

ats,h e inse s ,ns.82

t edt be

i l ls s ) .n t i n gainst

fr e e the

a n d s

of unit.

egah efe -pr

inet eibls oal. T

t 90%e r a li t h i n

fp a n y .x pird a t edu m

taxm p any

i scalm p aniea y no

ggri

t:

T md a t edard and

t r y onsu s t be

o r t uge h o l dinge asm p anie

x genn t l y wo m

r i o d eo n s olindsi n i m

e r g e r w beld t a k e over

o r wr p o r atio

me d bush

ou t o t h

e grax

te mh

e grou

of f i ce

at ll l correr i o d

o of f s e t ag

e ci des . < 25% tax

r b e d

co i n t h

e hn e ff co no

r

e a l i m

Miniss ee nts

's shar pea

mn pe t h

e cd cop t i f me ar

r i e d fs

• : 

  div

e r g e r tot i

on as os op ao f itsggrega

o n s olir e be

oui nge prh e a

h

e c

absolue

vaa xao sse

d d i t i

on thee are m

p a n i e s

of tincipale

nt in Pu s t bey . Aa t

e Ta

e me ctli e

s cu then l y be

f i l i ate exemo r

2 y

d that the mp a n y s

re s t r u ctur

f thei v e

y re quir

o mi r prm p any

y mo r p orano l idatio car

s e s fromt be

ookn.e gra bl of t

ef: tm p any

a

m fo l l o w

t s

of tu l d the

n.m p

anie5

%

s

• : 

  i n ai nga nag compior to

n can o by t h

e s

ns

ainst af

tratef theo m

thed tax5 %

r or ofis sho

i n

g cts oe biss i o n by

fo r fh e ce theidiarr indirt t o C

f cor n ed co

ar

e 9

of 25% f

e

nt com e and l

h a n 6e m

r l y i ng

and unde

tax

l d ing tax

i thho

r p o r ate

f o r wco

fith83

e s t ics .

Cen

od wme ar

I Rn l y if do

ive d ar

e ductioo v i dee 95%o r

2 y

r

l

y

i

ngs

:e d

it o

re ce

cr. D prnding

is

ceive d ar

r s hip, f

f undem p anie

nds

i deu l l tax tre a ty

o urr r e s po

co re cew ne

nt co tax at s

f thends

f 60% oe .i

th f

U : div

d by 25% o

l d ingi t on .i de

ne nt

quityntnd ifsa tiond if

ti tatioh ich

bt/ei

de re g a r d to

r m (a rise d

i

de fi x e d

all y

lue.e ifr a d e

e noi mf w

fo r

s t on-r e s i de

i ngr tee )

et vauntings of tt

he ra tiont 75% par nor m

ductiblptabl

a rki f t s ars o

a la nce co

d acco al fi x e

d de a nod

as a divth havy capital no

e deu r p o sel o wquityd

as a divh e r e is1 is

ptee s the nor ise

bt/en-r e s i de

r iseth. T

of

1:

e s arh e p

i n g d o mo s t

or mnse n

t and gh e r e isn al

e hiclr e isactem ' s l e ngr e s t and o

deall y acce a thinl

a noacte

Ke s .e . Tr e s t paid do ther ma tio

y toe .

edx ped for t

e

r of cr a l l y acce

s t er t ainmpre c iatioe r vs

u g h thet at ar of inte

s e ngeed, inte re char

re charm ' s l e ng

no ra teuding1 is nor e s t paid byidiar

t at ar but a rptabl

Un it

L o w

G e ne

principl

Moi n c u rreE n teductibl

def o r depasexc£ 12,000.A l tho

r a tiocan beit istheinclo f

1:I n tes ubscan benor a tioacce

s

d

s ou tf

fee r sT E84

s i ons t

on

e thos P of an

ns

r y inge hicl noe s .

s e ng

Provin conseed

st m ra tiod byitatio

carp a n ye r vi tioL imndition

t is

lue.r e t a xx pe

e

co theo m

nse s , 20% o

s e ngd e fo r

fo r pasquity gr ante

t y . ( et value a

h i c h exce f f i ciee r

e r agd toh

e cx pe acquisc r a f t and any

a rka nce wbt/e

l o ansd par l o

an co

. Co

e t

va of ttantiate

l o

w theth)a r e h old

i r mrk

and av re l a te

l e g a l

en and pass , theo s t ofu l e , dee l a tey if

e .e ifi neubs lim i t a t i on

n ale ng

n.

j e c t i v es

d t o maI F Oe s , ilaft and airr s

l s o a

e s ; f t h

e cs .r a l re s p e c

t tont r

m 's l

d os aone ned t o t h e sht appl

B. QUANTITATIVE ANALYSIS OF THE EFFECTIVE LEVELS OF COMPANY TAXATIONIN MEMBER STATES

INTRODUCTION

This part of the study presents effective corporate tax rates on domestic and transnational investments in the 15 EU Member States taking the tax systems in operation in 1999. Moreover, in view of the structure and magnitude of the German tax reform approved in 2000, the analysis developed in this section also takes account of the effects of this reform as of the 1

st January 2001.

Effective tax rates are tax rates which take into account not only the statutory corporate tax rates but also other aspects of the tax systems which determine the amount of tax effectively paid. In other words, they take into account the tax base and the manner (if any) in which corporate and personal tax systems are integrated.

Systems of taxing profits are, however, far too complex to be encompassed fully in the methodologies developed so far in order to calculate effective tax rates. A number of the special features of individual tax systems thus have to be ignored, for instance special sectoral incentives. However, the main features of the national tax systems are captured in the calculations presented in this report. The methodologies used in this study build, on the one hand, on a revised and extended methodology from the so-called King & Fullerton approach, set out by Devereux and Griffith (1998) and, on the other, on the "European

Tax Analyser model", set out by the University of Mannheim and ZEW (1999).

The purpose of the quantitative analysis developed in this part of the study is twofold. First, it gives summary measures of the overall relative incentive (or disincentive) provided by each country's tax law to undertake various types of investment at home or in the other EU countries. This provides an indication not only of the general pattern of incentives to investment that are attributable to Member States' tax law, but also of the extent to which taxation in each country discriminates in favour or against inward and outward investment. Second, it identifies the most important tax drivers influencing the effective tax rate, that is the weight of each of the most important elements of a tax regime. From the pure point of view of economic efficiency, decisions related to the location and the organisation of an investment in terms of choices of assets and sources of finances should not be driven by tax considerations. In order to highlight the policy issues involved in reducing potential tax-induced economic distortions to the allocation of resources, this part of the study investigates the contribution of particular features of taxation to the lack of neutrality in taxation systems.

mandate given to the Commission by the Member States. The main focus of this report is corporation tax including its interaction with some elements of personal tax.

In this part of the report, the cost of capital, marginal effective tax rates and average effective tax rates are computed for different types of domestic and transnational investments in the manufacturing sector in each Member State. The contribution of various features of Member States' tax laws to the lack of neutrality of the tax regimes is assessed by means of a series of simulations. Some cases of the effective tax burden of SMEs as well as some cases of tax planning are analysed separately.

This quantitative analysis relies heavily on the report "The effective levels of company taxation in the Member States of the EU", produced for the Commission by the Institute for Fiscal studies, the University of Mannheim and the Centre for European Economic Research (ZEW) of Mannheim. The calculations presented in the boxes "Tax Analyser" are based on the report "Computing the Effective Average Tax Burden for Germany, France, the Netherlands, the UK, Ireland and the USA using the "European Tax Analyser" Model" produced for the Commission by the University of Mannheim.

METHODOLOGY

Existing approaches to measure companies' effective tax burden: backward and forward-

looking concepts

The existing approaches to measure the effective tax burden are based on two types of analysis implying either backward-looking concepts or, alternatively, forward-looking concepts. Both approaches have their respective advantages and disadvantages and can lead to different quantitative results. Even if the results of the application of different methodologies are not directly comparable, the existence of tax induced distortions seems to be confirmed by a variety of studies regardless of the particular approach adopted. Nevertheless, the size of the observed differences as well as the relative situations of countries do vary depending on the methodology applied.

Backward-looking approaches

One approach to measure the effective tax burden in policy-making is based on aggregated data from existing firms. As this looks at the capital stock, profits or other relevant data accumulated in the past it is called a backward-looking approach. By referring to the observation of ex-post data, it measures "actual" rather than "hypothetical" tax rates. Within this framework, one can distinguish between approaches based on firm-specific data or on aggregated economic data.

indicators are imprecise indicators of the investment incentives of taxation. But, they do permit an assessment of effective actual tax burdens by firm size, sector or industry, which may be useful in addressing equity concerns.

Measures for the tax burden using aggregate economic data from national accounts are computed as a percentage of domestic corporate taxes (in general only corporate income tax) relative to various income measures, such as aggregated domestic corporate profits or the corporate operating surplus. Although these formula are mathematically correct, it is hazardous to make an international comparison of corporate tax rates on the base of aggregated economic data. On the one hand, the methods and definition of the National Accounting Systems differ between countries and, on the other, these data are not sufficiently developed to distinguish different sources of taxation. Moreover, as is the case for tax rates based on firm-specific data, tax rates based on macroeconomic data sometimes tend to show significant fluctuations from one year to another due to business cycle effects.

Forward-looking approaches

Consequently, the most commonly used indicators for analysing the impact of taxation on investment behaviour are based on forward-looking concepts and involve calculating and comparing the effective tax burden for hypothetical future investment projects over the assumed life of the project or, alternatively, the effective tax burden for hypothetical future model firm behaviours, using the statutory features of the tax regimes.

These approaches permit international comparisons and are especially tailored to "isolate" the effects of taxation thus providing an indication on the general pattern of incentives to investment that are attributable to different national tax laws. It is worth noting that the results of the application of these approaches rely on the assumptions underlying the definition of the hypothetical investment in terms of assets and financing and of the future firm behaviour in terms of total cash receipts and expenses, assets and liabilities over time. Moreover, these approaches do not take into account in the computation all the features of a tax system.

The results produced by the application of these approaches summarise and quantify the essential features of the tax system in a relatively straightforward manner. They provide an estimate of the discrimination of Member States' tax law between various forms of investment and different sources of financing as well as of the discrimination in favour or against inward and outward investment. They also identify the most important tax drivers influencing the effective tax burden. Therefore, these approaches can illustrate the distortive effect on the allocation of resources of a tax system for typical investments or typical firm behaviour, which may be useful in assessing the investment incentives of taxation and addressing efficiency concerns.

rates on capital income, and in particular by the OECD (1991), the Ruding Committee (1992), and Baker & McKenzie (1999)

36.

The computation of the effective corporate tax rate builds on two different methodologies which involve calculating the effective tax burden either for a hypothetical future investment project or, alternatively, for a hypothetical model firm behaviour. In technical terms, the analysis relies on a revised and extended methodology from the so-called King & Fullerton approach, set out by Devereux and Griffith (1998) and on the "European Tax Analyser" model, set out by the University of Mannheim and ZEW (1999). The main computations are based on the hypothetical future investment approach and they are supplemented by the "European Tax Analyser" model, which utilises the model firm approach.

Considering that each methodology is based on different hypothesis and restrictions, it has been considered useful to compare the results of these two different approaches in order to test them and, possibly, to confirm the general trend arising from the computations.

It is worth noting, however, that the analysis of a hypothetical investment is more complete, in the sense that it covers a broader range of cases for all the European Union Member States. For technical reasons linked to the availability of data and the nature of the model, the "European Tax Analyser" only covers a limited numbers of countries and cases.

The main body of the quantitative study thus relies on the application of the analysis of a hypothetical investment and is complemented, where relevant and possible, by results arising from the computation of the behaviour of a hypothetical model firm.

The taxation of a hypothetical investment

The King and Fullerton approach (reviewed by Devereux and Griffith) is based on the assumptions that all markets, especially production factors markets, are competitive and the production function has the usual properties, notably constant return to scale. In this situation, the decision to invest and locate somewhere is influenced only by capital taxation, not by taxes or contributions on other factors such as wages, energy etc. and the incidence of these other elements of the tax system is borne by other agents (see Annex A for a more detailed description of the methodology).

another country (transnational case). A marginal investment is one whose expected rate of return is just sufficient to convince the investors that the project is worthwhile. This minimum rate of return is widely referred to as the "break-even" rate of return. Given a post-tax rate of return required by the company's shareholder (for instance on interest earned in some alternative use of the capital), it is possible to use the tax code to compute the pre-tax rate of return of the hypothetical investment, that would be required in order to obtain the minimum post-tax rate of return. This is known as the cost of capital.

A company that is contemplating a new investment project has, on the one hand, to compute the overall cost of the asset, taking into account not just the initial outlay, but also any reduction of that outlay due to tax relief received as a result of the investment. On the other hand, the company must also calculate the after-tax returns that it expects the investment to generate in the future. The company would undertake the investment provided the present value of the after-tax profits from the investment is greater than the initial cost of the asset minus the present value of any tax relief. Hence, the principal impact of taxation on investment is through the cost of capital. The difference between the cost of capital and the required post-tax rate of return (expressed as a percentage of the cost of capital) is known as the effective marginal tax rate (EMTR) that is, the rate applied to a marginal investment.

For example, if the minimum rate of return required by the company's shareholder is 5% and the company must earn 6.67% before tax (the cost of capital) in order to pay this 5% rate of return to the investor, then the effective marginal tax rate is 25% (6.67%-5%/6.67%). The difference between 6.67% and 5% represents the impact of taxation on the cost of capital.

This approach is based on the presumption that companies will undertake all investment projects which earn at least the required rate of return. For a given required post-tax rate of return, the more severe the tax system, the higher is the cost of capital, i.e. the required pre-tax rate of return, and hence the less likely that any specific investment project will be undertaken. In comparing such investments in alternative locations, the underlying economic model would predict that, ceteris paribus, locations with a higher cost of capital or EMTR would have less investment.

• b) 

  Effective Average Tax Rate

The current study goes beyond this approach, however, to also consider the effective "average" tax rates (EATR) on various forms of incremental investment which are more profitable then the marginal investment explained above. The rationale for doing so is that often a company that has taken the decision to undertake a specific profitable investment has to choose between two or more mutually exclusive locations. Examples include the location decision of multinationals in choosing a site for one new factory, and the choice of investment projects in the presence of binding financial constraints. In this case, the impact of taxation on the choice is likely to be measured by the proportion of total income taken in tax in each location. The measure used in this study is computed as the net present value of tax revenue expressed as a proportion of the net present value of the income stream (excluding the initial cost of the investment). The literature commonly defines the effective average tax rate as the effective tax burden held by an infra-marginal (average) investment as opposed to the effective marginal tax rate, which is the effective tax burden held by a marginal investment.

Box 3:

Properties of the measure of effective average tax rate used in the computation

The properties of this measure have been explored by Devereux and Griffith (1999).

One attractive feature of the measure is that, in the absence of personal taxes, the EATR for marginal investments is identical to the EMTR. At the other extreme, for extremely profitable investments, the EATR tends to the statutory tax rate. An example of this is given in Figure 1, which presents a range of values of the EATR for different levels of profitability for Belgium. The figure shows the average EATR for the investments analysed in this study.

Figure 1 Effective Average Tax Rate and Profitability in Belgium

-

only corporation taxes

-average across all forms of investments,

45%

40%Statutory

Tax Rate

35%

30%

EATR25%

20%EMTR

15%

10%

5%

0%

0%20%40%60%80%100%

Level of Profitability ( p )

• c) 

  Hypotheses and assumptions

Estimates of the effective tax rates on domestic and transnational investments in the 15 EU Member States are presented as at June 30, 1999. In the transnational case, the analysis is extended to the case of investors located in the USA and in Canada. Calculations consider primarily corporation tax in each country, but also include the effects of personal income taxation of dividends, interest and capital gains.

Several assumptions need to be made in order to define the hypothetical investment project analysed, and the economic conditions under which it is assumed to take place. Besides these, the exercise is limited to parameters of the various tax regimes which can be captured in the context of the analysis of a hypothetical investment project. Thus, as in every study of this kind, the hypothetical investments analysed are rather simple manufacturing sector investments, and a number of detailed features of actual tax systems cannot be incorporated in the model as for instance different kind of provisions in the different Member States. The fact that the analysis is limited to the manufacturing sector is due to the impossibility to quantify, in the framework of the model the number of different specific provisions applying to the service sector across the EU Member States (e.g. the special provisions applying to the financial service sector). Moreover, this approach does not, for methodological reasons, take into consideration all the relevant features linked to the existence and functioning of different tax systems. It does not, for instance, quantify the effects on the tax burden of the possibility of consolidating profits and losses throughout the EU because, by definition, it only takes into account investments which make profits. The quantification of compliance costs is also impossible.

The computation is also based on the hypothesis that all taxes due are paid and therefore that the results are not affected by different levels of tax enforcement. In fact, there is no reason to believe, nor is there any empirical evidence, that possible shortcomings in the enforcement of tax laws have a significant impact on the location of business activities within the EU.

The assumptions and parameters underlying the computation are given in Annex B. Sensitivity analysis investigates the impact of the assumptions and of some elements of tax systems on the results.

It is worth noting that, for the sake of comparison, the definitions of investment and of the economic variables underlying the computations are the same for all countries considered. The purpose of the analysis is to understand how taxation influences the profitability of the same hypothetical investment in different countries and not to give a picture of the actual economic situation for each country.

purpose to test and, possibly, confirm the general picture arising from the application of the "hypothetical investment" approach. (The methodological framework and the hypotheses and assumptions of the Tax Analyser model are given in Annexes G and H).

The calculations are based on an industry-specific mix of assets and liabilities taking as a base case a typical medium-sized manufacturing company. Based on this (in general, existing) capital stock, the future pre-tax profits are derived on the basis of estimates for the future cash receipts and cash expenses associated with this initial capital stock. In order to determine the post-tax profits the tax liabilities are derived by taking into account the tax bases according to the national rules and then applying the national tax rates.

This approach does not need to characterise optimal investment behaviour but it relies heavily on the particular characteristics of the model firm, in particular the initial capital stock and the expected development of the capital stock over the simulation period.

• a) 

  Average effective tax rates

The tax effects of infra-marginal investments, i.e. of investments that are more profitable than the marginal investment, are central to this model and the taxation of an existing capital stock is analysed. Consequently, this model only computes effective average tax rates which measure the effective tax burden of projects that earn more than the capital costs.

The effective average tax rate is expressed by the difference between the pre-tax and the post-tax return of the capital invested in the corporation divided by the pre-tax return.

• b) 

  Hypothesis and assumptions

Estimates of the effective average tax rates for Germany, France, the Netherlands, the UK, Ireland and the USA are presented as at 1999. The calculations consider primarily corporation tax in each country, but also include the interaction of corporate and personal income taxes, the individual income tax rates including surcharges and capital taxes at the shareholder level. The effective average tax rate is derived by simulating the development of a medium-sized manufacturing company over a ten year period.

Box 4

Tax provisions taken into account in the models

Hypothetical investmentTax Analyser

• - 

  Depreciation (methods and tax period for all considered assets);- Depreciation (methods and tax period for all considered assets, extraordinary depreciation);
• - 

  Inventory valuation (production costs, lifo, fifo, and weighted average);- Inventory valuation (production costs, lifo, fifo and weighted average, inflation reserves);
• - 

  Development costs (immediate expenses or capitalisation);
• - 

  Taxation capital gains (roll-over relief, inflation adjustment, special tax rates);
• - 

  Employee pension schemes (deductibility of pension costs, contributions to pension funds, book reserves);
• - 

  Provisions for bad debts;
• - 

  Elimination and mitigation of double taxation of foreign source of income (exemption, foreign tax credit, deduction of foreign taxes);
• - 

  Elimination and mitigation of double taxation of foreign source of income (exemption, foreign tax credit, deduction of foreign taxes);
• - 

  Loss relief
• - 

  Investment incentives (extraordinary depreciation, special tax credits, special tax incentives). These are considered only in the sensitivity analysis (section 5).

That said, as the German corporate tax reform approved in 2000 addresses in both a quantitatively and qualitatively significant manner all the main relevant characteristics of the German corporate tax system, the analysis developed in this section also takes account of the effects of this reform.

Therefore, an additional separate set of effective tax rate data for Germany as at the 1st January 2001

have been computed, as well as additional comparative tables which take into account the 2001 situation for Germany and the 1999 situation for the other countries (see annexes E and J). Where relevant for the analysis of the effects of effective tax rate differentials, the inclusion of the new German situation in theEU context is commented on in this section. All the tax reforms introduced in other Member States are less significant as far as corporate taxes are concerned, both quantitatively and qualitatively.

The simulations of the harmonisation of particular features of taxation are based on the 2001 situation for Germany and on the 1999 situation for the other countries. As such policy simulations refer to a hypothetical future situation, the use of a consistent basis is less relevant here and indeed to ignore the German reform would be highly misleading.

The effective tax burden of SMEs and the tax planning cases are analysed on the basis of 1999 data for

all countries considered.

It is worth mentioning that France introduced a tax reform in 2000 aiming at abolishing the surcharges on corporation tax by the year 2003. Due to the particular structure of these surcharges, which are determined partly by the amount of wages and salaries, and the period over which the changes will be implemented, this reform has not been modelled. Section 10 of this part of the Study presents an updated computation of the effective tax rates on domestic investments for all Member States, taking into account the tax regimes of 2001. This permits an analysis of the impact of national reforms of the corporate tax regimes on the effective tax burdens.

THE TAXATION OF DOMESTIC INVESTMENTS

This section considers the influence of domestic tax regimes on the organisation of companies' investments and the way in which national tax codes can affect the international competitiveness of resident companies and, under certain assumptions, the location choice of multinationals.

Box 5:

The role of personal taxation

The influence of personal taxation on company investment behaviour depends on the functioning of international capital markets and, in particular on the extent to which international portfolio capital is mobile. If companies can only raise money domestically, then changes in the personal tax treatment of investment income will alter company behaviour. Instead, if companies are able to finance their investments on the international capital markets, the influence of personal taxation on investment income varies according to the degree of integration of capital markets.

If this market is so integrated that the world interest rate is unaffected by the domestic amount of saving, personal taxes do not and should not affect the investment behaviour of companies. In fact, a personal tax on all forms of interest income will result in a lower post-tax return to savers; consequently they will save less. But assuming that domestic saving is small relative to the world supply of saving, the world interest rate will be unaffected, and so the investment decisions of the domestic corporate sector will be unaffected. In contrast, taxes on corporate income generated in a particular country will affect corporate behaviour, regardless of how the project is financed. In such a case, due to capital mobility, personal taxes in small open economies like the individual EU Member States do not affect investment decisions of companies. From this point of view the taxation of shareholders or more generally, the taxation of suppliers of finance, would not be relevant for a comparison of business tax burdens.

This conclusion, however, depends on the assumption that internationally mobile portfolio capital always exists. But this assumption could be questioned on the grounds that all companies raise at least some money domestically and small and medium sized companies may even have no access to international capital markets. The literature is not unanimous on whether the assumption of perfect international capital market mobility is pertinent for all type of economic agents.

Moreover, structural differences between national tax systems are mainly caused by the differing corporation tax systems and the different ways in which the corporation tax and income tax interact. For this reason, the level of taxation not only for retained, but also for distributed profits differs among countries.

A practical difficulty also arises in seeking to take into account personal taxes. That is, the company may have many shareholders, facing different rates of tax from each other. Which set of personal taxes should a company take account of in these circumstances? Economic theory suggests that a company should act in the interest of the "marginal" shareholder that is, the shareholder who is just indifferent between owning and not owning the company's share. Unfortunately, in practice, it could be impossible to identify "the" marginal shareholder.

In order to consider all these arguments, the present study shows a separate analysis of the impact of personal tax in the domestic case. The central case analysed takes into account only corporate taxes on the hypothesis that the company does not know the identity of the marginal shareholder. Then, in order to provide a comprehensive analysis of the impact of Member States' national tax systems on investment and financing decisions, personal taxes are added. However, as far as the effective average tax rate is concerned, since its main focus in this analysis is on the choice of location, an implicit underlying assumption for this case is that economies are open to flows of mobile capital. In this situation it is very difficult for firms to allow for the tax positions of their shareholders. Nevertheless in the framework of the "Tax Analyser" model, personal taxation is considered and average effective tax rates including the effects of personal taxation are separately presented.

The influence of domestic tax regimes on the organisation of companies' investment by assets

and sources of finance

Domestic tax regimes can influence the organisation of companies' domestic investments by creating incentives both as to how to finance the investment and the overall mix of assets. In fact, different forms of investment or different sources of financing may face very different tax treatments. Such variations constitute a potential source of distortion in the allocation of resources and may therefore impact overall efficiency. If the impact of differences in tax treatment favours one particular form of investment or financing, then the economic activity may not be organised in the most efficient economic way. Although these differences may be secondary to the main focus of this section, which is the impact of taxation on the incentives to locate investments, it is useful to have an indication of the effects of tax regimes on the organisation of investments in the EU as a starting point.

Relevant economic measures: cost of capital, EMTR and EATR averaged across the EU

The first case analysed is the simplest case in which there are no personal taxes. Separate investments in five different assets are considered: intangibles (e.g. purchase of a patent), industrial buildings, machinery, financial assets and inventories. In presenting averages over different forms of investment, these assets are weighted equally. Three sources of finance for investment in each asset are separately

considered: retained earnings, new equity and debt. The weights used are taken from OECD (1991): retained earnings 55%, new equity 10% and debt 35%. Thus, calculations are made for 15 different types

of investments.

Tables 1 and 2 present the cost of capital, the effective marginal tax rate and the effective average tax rate for each type of investment averaged across the 15 Member States. This is an unweighted average. That is, it does not take into account differences in the size of each country (or any other factor). As such, it gives an indication of the average effect of tax regimes in the EU and it is not an attempt to measure the "average" taxes in Europe, where the size of countries and hence the numbers of investments facing each specific tax regime would need to be taken into account.

Tables 1 and 2 capture the extent to which corporate taxation in the EU affects the incentives to undertake particular kinds of investments by responding to the following questions, respectively for Table 1 and Table 2.

• A) 

  The case of a marginal investment

TABLE 1Cost of Capital and Effective Marginal Tax Rate

-

average across all 15 EU member states

-only corporation taxes

Cost of CapitalIntangibles IndustrialMachinery FinancialInventoriesMean

(upper line)BuildingsAssets

EMTR

(lower line)

%

Retained Earnings6.68.06.78.67.97.6

20.035.223.339.935.532.6

New Equity6.47.86.68.47.87.4

18.534.222.039.334.631.6

Debt3.34.43.54.94.34.1

-67.3-22.2-50.9-3.8-16.8-24.6

Mean5.46.75.67.36.76.3

3.623.38.329.824.120.2

Note. Each number in the Table is an unweighted average over the equivalent number for each member state. This is true for both the cost of capital and the EMTR. For a specific investment in a specific country, the EMTR is the percentage difference between the equivalent cost of capital and the post-tax required rate of return of 5%. For example, a cost of capital of 7.5% generates an EMTR of (7.5-5)/7.5=33.3%. However, taking an average of the costs of capital, and a separate average of the

EMTRs implies that the average EMTRs presented in the table are not precisely the percentage difference between the

average cost of capital and 5%.

Generally, according to Table 1, there is considerable variation in the tax treatment of different forms of investment within the EU and, therefore, the EU tax regimes effectively seem to create incentives as to how to organise investment in the EU. Annex C (country tables) shows that there is a remarkably similarity between countries in the pattern of tax incentives for domestic investments even if the range of values across countries gives an indication of differences between EU Member States in their treatment of specific forms of investment (see sections 4.2 and 4.3).

Second, in the absence of personal taxation, there is almost no difference in the cost of financing the investment by giving up one unit of dividend income as opposed to contributing one extra unit of new equity

37.

When considering the assets, considerable variation in the average treatment can be observed too.

Financial assets are the most heavily taxed. In fact, financial assets are assumed not to depreciate and hence not to benefit from any allowance. Any income generated from the asset is generally taxed at the full statutory tax rate. Moreover, this rate is applied to the nominal return defined as the real interest rate plus inflation rate (set at 2% for each country in this analysis), rather than to the real return, and for this reason the effective marginal tax rate exceeds the statutory tax rate. Hence, the higher the inflation rate, the higher the EMTR. In the case of financial assets financed by debt, the fact that nominal interest payments are deductible from tax generally compensates for the fact that the nominal interest receipt is taxable

• 38. 

  In such a situation, the value of both tax and economic parameters plays no role.

In general, the cost of the other assets can be offset against taxable profit over a period of time. Typically, the rate at which the cost can be offset is related to the economic depreciation rate of the assets. For a given true economic depreciation rate, the more quickly the cost can be set against tax, the more valuable the allowance and hence the lower the effective marginal tax rate. The EMTR thus reflects the difference between the true economic rate of depreciation and the rate of allowance permitted

in the tax code.

Differences between the remaining four assets therefore reflect not only the generosity of the tax systems with respect to the allowance rates for the four assets, but also the assumptions made about the true rate of economic depreciation. However, even allowing for this dependence, significant differences seem to persist within the EU. In general, industrial buildings and inventories have the highest cost of capital and effective marginal tax rate, while intangibles and machinery are rather lower.

• B) 

  The case of a profitable (infra-marginal) investment

Table 2 presents estimates of the effective average tax rates for each of the same 15 investments analysed above under the assumption that the pre-tax real rate of return is 20%.

Table 2Effective Average Tax Rate - average across all the 15 member states

• - 

  only corporation taxes

EATRIntangibles IndustrialInventoriesMean

%BuildingsMachinery Financial

Assets

Ret Earnings30.635.131.035.634.933.5

New Equity30.234.730.735.234.533.1

Debt

20.023.820.723.623.522.3

Mean26.831.127.431.430.929.5

In this case, it is important to remember that, as explained in Box 3, the EATR varies according to the expected level of profitability of the investments. In particular, in the absence of personal taxes, theEATR is identical to the EMTR for marginal investments, and it rises when the profitability rises because allowances against the cost of the investment become relatively less important when the cost of the investment becomes smaller relative to the returns.

Since the assumed real rate of return of 20% is not high enough to mean that allowances and deductions are too small to have much impact, the pattern of the EATR in Table 2 bears some resemblance to that of the EMTR in Table 1.

variation in the tax position of different shareholders, which may make it impossible for a company to maximise the post-tax earnings of all shareholders. Table 3 presents the cost of capital and the effective marginal tax rates, averaged across the 15 EU Member States, for the 15 hypothetical investments in the case where companies aim to maximise the wealth of top-rated qualified shareholders, taking into account their personal tax liabilities on the hypothesis that these are known by the company

• 39. 

  A

qualified shareholder is a shareholder who holds a substantial part of the shares of the company. Three personal taxes are introduced in this section: on interest received, on dividend income and on capital gains.

This table, therefore, captures the extent to which corporate taxation and these three forms of personal taxation affect the incentives to undertake the particular forms of investment considered in this study, assuming that the investments will not raise extra-profits.

For the theoretical reasons explained in Box 5 the analysis of the impact of personal taxation is restricted here to the case of a marginal investment.

Table 3Cost of Capital and Effective Marginal Tax Rate

• - 

  average across all 15 EU Member States

• - 

  top personal tax rate, qualified shareholder

Cost of capitalIntangibles IndustrialMachinery FinancialInventoriesMean

(upper line)BuildingsAssets

EMTR

(lower line)

Retained Earnings4.15.24.35.64.94.8

51.061.453.363.460.359.3

New Equity4.75.94.96.35.65.5

56.764.558.868.165.564.0

Debt3.54.63.84.94.34.2

30.944.634.952.447.544.7

Mean3.95.14.25.44.84.7

Table 3, shows, that when personal taxation is taken into account, the differences observed in Table 1 still exist, even if a different treatment of the sources of finance can be observed. But the most striking feature of this table is that the taxation of the investment backflows in the hand of the shareholders considerably reduces the EU average cost of capital and increases by more than twice the effective marginal tax rates.

The most important reason for the decrease of the cost of capital is the impact of the personal tax on interest. In fact, the post-tax rate of return required by the shareholder depends on the post-tax rate of return of an alternative financial investment. Assuming the alternative to be lending, then any tax on interest -the return on lending- reduces the post-tax return to lending. Consequently a lower post-tax rate of return is required from equity investment. In fact, on average, investment financed by retained earnings and debt have a cost of capital less than the real interest rate of 5%. Of the five assets, only industrial buildings and financial assets have an average cost of capital above 5%.

Personal taxes do not generally affect the cost of capital for investments financed by retained earnings. This is because they affect the net cost of the investment in the exactly same way as the net return to the investment. Suppose, for example, that the tax rate on a dividend payment is 30%. And suppose that a company finances the purchase of an asset costing 100 euros by reducing dividends. The net cost to the shareholder is therefore 70 euros. Suppose also that the investment generates a gross rate of return of 10%, being worth 110 euros after one period (ignoring taxes). When this amount is distributed as a dividend, it generates post-tax income of 77 for the shareholder. But this represents a post-tax rate of return to the shareholder of 10% -the same as the pre-tax rate of return. The impact of the dividend tax is negated by the fact that it affects both the net cost of the investment and the net return. There are much smaller effects on the cost of capital for investment financed by debt.

The underlying reason why the average EMTRs in Table 3 are considerably higher than those in table 1 is that the alternative opportunity open to each individual is to lend an equivalent sum. Therefore theEMTR compares the cost of capital with the post-tax rate of return on lending rather than to the real interest rate. Since this post-tax rate of return is lower as a result of taxes on interest received, the effective marginal tax rates are higher.

that ignore personal taxes (perhaps because they do not know the identity of their marginal investor) and that are able to leave effective tax rates close to those of the host country, due to the provisions of international tax codes or to the tax planning activity of the company.

This section considers these questions by looking at the EU range of values for the 15 types of hypothetical investment considered earlier.

With regard to the location of investment, it should be noted again that the data arising from the application of the theories underlying this analysis give summary measures of the incentives (or disincentives) to undertake different types of investments and do not provide evidence of the impact of taxation on actual economic decisions. Box 6 presents a short survey of the empirical studies which have attempted to measure the impact of tax differentials on actual location choice in recent years.

Box 6:

Links between business taxation and companies' location decisions

The mandate given to the Commission by Member States requests an assessment of the effects of differences in the EU Member States' effective tax burdens on the location of economic activity and investments. The methodologies applied in the current study assess the relative incentives (or disincentives) provided by each country's tax law to undertake various types of investment at home or in the other EU countries. To what extent taxation has an impact on actual investment decisions depends, however on the extent to which tax incentives lead to changes in actual behaviour.

If taxation were the only element influencing location decisions, that is, ceteris paribus, differences in the effective tax burdens between countries would be the only factor determining location decisions, investment should be located in countries where taxation is lower. But taxation is only one of the elements affecting location decisions. Several differences arising from the macro and micro economic framework of each country contribute to determine the actual behaviour of companies.

The fact that differences in the effective tax burdens between countries persist, shows that the arbitrages are not perfect and that taxation is not the only element affecting location decisions.

(A) According to many authors, one implication of the hypothesis of tax competition amongst governments to attract business should logically be the convergence of the observed tax rates. Empirical evidence of such convergence is, however, not very strong.

Moreover, the relative convergence of statutory tax rates on companies' profit and on individual investors' income and capital gains that has been observed in the European Union does not, in itself, give much indication of the extent to which tax competition and effective differences amongst Member States in the treatment of such income remain.

In fact, the effect of differences in taxation on economic decisions depends on their marginal impact on the rates of return to investment, which, in turn, depends not only on apparent tax rates, but also on rules determining the tax base. So, the convergence of effective marginal tax rates should be evident over a period of time.

Unfortunately, it is not possible to study this possible convergence due to the lack of appropriate data based on forward-looking methods covering a long enough period of time.

(B) One obvious direction for testing the consequences of differentials in national business taxation on location decisions is foreign direct investment, where, by definition, capital is internationally mobile.

The empirical relevance of tax considerations in investment decisions has been mostly studied by looking at the investment location decisions in multinational corporations. A number of empirical studies show that tax considerations are relevant in investment decisions. Nevertheless, the size of the correlations varies according to the specific methodology applied.

Devereux and Griffith (1998), using individual firm activity data of US multinationals investing in Europe (restricted to the UK, France and Germany) show that the choice of the location, conditional on the decision to produce abroad rather then to export, is driven by taxation and other cost-related factors.

Friedman, Gerlowsky and Silberman (1992) consider the establishment of new manufacturing plants of European and Japanese firms at the state level in the USA. They find that per capita state and local taxes are strong determinants for location.

Therefore, the fact that the existing literature has been rather deceptive in furnishing coherent results on the size of the impact of taxation on capital flows or location decisions seems mainly due to sample biases or data shortcomings. A way to properly deal with the difficult issue of the interaction between taxes on various factors would be to use a general-equilibrium framework allowing for imperfect competition on some markets. An alternative way of proceeding would be to confront net bilateral foreign direct investment flows with series indicators of average effective tax burden.

(C) A number of studies based on European data have focused on the hypothesis of tax competition amongst local governments within national economies. This restricted field has the advantage of more closely resembling the US context in which the original empirical work was initiated. In addition it also restricts the number of differences potentially interfering with business tax differentials in the firms' location decisions: for instance, within the national economy, social contributions are uniform. However, even in this environment of restricted differentiation amongst jurisdictions, they are potentially many factors influencing the location of firms, and taxation is only one of these.

Conventional wisdom currently rests on the apparently sound hypothesis that firms' location decisions, be it when a firm relocates or when it decides to open a new plant or office, are made according to many factors, that are treated in a hierarchical way. A firm first chooses the region -the so-called "macro location" decision- based on such factors as market for product, labour market conditions and labour costs, etc. Only then, in the so-called "micro- location" decision - i.e. when choosing the precise locality in which to settle- will local tax differentials influence the firm's choice (Jayet, 1993; Jayet and Wins, 1993; Conseil national des impôts (F), 1997, Madiès, 1997a and b; Houdebine and Schneider, 1997;

Paty, 2000).

To conclude, the empirical studies show, to different degrees, that there is a negative correlation between the size of taxation and location decisions. Nevertheless, most of the empirical studies suffer methodological weaknesses or are tailored to study just the effect of local business taxation. It is therefore difficult to have "the" quantitative measure of this impact even if the existence of such a relation is generally undisputed.

Because of the weaknesses of the existing methodologies and the severe limitation due to lack of available data, it is considered that none of the existing approach could be usefully adopted in the current analysis, without considerably extending the range of the work. Taxation has certainly an influence on the location of economic activity but it is very difficult to correctly isolate this influence.

Relevant economic measures: range of the cost of capital and EATR values across the EU

As in the previous analysis on the influence of domestic tax regimes on the organisation of companies' investments, the first case analysed in this section is the simplest case in which there are no personal taxes. The calculations are made for the same 15 different types of investment.

Tables 4 and 5 capture the range of values across EU Member States and therefore present the highest and lowest values observed when analysing separately individual EU Member States, respectively in the case of marginal investments and in the case of investments whose assumed rate of return is 20%. These ranges give some indication of the differences between EU Member States in their treatment of specific forms of investment. The individual situation of each EU Member State is presented in section 4.3

40.

• A) 

  The case of a marginal investment

TABLE 4Cost of Capital-

- maximum and minimum across the EU

- only corporation taxes

Maximum

MinimumIntangibles Industrial

BuildingsMachinery Financial

AssetsInventories

Retained Earnings8.410.19.812.610.5

3.45.14.35.85.5

New Equity8.410.19.810.49.0

3.45.14.35.85.5

Debt4.86.35.75.75.0

1.62.22.23.03.6

• B) 

  The case of a highly profitable (infra-marginal) investment

TABLE 5Effective Average Tax Rate - maximum and minimum across the EU

- only corporation taxes

MaximumIntangibles IndustrialMachinery FinancialInventories

MinimumBuildingsAssets

Retained Earnings40.646.344.354.348.2

10.117.09.411.011.0

New Equity40.245.444.347.942.0

10.117.09.411.011.0

Debt26.931.532.234.828.8

6.613.56.02.57.5

This table shows that there is even greater variation in the effective average tax rates across Member States than there is in the cost of capital. To take one example, the EATR on investment in machines financed by retained earnings and new equity ranges from 9.4 to 44.3%.

In contrast to the case of a marginal investment, differences in the EATR are less likely to generate a competitive advantage (or disadvantage). In fact, the EATR is based on the assumption that the investment will generate a rate of return in excess of the minimum required. If a company in this situation (resident in one country) finds itself undercut by its competitors, then there is nothing to prevent it from responding by reducing its prices, and lowering its pre-tax rate of return.

The introduction of personal taxation

When personal taxation is introduced into the analysis of tax differentials in the domestic case, it is worth noting that differentials in Member States' effective tax burden are still relevant in affecting the relative competitiveness of domestic operators, as was the case for the analysis of the cost of capital, but they are no longer relevant for the choice of location. In fact, in this case the objective of companies is to maximise the wealth of their domestic shareholders. This could be incompatible with the hypothesis that the company aims at leaving effective tax rates close to those of the host country by means of tax planning

Table 6 shows the range of values of the cost of capital across the 15 EU Member States, using as an example a qualified shareholder taxed at the highest personal tax rate. Once again, the impact of personal taxation is restricted to the case of a marginal investment.

Table 6Cost of Capital-

- maximum and minimum across the EU

- Top personal tax rate, qualified shareholder

MaximumIntangibles IndustrialMachinery FinancialInventories

MinimumBuildingsAssets

Retained Earnings6.57.16.49.67.1

0.1-0.30.2-0.8

New Equity8.98.98.810.08.8

0.20.4 0.5-0.5

Debt4.86.46.25.75.0

2.02.52.63.03.6

The position of the EU Member States

One of the main purposes of this section is to analyse the differences in the effective tax burden borne byEU companies located in different Member States and thus appreciate the possible effects of such differentials on competitiveness and investment decisions. Under the hypothesis and assumptions already mentioned, Tables 7 and 8 present country data where only corporate taxation is taken into account : this includes the statutory tax rates, the surcharges and the local taxes. (The list of such taxes for each Member State is given in Annex B). Section 4.3.2 will present data including the forms of personal taxation already considered above.

Relevant economic measures: cost of capital, EMTR and EATR by Member States

• A) 

  The case of a marginal investment

Table 7 shows the cost of capital and the EMTR for the level of corporation and examines more closely the dispersion across the EU and the relation between the effective tax burden and the national tax rate

on

profits.

Debt

-25.0-42.9-13.6-11.1-8.7-56.2-47.13.8-38.9-35.1-21.9-28.2-21.9-39.5-25.0

TRquityNew e33.337.533.330.544.435.534.215.210.035.135.136.735.125.435.1

E M

earnings

Retained33.337.533.330.544.448.434.215.210.035.135.136.735.125.435.1

Debt4.03.54.44.54.63.23.45.23.63.74.13.94.14.34.8

quityNew e7.58.07.57.29.07.67.65.95.57.77.77.97.76.77.7

earnings

Retained7.58.07.57.29.09.77.65.95.57.77.77.97.76.77.7

AL

I T

APInventories6.36.77.16.87.47.97.45.55.06.56.96.56.46.66.9

C

T OF

Assets

C OS

Financial7.38.07.16.88.010.05.15.57.77.77.47.77.46.66.9

Machinery5.95.35.45.68.45.86.15.23.85.35.95.25.45.05.6

r y

l

ts an

en

m n repres

fi n a n

ce colu

d 35% debt. Each

an

ity

equ

e whts.

e ig

e w

a m

i n g s , 10% ne s

th

ith

ed earn

e n t , w

55% retainv e s t m110

in

of

h t s of

e igpes

w

ith

s all 15 ty

Table 7 shows that there is considerable dispersion across the EU. Focussing first on the overall average, it is notable that 7 Member States have an average cost of capital between 6.3% and 6.5%. Three others (Greece and Finland and the UK) are very close to this range, leaving five significantly outside: France and Germany have an average of 7.5% and 7.3% respectively, and Sweden, Ireland and Italy have an average of 5.8%, 5.7% and 4.8% respectively.

This EU wide spread cannot be explained by one feature of the national tax systems alone, but it can be observed that Germany had in 1999 the highest statutory tax rate on profit, and France by far the highest non-profit taxes (local taxes on corporations). On the other hand, statutory tax rates in Sweden, in Ireland and for certain categories of income (see below) in Italy, are by comparison the lowest. Even Finland, which has one of the lowest statutory tax rates, shows a relatively low cost of capital.

All the EU Member States, except Ireland, have an EMTR lower than the overall corporate tax rate.

The fact that Ireland has an EMTR higher than the corporate tax rate is fundamentally due to the relatively high (relative to the corporate tax rate of 10% applying to the manufacturing sector) real estate tax rate (1.58%) applicable to industrial buildings (see Table 3 of Appendix B). The influence of the real estate tax in the EMTR for Ireland is shown in Table 7, where the cost of capital for industrial buildings is 6.8 and therefore the related EMTR is 26.8%. This influences the overall mean.

In order to appreciate the dispersion across Member States and the relation between overall nominal profit rates and EMTR it is useful to look at the situation of the specific types of investment both by type of assets and by source of finance.

Table 7 clearly confirms that, from the point of view of the corporation, the most tax-efficient way of financing is debt. The major reason is that deduction of nominal interest payments from the corporation tax base significantly reduces the effective tax burden on investments financed by debt. The effect of interest deduction is high in countries where the corporation tax rates are higher. Furthermore, in certain countries, debt-financed investments are subsidised if, relative to other countries, assets receive "accelerated" depreciation as is the case for example in Belgium, Greece, Italy and Sweden.

Financing through new equity and retained earnings is disadvantageous, as no deduction from the taxable base for the corresponding payments (dividends) is allowed. The national effective tax burden for both forms of financing almost equal the tax rate on profit. Given the close relation of the effective tax burden on new equity and retained earnings to the tax rates on profits in most of the EU Member States, it can be inferred that "normal" accounting rules for profit computation, in so far as they are considered in the model, in general do not have a great impact on the effective tax burden and on the ranking of the countries, as they only result in "timing differences". Rather, it is likely that the different tax rates on profit explain most of the differences in EMTR between countries.

With regard to assets, in general, intangibles and machinery are taxed quite generously. The only exception is Greece, where buildings are depreciated for tax purposes over a period of ten years and where financial assets benefit from a very favourable tax rate. Reasons for the general disadvantageous treatment of buildings include the comparatively long lifetimes for tax purposes and the obligation to use straight line depreciation (except in Finland and Sweden).

As far as non depreciable assets are concerned, inventories are, in general, more heavily taxed. It is very difficult to draw general conclusions concerning the relative treatment of depreciable and non depreciable assets. In fact, as already pointed out when analysing Table 1, estimates of the effective tax burden could be sensitive to the assumptions made for true economic depreciation. Section 5 of this section examines the sensitivity of the results commented on in this section to the assumptions made.

• B) 

  The case of a profitable (infra-marginal) investment

Table 8 presents a summary of the effective average tax rates for each Member States for investments whose pre-tax real rate of return is 20%

Table 8Effective Average Tax Rate by country -by asset, source of finance and overall

-only corporation taxes

EFFECTIVE AVERAGE TAX RATES

t y

quibt

w eDe

CountryCorporateO v erallMean

tax rates (1)IndustrialB u ildingsA ssets

IntangiblesMachineryFinancialInventoriesR e tained

earnings

Ne

Austria34.0029.828.629.228.433.229.933.933.922.3

Belgium40.1734.530.736.131.039.235.339.139.125.8

Denmark32.0028.821.334.725.331.231.232.332.322.1

Finland28.0025.524.824.823.127.327.328.828.819.3

France40.0037.530.640.640.139.037.142.142.128.8

Germany52.3539.133.939.034.946.840.846.140.127.7

Greece40.0029.635.530.433.411.637.134.434.420.8

Ireland10.0010.58.915.88.29.89.811.711.78.2

Italy41.2529.824.929.827.436.131.131.831.826.1

For all countries bar one (see below), the EATRs are higher than the effective marginal tax rates but still lower than the overall nominal profit tax rate. The effective tax burden rises when the profitability rises because allowances against the cost of the investments become relatively less important when the cost of the investment becomes smaller relative to the returns.

It is worth noting that the only country which has an EATR lower than the EMTR is Ireland. For this country the average tax burden of the investment decreases when profitability rises and marginal investments are relatively more highly taxed than profitable investments. In fact, as was shown when analysing Table 7, for Ireland the marginal tax rate is higher than the national profit tax rate because of the relatively high real estate tax rate applicable to industrial buildings. When profit rises, the weight of this tax becomes relatively less important because the tax is levied not on profit but on the value of the industrial buildings. As a result, when profit rises, the effective tax burden for industrial building diminishes.

Table 8 shows that, when profitability is set at 20% the EATR for industrial buildings is 15.8%. This is lower than the EMTR for industrial buildings (26.8%) but still higher than the corporate profit tax rate. This influences the overall mean.

There appears to be rather more dispersion in the overall average EATR for each Member State than there is in the equivalent EMTR. The Irish average rate is only 10.5%. Other rates range from 22.9% in Sweden to 39.1% in Germany.

The other main conclusions of the previous section are equally applicable in the case of a profitable investment. The link between effective tax burden and national profit tax rate is now even stronger and it is noteworthy that the ranking of Member States is almost the same when considering the EATR and national profit rates. The only exceptions are Italy and Greece. For Italy, again, the existence of "dual income" tax system tends to reduce its EATR in comparison to the national profit rate when the profitability is set at 20%. At the same time, this system plays now a less important role than in the case of a marginal investment. Italy has the lowest average cost of capital, but only the seventh lowest EATR. For Greece, the higher difference between its national corporate rate and average effective tax burden seems to depend on the generous capital allowances granted for depreciable assets and on the very favourable treatment of financial assets, which benefit from reduced rates.

Box 1 Tax Analyser:

Effective Average Tax Rates (corporation level) across 5 EU Member States and the USA.

This box gives measures of the effective tax burden of companies as computed by the application of the model "Tax Analyser", which was described in section 3. (The hypothesis and restrictions of this model are given in Annex H).

For the sake of international comparability and in order to isolate the effects of taxation, the comparisons of the effective average tax rates are made under the assumption that the weights of assets and liabilities of the model firms are identical in all countries. The Tax Analyser model refers, as a base case, to a typical medium-sized German manufacturing company with data taken from published German statistics. (Annex H explains the structure of the balance sheet of the model firm).

The effective average tax rates for the scenario in which only taxes at the level of the corporation are taken into account, over a calculation period of ten years, are shown in the following table.

TABLE AEffective Average Tax Rate across 5 EU Member States and the USA

• - 

  Only corporation taxes

FDIRLNLUKEU-5USA

Average

EATR - (corporation)39.732.88.324.021.025.229.7

relative in %100.0100.0100.0100.0100.0100.0100.0

Corporation tax and surcharges54.377.077.298.588.679.180.1

As was the case for the analysis of a hypothetical investment, this table shows that there is considerable variation in the EATR, with a range of 31.4 percentage points between the highest and lowest rate. In the case of a hypothetical investment the range was 28.6 percentage point. Therefore, the application of the

Tax Analyser model confirms the magnitude of the variation inside the EU.

If we compare the results of Table 8 with the results of Table A, the most striking result is that, with the exception of France and Germany, the ranking of the countries from the highest to the lowest EATR is the same for each model and this ranking closely follows the country ranking according to statutory tax rates. The difference in the relative position of France, which is the highest EATR country here and the second in the ranking in Table 8 can be explained by the fact that, in contrast to the hypothetical investment case, a higher portion of non-profit taxes is included in the Tax Analyser model. That is, the three "employer taxes" and the personal expenses forming part of the base of the "taxe professionnelle" are included in the Tax Analyser model.

The differences between EU countries' EATR are self-evidently due to different national tax systems, tax bases and tax rates. But, it is important to stress that the Tax Analyser model takes into account almost all the elements that affect the tax base. Only special incentives are excluded.

The results of Table A reveal that there is considerable variation in the structure of the national tax systems. In all countries, profit taxes have the highest impact on the EATR. By contrast, the impact of non-profit taxes on the EATR is low on average. The only exception is France where the weight of non- profit taxes is 45.7%.

When the corporation tax bases are compared for a typical medium sized company, Figure A below reveals a large variation between the countries. This considerable range is mainly caused by the low tax base in France, as notably a consequence of the relatively high social contributions, and, to a lesser extent in Germany. These differences are, however, smaller than would be expected from the findings in many qualitative analyses of these tax bases.

FIGURE AComparison of overall corporation tax bases of 5 EU Member States and theUSA

The analysis of the tax bases shows that all tax regimes are designed as more or less integrated systems. This means that there is a particular relationship between the tax rate and the tax base. High rate countries tend to compensate it through a reduced tax base and vice versa. Nevertheless, the cases of France and Germany, which have the highest EATR, and the highest overall corporation rates, clearly illustrate that higher tax rates more than compensate for reduced tax bases. On the other hand, despite having the highest corporation tax base, Ireland benefits greatly from its comparatively very low rate, which results in by far the lowest effective corporation tax burden. In general, tax rate differentials more than compensate for differences in the tax base.

Therefore, the effects of the different tax bases, even if these differences partly compensate tax rate differential, have therefore only a comparatively minor impact on EATR.

The introduction of personal taxation

The introduction of personal taxation in the case of a marginal investment increases the effective tax burden significantly. This is caused by the taxation of the investment backflows in the hand of the shareholders. But personal taxes do more than simply increase the wedge driven between the initial return on the investment and the return finally received by the financier. They alter the whole structure of incentives to use one form of finance rather than another.

Table 9 shows the range of values of the cost of capital and the EMTR in each Member State for a qualified shareholder taxed at the highest personal tax rate.

Table 9Cost of Capital and EMTR by country

• - 

  by asset, source of finance and overall

• - 

  top personal tax rate, qualified shareholder

OverallCost of Capital

Mean

t y

quibt

a pitalssets

w eDe

CountryCost ofCEMTR

IndustrialB u ildingse tained

IntangiblesMachineryFinancialA

InventoriesRearnings

Ne

Austria5.843.55.45.65.46.75.76.57.64.1

Belgium5.730.24.66.34.77.15.86.58.13.7

Denmark4.178.42.45.93.34.54.53.84.64.6

Finland5.460.25.25.34.85.95.96.14.64.6

France5.372.53.56.46.65.44.75.17.84.9

Germany5.479.54.05.34.37.75.66.84.13.5

Greece5.027.85.64.15.04.16.05.95.63.5

Ireland4.156.43.85.33.73.93.93.16.05.2

Italy5.118.83.14.94.08.05.46.05.63.5

Luxembourg4.170.33.44.53.55.14.04.14.74.0

Netherlands2.895.72.33.12.63.42.92.02.24.4

personal taxes were not taken into account. In both cases the low cost of capital can be traced to the difference in the tax treatment between dividends and interest received by the shareholders. In the Netherlands, for example, qualified shareholders can pay tax on interest at a rate of 60% on nominal interest receipts, but pay tax on dividend income at a rate of only 25%. The high tax rate on interest income means that the shareholder demands a much lower rate of return on their investment in equity, which is reflected in the cost of capital.

However, the ranking of countries by the EMTR is very different. This is because the EMTR is defined as the percentage difference between the pre-tax rate of return earned on the investment and the post tax rate of return earned by the shareholder- that is, the difference expressed as a proportion of the cost of capital. Given a very low cost of capital in Spain, it is quite likely that the EMTR will appear very high. This is compounded in the case of Spain by the fact that the high tax rate on interest income implies that the shareholder is willing to accept a very low post tax rate of return on equity investment; in this case actually less than zero. This results in an EMTR over 100%. The fact that countries with the same cost of capital nevertheless have a different EMTR is because the post-tax rate of return differs between countries.

As far as the whole structure of incentives to use one form of finance over another is concerned, it is no longer true that debt is always the form of finance which minimises the effective tax rate. For a majority of Member States debt is still the most favoured form of finance and new equity the less efficient form, with retained earnings in between. The reasons for the different relative treatment of finance are the different corporation tax systems, the taxation of capital gains from the disposal of shares from the shareholders and, in certain relatively generous cases, final withholding taxes on interest income. These three tax-drivers have an interactive impact on the relative position of the forms of finance.

For example, the reason for the lower tax burden on retained earnings when personal taxation is considered, is simply that capital gains are either not taxed at the personal level, (whereas debt or dividends are) or else that capital gains are not taxed on accrual but only on realisation, whereas with dividends and debt any tax due cannot be postponed in this way.

Box 2 Tax Analyser:

Effective Average Tax Rates (overall level: corporation and shareholders) across 5 EU Member States

and the USA

Compared to the tax burden at the level of the corporation (see Box 1 Tax Analyser), the overall tax burden including personal taxes is higher, with no changes in the country ranking.

TABLE BEffective Average Tax Rate across 5 EU Member States and the USA

• - 

  Corporate and personal taxes

FDIRLNLUKEU-5

AverageUSA

EATR - (corporation and shareholder)

48.837.417.232.025.632.232.0

The range of the EATR is also higher. This suggests that there is even more variation in personal taxes than in corporate taxes.

Besides corporate taxes, these differences at the overall level come from the interaction of corporate and personal income taxes, the individual income tax rates including surcharges and capital taxes at the shareholder level.

The impact of the German tax reform

The analysis so far has been based on the tax regimes which were in place in 1999. For the purpose of comparison across countries it is important to choose a particular time at which to make a comparison. Clearly, tax reforms in the EU can and do occur almost continuously, and so it is impossible for any set of results to reflect the long-term position. However, the German tax reform, which came into effect on January 1, 2001, is such a substantial and important reform affecting at the same time the system, the taxable base and the rate, that it is useful to investigate how the main conclusions of the analysis are affected by it. Therefore, this section investigates the impact of the German tax reform on the main results of the previous analysis.

Box 7:

Description of the major tax changes in Germany

With effect from January 1, 2001, the German tax reform has changed the corporation tax system, reduced corporation and personal income tax rates and broadened the tax base.

• - 

  Corporation tax system : The full imputation system that has been in force since 1977 has been abolished and instead a

shareholder relief system has been introduced. Under the new system (which is similar to the system in Luxembourg), only one half of the dividends received by a private shareholder are subject to personal income tax. At the same time, all deductions connected with dividend income from the income tax base are halved. However, other elements of private capital income such as interest receipts are still taxed at the full rate. The abolition of the (full) imputation system follows an international trend. After the German tax reform only five EU Member States remain that apply an imputation system: Finland, France, Italy, Portugal and Spain. The vast majority of the Member States now utilise shareholder relief systems.

• - 

  Corporation tax rates : The changes in the corporation tax rate cover both the structure and the level of the tax rate. The

split-rate that distinguished between retained (40%) and distributed profits (30%) has been abolished and a single uniform tax rate of 25% has been introduced. Although the 25% corporation tax rate is the second lowest of the countries considered in this report (and within the EU), the solidarity levy of 5.5% and the trade tax with an average rate of 17.56% remain. This has reduced the national tax rate on retained earnings from 52.35% to 39.3%. Although this is a significant reduction, the statutory tax rate is still high by EU standards. Only France, at 40%, has a higher tax rate.

Relevant economic measures: cost of capital and EATR before and after the reform

The effects of the reform both on the cost of capital and the EATR for domestic investments in the case in which there are no personal taxes are summarised in Tables 10 and 11.

• A) 

  The case of a marginal investment

Table 10 presents the cost of capital for domestic investment in Germany both before the reform (based on the 1999 tax regime) and after the reform has been implemented. For each of the 15 types of investment, the upper number represents the case before the reform, and the number below represents the cost of capital after the reform.

Table 10Germany Cost of Capital before and after reform

• - 

  only domestic investment

• - 

  only corporation taxes

Before reform

IntangiblesIndustrial Buildings

MachineryFinancial

Mean

(upper line)AssetsInventories

After reform

(lower line)

%

Retained Earnings7.89.88.012.610.59.7

6.68.47.49.58.28.0

5.87.66.110.48.27.6

With respect to the taxation of different types of assets, investment in all types of assets benefits from the tax rate reduction. However, only investment in buildings and in machinery suffers from the reduction in depreciation allowances. Overall, given the assumptions made here, the average cost of capital across assets is generally reduced; however, it rises for investment in machinery.

It is worth comparing the relative position of Germany after the tax reform with its position before the tax reform. The overall average cost of capital for Germany before the reform was 7.3% . This is the second highest in the EU after France. After the reform the overall average cost of capital for Germany falls to 6.8%. Given the economic assumptions used in the computation, then, the average cost of capital for domestic investment in Germany does fall, but not by enough to change its ranking in the EU. The reason could be the overall national tax rate, which at 39.3% is still the second highest in the EU.

• B) 

  The case of a highly profitable (infra marginal) investment

Table 11 presents the impact of the tax reform on the EATR for domestic investment in the absence of personal taxes. As noted elsewhere, the EATR depends more closely on the statutory tax rate than does the cost of capital. As a result, the EATR on retained earnings falls substantially - on average across the 5 assets from 46.1% to 38.7%. The EATR on investment financed by new equity also falls - unlike the cost of capital for such investment- reflecting the lower statutory corporation taxation tax rate for distributions. Finally, on average the EATR for investment financed by debt is almost unaffected, although there are differences across assets.

Table 11 Germany EATR before and after reform

- only domestic investment

- only corporation taxes

Before reform

(upper line)IntangiblesIndustrial

Buildings

MachineryFinancial

AssetsInventoriesMean

After reform

(lower line)

%

Retained Earnings40.646.341.254.348.246.1

34.439.936.643.239.338.7

New Equity34.940.135.847.941.940.1

34.439.936.643.239.338.7

Debt23.127.224.834.828.827.7

23.928.626.131.727.927.6

33.939.034.946.840.839.1

Mean

30.835.932.939.235.334.8

The overall average impact of the German tax reform on EATR is to reduce it from 39.1% to 34.8%. Comparing Germany to other Member States (shown in Table 8), pre-reform, Germany had the highestEATR, ahead of France, (37.5%) and Belgium (34.5%). The effect of the reform is to shift Germany's ranking in terms of the EATR from the highest to the second. As with the cost of capital, the reform appears to have some impact on the effective tax rates faced by domestic German companies, but it has little effect on Germany's overall position relative to other EU Member States.

The introduction of personal taxation

Table 12 presents the cost of capital when the firm is owned by a qualified shareholder taxed at the highest personal tax rate. The addition of personal taxes has several effects. First, the reduction in the tax rate on interest income raises the post-tax return available on lending by the shareholder, which in turn raises the post-tax required return on equity. This tends to raise the cost of capital. Second, the abolition of the imputation system tends to raise the cost of capital for investment financed by new equity. However, this is offset by the reduction in the personal tax rate on dividend income. Third, the reduction in the effective capital gains tax rate tends to reduce the cost of capital, especially for equity financed investment. These effects are reflected in the results in Table 12.

Table 12Germany Cost of Capital before and after reform

• - 

  only domestic investment

• - 

  top personal tax rate, qualified shareholder

Before reform

IntangiblesIndustrial Buildings

MachineryFinancial

(upper line)AssetsInventoriesMean

After reform

(lower line)

%

Retained Earnings5.46.85.69.37.16.8

3.64.74.15.64.44.5

New Equity2.83.93.26.44.24.1

4.25.34.76.25.05.1

Debt2.33.32.65.73.63.5

3.74.84.25.74.54.6

Box 3 Tax Analyser:

The effects of the German tax reform

This Box highlights the impact of the German tax reform on the EATR both at the level of the corporation and overall (corporation and shareholder) based on the behaviour of a typical medium-sized German corporation in the manufacturing sector over the calculation period of ten years.

• A) 

  Corporation level

The following table compares the tax burdens for the base case corporation in the manufacturing sector in 1999 and after the tax reform (denoted as 2001).

TABLE CEffective Average Tax Rate in Germany before and after the reform

• - 

  only corporation taxes

GERGER

19992001

EATR (corporation)

• - 

  effective in %32.830.1

Relative in %100100

Corporation tax incl. Surcharges77.065.6

Trade tax22.333.4

Real property tax0.71.0

dividends are no longer accompanied by a tax credit, the corporation must increase its cash distribution in order to pay the same amount of dividends to the shareholders as before the reform.

TABLE DGerman Tax Reform

• - 

  increases and decreases attributed to different changes in taxation

• - 

  only corporation taxes

in % of

1999

EATR

Reduction of corporation tax rate to 25%-23.4

Reduction of straight-line depreciation for buildings1.3

Reduction of declining balance depreciation4.2

Abolition of full imputation system11.0

Altogether, the EATR reduction is too low to improve the relative position of Germany in the country ranking. Before the reform, German corporations' EATR (32.8%) was the second highest after France (39.7%). After the reform, Germany is still second highest (30.1%) now closely followed by the USA (29.7%). There is still a considerable gap to the Netherlands (24%) in fourth position.

• B) 

  Overall (corporation and shareholder) level

If we consider the overall level including personal taxes of the shareholders, the German tax reform reduces the effective tax burden significantly. In addition to the reduction of the marginal (and average) income tax rate, which affects both the dividend and interest income, this result can be attributed to the introduction of the new corporation tax system. According to the German method of shareholder relief only one half of the dividends is subject to personal income tax. The results in Table E, which combine corporate and personal taxes, show that the overall EATR falls from 37.4% in 1999 to 30.1% in 2005.

As a result Germany improves two positions in the country ranking to third lowest place behind Ireland (17.2%) and the UK (25.6%).

As a result, overall it seems reasonable to conclude that a typical medium-sized corporation with a low number of shareholders benefits from the German tax reform.

The results presented in this box are in line with the results presented in section 4.4. Despite the considerable reduction of the EATR, the German tax reform has only minor effects on the relative position of Germany in the country ranking.

Neutrality and distortion effects: concluding remarks from the domestic analysis

The analysis of the domestic case shows that European tax codes have an influence on the incentives to investment and the choice of the way of financing the investment. The analysis suggests that, in practically every situation analysed, on the one hand, tax systems tend to favour investment in intangibles and machinery and, on the other, debt is, by far, the most convenient source of financing. Moreover, the data tend to show that financing by retained earnings implies a lower cost than financing by new equity when personal taxation is taken into account. Thus, tax regimes as such are clearly not neutral to the extent that they tend to distort investment and financing decisions compared to a situation without taxation. From a purely economic point of view this has an impact on the efficiency of the allocation of resources within Member States and within the EU as a whole.

The previous analysis also suggests that there is considerable variation in the effective tax burden faced by investors resident in the different EU Member States. However, the Member States' tax codes tend to favour the same forms of investment by assets and sources of finance. Differences between the effective tax burden in the EU Member States can affect the competitiveness of companies competing in the same external markets and may affect, under certain conditions, the location choice of multinationals.

The wide spread within the EU cannot be explained by one feature of the national tax systems alone. However, the analysis presented above tends to show that the different overall nominal tax rates on profits (statutory tax rates, surcharges and local rates) can explain most of the differences of EMTR between countries. Therefore, although tax regimes are designed as more or less integrated systems (in general high tax rates on profit correlate with lower taxable bases and vice versa) tax rate differentials more than compensate for differences in the tax base. These conclusions are to be considered when discussing the compensatory effects of a broad tax base compared to a relatively low tax rate on the effective tax burden. The relative weight of rates in determining the effective tax burden of companies rises when the profitability of the investment rises. The policy simulations presented in section 7 will allow a better appreciation of the influence of particular features of taxation on the effective tax burden differentials between Member States.

tax, it is more difficult to achieve a neutral corporate tax system. More generally, a non-neutral tax regime may be justified from the point of view of economic efficiency in order to encourage or discourage certain activities insofar the activities in question render positive or negative side-effects.

On the other hand, concerns related to the non-neutrality of tax systems derive from the fact that taxation is one instrument for the creation of an appropriate business environment, and that its various other goals can often be furthered more effectively by other policy means. In this case the taxation system must not act as an obstacle to market efficiency. Moreover, a system characterised by large differences may often offer unintended opportunities for tax avoidance.

TESTING THE ASSUMPTIONS OF THE MODEL

Various assumptions have been made to generate the results given so far. This section examines the effects of altering these assumptions, thereby illustrating the sensitivity of the results to the assumptions made. In fact, the previous analysis is based on a set of very specific hypothetical investments under specific economic conditions. Therefore, the data presented in the previous sections should not be regarded as the universally valid values for the effective tax burden in different countries. It is therefore legitimate to ask to what extent the general results shown above depend on the assumptions and, in particular if changes in the parameters defining the investments or in the economic variables alter the general conclusions of the previous sections. In fact, even if there are no universally valid values, it is important to check whether it is possible to make generally valid statements regarding differences in the effective tax burden. In order to answer this question, this section conducts a sensitivity analysis which recalculates the cost of capital and each effective tax rate several times, each time varying the main parameters of the model. This is done for the average values across the EU and for the different countries separately. For the purposes of this exercise, only corporate taxes are considered.

Sensitivity of the average EU cost of capital and EATR to the changes in the economic model

or level of taxes

Tables 13 and 14 present an average across all Member States of the cost of capital and EATR for the different forms of investment, corresponding to the tables in the previous sections. The first row in each of these tables summarises the position in the base case, that is the averages shown respectively in Tables 1 and 2. The other rows consider separately the effects on the overall average of changing one parameter or set of parameters at a time. These changes involve the economic variables (real interest rate, rate of inflation and level of profitability of the investment), the weights assigned to the assets and the impact of local taxes and special investment incentives. There is no row numbered 4 in Table 13 (cost of capital), since that would refer to the change in profitability which is relevant only for theEATR.

Table 13Cost of Capital

• - 

  average across all the 15 EU Member States

• - 

  only corporation taxes

t y

Cost of capital (%)ssetsquibt

O v erallmean

IntangiblesIndustrialB u ildings

MachineryFinancialA

InventoriesR e tained

earningsw eDe

Ne

1 Base case6.35.46.75.67.36.77.67.44.1

2 Real interest rate: 10%12.511.013.111.213.913.414.614.48.6

3 Rate of inflation: 10%6.75.46.15.79.66.59.28.92.1

5 OECD/Ruding weighs6.1-6.75.6-6.77.37.23.9

6 BACH average weights6.04.76.04.96.55.98.07.84.4

7 Service sector weights6.14.65.84.86.45.88.38.14.6

8 Equal weights6.45.46.75.67.36.77.67.44.1

9 High level of local taxes6.55.47.35.77.46.77.87.64.2

10 Low level of local taxes6.25.46.25.57.26.67.47.24.0

11 Tax incentives for new investments5.14.15.72.27.16.56.36.13.0

Note. Each asset column represents an average across all three types of finance, with weights of 55% retained earnings, 10% new equity and 35% debt. Each finance column represents an unweighted average across all 5 assets. The overall average is an average across all 15 types of investment, with the same weights. Note also that the OECD report and the Ruding report considered only three assets: industrial buildings, machinery and inventories.

Table 14Effective Average Tax Rates

- average across all the 15 EU Member States

- only corporation taxes

t y

ssetsquibt

e tained

EATR (%)earningsw eDe

O v erall meanIndustrialB u ildings

IntangiblesMachineryFinancialA

InventoriesR

Ne

1 Base case29.526.831.127.431.430.933.533.122.3

2 Real interest rate: 10%24.920.327.020.728.628.031.731.012.5

3 Rate of inflation: 10%30.727.029.427.638.930.638.737.916.1

4 Level of Profitability: 40%31.630.432.630.731.732.533.533.328.0

5 OECD/Ruding weights29.2-31.127.4-30.933.032.722.2

6 BACH average weights28.024.728.925.329.028.634.434.023.0

7 Service sector weights28.925.329.225.829.529.236.135.724.0

8 Equal weights29.626.931.227.531.531.033.533.122.3

9 High level of local taxes30.527.333.628.231.931.434.434.023.3

10 Low level of local taxes28.626.328.926.530.830.332.532.121.4

11 Tax incentives for new investments25.722.828.017.030.530.329.429.019.0

The rise of the profitability of the investment from 20% to 40% increases the value of the EATR. This is in line with the analysis in Box 1. The higher the profitability of the investment, the closer is the EATR to the statutory profit tax rates. In general, this tends to reduce differences both between types of investments and types of financing and between countries. Figure 2 illustrates that for a level of profitability of 40% the individual countries EATR have already moved much closer to the national nominal profit tax rates. Italy and Greece show wider differences for the reasons explained in section 4.3 (commentary on Table 8). Annex D shows the distribution of the EATR in each EU Member State. It therefore presents the range of values of the EATR for all possible levels of profitability.

a t e

at 20%

at 40%y R

utor

E A TRE A TRS t at

y

an

rm

Ge

ly

yIta

u t or

t atm

lgiu

Be

and S

ce

ee

• & 

  40%

Gr

f 20%e

nc

Fra

ber States

e t u r n s org

ou

ax rmb

e txe

Lu

pral

rtug

age atPo

v erain

Sp

i v

e A

ds

Ef f ectan

erl

th

Changing the weights of assets (rows 5 to 8) has almost no effect on the average cost of capital or the average EATR. This suggests that the weights of the assets are not likely to be very sensitive for the purpose of the overall analysis.

Row 5 considers only the 3 assets and the respective weights considered by the OECD/Ruding studies: industrial buildings: 28%, machinery: 50% and inventories: 22%. The weight of the sources of finance is held constant at the levels of the base case scenario. These weights are taken from the OECD (1991) and Ruding (1992): retained earnings: 55%, new equity: 10% and debt: 35%.

Rows 6 and 7 use weights generated from accounting data (the BACH database). This approach uses accounting data for a large number of companies from different Member States. The numbers in the table below reflect the relative importance of each of the 5 assets, and each of the 3 sources of finance.

BACH weight (%)

n eryd

taine

Debt

IntangiblesIndustrialBuildingsMachiFinancialassets

InventoriesReEarnings

New Equity

Manufacturing2.914.319.345.617.924.720.554.8

Services3.212.311.954.28.421.619.459.0

The high weight in the BACH database attached to debt for both the manufacturing and services sector implies that the average cost of capital and EATR is lower for the 5 assets. The overall averages are lower mainly as a result of the high weight for debt.

Considering that the purpose of the quantitative analysis in this study is to "isolate" the impact of taxation on the same identical investment in each EU country, indicators based on the weight of each individual Member State have not been computed.

Rows 9 and 10 in each of the Tables relate to local taxes. In the analysis in the base case, "typical" values of local taxes were used: these are detailed in Appendix B. However, by their very nature, local taxes vary within a country. Hence two more extreme cases are considered - where local taxes are 50% higher than those in the base case and where they are half those in the base case

Depending on the country, the incentives considered might be extraordinary depreciation, special tax credits or special tax incentives. The precise incentive for each country is given in TABLE 12 of Appendix B.

Not surprisingly, such incentives do reduce the average cost of capital and the average tax burden without altering the relative position of the source of finance.

All in all, the previous analysis demonstrates that in most cases the parameters used in the model tend to have a little effect on the overall EU values of the cost of capital and the EATR. More importantly, changing the parameters does not alter the nature and the broad size of differences observed in section 4 as far as the overall EU values are concerned.

Impact of the sensitivity analysis on the relative position of Member States

The exact values of the effective tax burden of each Member State can, however, vary according to the definition of the investment and, as mentioned above, there is no universally valid value in one country. The purpose of this section is therefore to test if the ranking of Member States according to their average cost of capital and the average EATR arising from Tables 7 and 8, would be affected by changes in the assumptions used in the base case. The analysis is made for the 1999 situation.

Tables 15 and 16 show the ranking of Member States from 1 to 15, with the country with the highest cost of capital or EATR having the rank 1, and the lowest having the rank 15. Each of the cases presented in Tables 13 and 14 are presented; these are numbered 1 to 11 with 1 corresponding to the ranking relative to the base scenario. The column numbered 4 is blank in Table 15 (cost of capital), since that refers to the change in profitability which is relevant only for the EATR. The first column presents the average ranking over columns 1 to 11 and therefore gives the position of each Member State when all the different sensitivity analyses are taken into account.

Table 15 Ranking of Member States by Average Cost of Capital

• - 

  highest=1, lowest=15

• - 

  only corporation taxes

CountrySensitivity Analysis No.

1234567891011

n k

Ra

(Average)

Austria9989./.8101091076

Belgium57510./.7635584

Denmark48114./.44484101

Finland1111126./.101112111193

France2121./.11211113

Germany1212./.3212225

Greece12121012./.6151512121114

Ireland14141414./.13131114131410

Italy15151515./.15141415151515

Luxembourg1010911./.9771061212

Netherlands6565./.5567949

Portugal7638./.12986852

Spain8447./.118947311

Sweden13131313./.1412131314138

UK3373./.2353367

Table 16Ranking of Member States by Average EATR

• - 

  highest=1, lowest=15

• - 

  only corporation taxes

Sensitivity Analysis No.

Country

n k

Ra1234567891011

(Average)

Austria881081010998977

Belgium333334333332

Denmark101111118910101110116

Finland131313121213121313131311

France222221222218

Germany111112111121

Greece11109911314151011813

Ireland151515151515151215151515

Italy99413131166981014

Luxembourg556565445494

Netherlands678756777765

Portugal445447554543

Spain767678886659

Sweden141414141414131414141412

UK12121210912111112121210

Box 4 Tax Analyser:

Testing the importance of the assumptions

In order to test the robustness of the results presented in the previous boxes "Tax Analyser" the effects of alternative assumptions of the input data on the EATR are tested by means of sensitivity analysis covering data both at the level of the corporation and the level of the shareholder.

• 1) 

  Level of the corporation

The impact of changing assumptions is investigated by varying on the one hand the tangible fixed assets to total balance sheet ratio of the model firm and on the other, the weighting of the sources of finance.

• a) 

  Investment policy

This variation takes into account a change in the model firm's capital intensity. The quantity of tangible assets as a percentage of total assets is raised or reduced first by 10% and then by 20% in comparison with the base case. The ratio in the base case is 22.9%.

The results in figure B show that, with the exception of Germany and the Netherlands, the EATR increases with the capital intensity.

FIGURE BEffective Average Tax Rate across 5 EU Member States and the USA

• - 

  variation of tangible fixed assets to total balance sheet ratio

• - 

  only corporation taxes

FDIRLNLUKUSA

40

In Germany and the Netherlands the EATR decreases due to a shift from less generous rules for non- depreciable assets (i.e. financial assets) to a more generous capital allowances practice for depreciable assets. In the case of Ireland, the UK and the USA, it is above all the higher level of real property tax in the overall tax burden that overcompensates for the effects of the capital allowances and is therefore decisive in causing the increase. The EATR increase in France is noticeable. This can attributed to the structure of the French "taxe professionnelle". In fact, the basis of this tax includes tangible fixed assets but exempts intangibles and financial assets.

Although there is no change in the country ranking, it is worth noting that France and Germany have very similar EATR in the case of low capital intensity.

• b) 

  Structure of finance

In order to investigate the impact of changing assumptions regarding corporate financing on the EATR, the weighting of the sources of financing is gradually changed by increasing the equity to total capital ratio from 25% to 100%. This increase is accompanied by a reduction of the interest expenses for long term debt.

FIGURE CEffective Average Tax Rate across 5 EU Member States and the USA

-variation of equity to total capital ratio

- only corporation taxes

FDIRLNLUKUSA

50

45

40

This discrimination against equity financing at the level of the corporation is more evident in Germany and in the USA. Besides the high level of the corporation tax rate, this result is caused by the levying of other taxes that do not treat the payments for debt and equity capital equally. By contrast, the increase in the EATR is lower by far in Ireland and the UK. Since both countries apply the lowest corporation tax rate and levy no other taxes that discriminate against a particular source of finance, the reduction or saving of taxes due to the deductibility of interest is also the lowest.

Since the discrimination against equity financing in contrast to debt financing is common in all the tax systems that are under review, the EATRs neither cross nor converge. Therefore, differing assumptions about the debt to equity ratio do not change the country ranking of the base case. The level of variation within the sources of finance depends to a great extent on the level of the statutory corporation tax rate. Low tax rates tend to reduce such variations.

• 2) 

  Overall level (corporation and domestic shareholder)

Just as at the corporation level, the tax burden at the overall level, that is including personal taxation, is influenced by the assumptions about the economic data. Among the variables which have a large impact on the overall EATR, the distribution policy of the company and the sources of company financing provided by the shareholders are the most relevant.

• a) 

  Dividend policy

In order to work out the impact of changing distribution policy assumptions on the overall EATR, the corporation's distribution rate is gradually increased from zero (full retention of profits) to 100% (full distribution of profits).

Figure D below shows that if the profits are fully retained in the corporation the overall EATR is above all influenced by taxes at the level of the corporation. Due to the differences in the tax burden of the corporation, already explained in previous sections, the overall EATR is the highest in France and the lowest in Ireland.

The overall EATR increases with the rate of distribution in all countries. However Germany improves its relative position while the position of France, Ireland and the USA deteriorate. Ireland and the USA even lose one place in the ranking. This result can be attributed to a great extent to the different corporation tax systems and the degree of progression of income tax rates.

It is interesting to note that the EATRs converge with an increased rate of distribution. The combination of different levels of corporate and personal taxes and the interactions of these taxes with the different corporate tax systems thus reduces the dispersion of the EATR at the level of the shareholders to a great extent.

• b) 

  Equity to total capital ratio

In order to investigate the impact of changing assumptions about the financing of the corporation on theEATR at the shareholder level, it is assumed that the corporation is entirely financed by its shareholders with debt or equity capital (i.e. the corporation does not raise any funds from third parties). In the case of debt financing, the shareholders receive interest income from the loan granted to the corporation at a fixed rate. Whereas, in the case of equity financing, the profits are fully distributed to the shareholders. The weighting of the sources of finance are gradually changed by increasing the equity to total capital ratio from 25% to 100%. Figure E shows the results.

FIGURE E Effective Average Tax Rate across 5 EU Member States and the USA

- variation of equity to total capital ratio

- corporate and personal taxes

Altogether, the results show that taxation is not entirely neutral towards the financing of a corporation in

FDIRLNLUKUSA

of the countries at the overall level depends on the assumptions regarding the equity to total capital ratio.

In general, such dispersions are lowest in countries that either apply a full imputation system (e.g. Germany in 1999) or-irrespective of the corporation tax system-apply a low corporation tax rate (e.g. Ireland). In addition another prerequisite for neutrality towards company financing is an equal treatment of the dividends and interest payments with respect to income tax (and of shares and loans with respect to private property tax).

THE TAXATION OF TRANSNATIONAL INVESTMENTS

Section 4 examined the impact of taxation on the incentives to invest domestically. This section uses the same approach to consider the impact of taxation on the incentives to undertake transnational investments, i.e. to invest across country borders.

It describes how the framework used to analyse the domestic corporate tax systems of the EU Member States can be extended to cover investments located in one country by companies residents in another. The purpose of this section is to analyse whether there is an incentive for EU companies to choose specific forms of investment and the tax-favoured locations for their investments (which may not be the most favourable locations in the absence of taxes). To the extent that companies respond to such incentives, the tax system may create a global misallocation of resources as activities may be financed or undertaken in high cost locations because they are tax-favoured.

In order to understand the effects of the different Member States' tax systems on investment coming from two of the main EU economic partners in the transnational case, the analysis also considers inward investment into each Member State from the USA and from Canada, which respectively have a credit system and an exemption system.

As was the case for the analysis of domestic investments, this section gives summary measures of the potential distortions by tax systems of transnational investments and does not provide a measurement of the impact of these potential distortions on international investment patterns. (Box 6 in section 4 presented a short survey of the empirical studies which have attempted to measure this impact). The fact that companies may use more complex financial arrangements and group structures in order to minimise tax burdens indicates that the potential distortions reported in this section can be sufficiently large to alter company behaviour from that which would otherwise prevail. It should also be recalled that there are costs associated with these complex financial arrangements. Moreover, to the extent that these arrangements can imply tax evasion, this may create new distortions.

The position of the subsidiary is essentially the same as the independent firm analysed in the domestic case. It may invest in one of the five assets, and it is financed in one of the three ways: retained earnings, new equity and debt. The parent company also raises finance in one of these three ways.

As in the analysis of domestic investments, primarily corporate taxes are considered. Given that it is assumed here that cross border flows of capital are possible, it seems reasonable also to suppose that parent companies can be financed on the international market. But in this case, for the reasons explained in Box 5, it is less plausible that personal taxation affects investment decisions and that the identity of the shareholder is known. Most of the analysis in this section is therefore based on the comparison of taxes paid by the corporation only.

Nevertheless, section 6.5 discusses the role of personal taxes, especially in the context of whether tax credits associated with a dividend payment by the parent to the domestic shareholder are available if the underlying source of income is from abroad.

It is worth noting that the introduction of transnational investments considerably increases the number of cases to be dealt with

• 42. 

  Therefore, this section summarises these cases (by calculating simple averages)

and highlights only the main issues arising from the transnational investment additional to those already discussed in the context of the domestic investment.

Transnational effective tax rates: detailed positions (cost of capital and EATR) of the EU

Member States

The tables in this section summarise the effective tax burden of each possible transnational investment, averaged across the five different assets and the three sources of finance of the parent company. A separate table is used for each of the three different ways in which the subsidiary can be financed.

These tables address the question of differences in the effective tax burdens across possible locations of the investment for a given State of residence of the parent company or, alternatively across possible States of residence of the parent, for an investment in a given location. Moreover, they identify differences in taxation which arise solely because of the way the investment is financed.

• A) 

  The case of a marginal investment

rate of return of 5% (see Table 17); whereas if the Belgium subsidiary were financed by funds lent by the parent, the subsidiary would need to earn 6.0% (see Table 19).

Table 17 presents the cost of capital for investment financed by retained earnings in the subsidiary. As was the case for a domestic investment discussed in section 4.1, the cost of capital in this case is not influenced by the taxation of dividends paid by the subsidiary to the parent. In fact, since the parent forgoes dividends to finance the investments and receives higher dividends as the return from the investment, the tax rate on such dividend flows nets out of the analysis.

Given the assumption of no personal taxes, there is in general also no differences between investments financed by the parent from retained earnings and new equity. Within the different sources of finance used by the parent, then, only debt financing introduces any element of the parent (home) country tax regime.

This implies that within each column in Table 17 - i.e. considering a single subsidiary (host) country - differences in the cost of capital across different home countries arise only in the different treatment of debt in the parent company. Some countries, Austria, Denmark, Netherlands, Spain and Canada do not permit interest paid on loans used for outbound investments to be deductible; these countries have a higher cost of capital. Other countries do permit this. In their case the tax rate determines how valuable this deduction is, and hence how low the average cost of capital is. It is worth noting that the higher the tax rate, the more valuable the deduction.

By contrast, the differences within each row reflect primarily the host country tax system. These differences are likely to affect the location choice of the parent companies. The relative ranking of host countries in Table 17 is very close to that shown in the case of domestic investments. Thus, on average, Italy is the most attractive host location and France and Germany are the least attractive host locations.

Summing up the results of table 17, it can be observed that when the subsidiary is financed by retained earnings, the differences in the cost of capital of subsidiaries located in the same country largely depend on the treatment of debt financing of the parent and on the tax rates applied in the home countries. On the other hand, the incentives to locate faced by the parent company largely depend on the domestic tax systems of the possible host countries.

Mean

7.66.16.46.66.05.56.07.36.46.27.56.26.36.76.47.6

Kingdom

United

7.76.36.66.76.35.96.17.36.46.47.76.46.56.8./.7.7

edenSw

6.75.45.75.85.45.05.16.45.55.56.75.55.6./.5.76.7

Spain

7.76.36.66.76.35.86.27.46.46.47.76.4./.6.86.77.7

Portugal

7.96.36.76.86.35.96.37.56.56.57.9./.6.56.96.77.9

ndsrlaNethe

7.76.26.66.76.35.86.27.46.46.4./.6.36.56.86.67.7

Luxembourg

7.76.26.56.76.25.76.17.36.3./.7.76.36.46.76.67.7

i ngs.

yItal

earn5.53.94.24.43.93.43.95.1./.4.05.54.04.14.44.35.5

ne d

Ireland

5.94.85.15.24.84.54.3./.4.94.95.94.95.05.24.95.9

retaii nance

i th

wGreece

7.66.16.46.66.15.5./ .7.26.26.27.66.26.36.66.57.6

ced parent f

i na nGermany

s f9.78.18.48.68.1./.8.19.38.28.29.78.28.38.78.59.7

i a

ry iFrance

./.

Mean

6.36.57.46.16.36.46.05.86.97.26.47.47.46.26.16.56.58.6

Kingdom

6.46.6United

7.76.56.66.76.46.17.77.36.67.77.76.56.56.8./.7.7

edenSw

5.45.76.75.65.75.85.55.27.26.45.76.76.75.65.6./.6.17.2

Spain

6.36.77.76.56.66.76.46.07.07.46.67.77.76.6./.6.86.79.6

Portugal

6.36.87.96.56.76.86.46.16.87.56.67.97.9./.6.56.96.712.4

ndsrlaNethe

6.46.67.76.56.66.76.46.07.07.46.67.7./.6.56.56.86.68.3

Luxembourg

6.36.67.76.46.56.76.36.06.67.36.5./.7.76.56.46.76.68.3

yItal

4.04.35.54.14.24.44.03.64.05.1./.5.55.54.24.14.44.37.4

4.65.0y

.Ireland

5.95.05.15.24.94.78.2./.5.15.95.95.05.05.27.15.9

quit147

6.26.5w ei nanceGreece

7.66.36.46.66.26.6./ .7.26.47.67.66.46.36.66.87.6

h ne

8.28.6 witGermany

e d parent f7.66.26.36.56.1./.6.07.26.37.67.66.36.26.56.49.7

France

./.

Mean

7.06.66.36.86.25.96.77.77.14.86.66.56.46.66.45.86.17.4

Kingdom

United

7.16.96.97.36.86.57.38.17.75.47.17.17.07.17.06.4./.7.6

edenSw

6.66.06.36.76.25.96.77.57.24.96.56.56.36.56.3./.6.17.1

Spain

7.56.96.46.96.36.06.87.77.04.86.76.66.56.7./.5.96.17.7

Portugal

8.26.96.26.76.15.86.77.66.85.16.56.56.3./.6.35.75.98.3

ndsrlaNethe

6.96.86.46.96.36.06.87.77.04.86.66.6./.6.76.55.96.17.3

Luxembourg

6.96.76.16.65.95.76.57.46.64.46.3./.6.16.46.15.55.87.0

yItal

4.64.56.16.65.95.66.57.56.54.3./.6.36.16.46.15.55.87.4

Ireland

7.65.66.97.26.86.67.27.97.4./.7.07.06.97.16.96.57.17.4

148

6.76.8bt.

i nanceGreece

5.76.25.65.36.28.2./ .5.16.05.95.86.05.85.25.56.4

h de

6.76.6 witGermany

e d parent f5.86.35.75.36.3./.6.34.06.16.15.96.15.95.25.57.2

France

7.17.66.96.6./.8.57.55.37.37.37.17.47.16.56.88.0

6.66.4

7.17.0

6.66.4

6.76.4

6.76.3

6.66.4

6.36.1

6.36.1

7.17.1

149

5.95.8

6.05.7

7.37.1

Table 18 considers the case in which the subsidiary is financed by new equity from the parent. Since the return from this investment is assumed to be paid to the parent as a dividend, this adds another ingredient to the cost of capital calculation: the taxation of such dividend flows

• 43. 

  The differences between Tables

17 and 18 therefore reflect the taxation of the dividend receipts in the hand of the parent (and the German split rate system when Germany is the host country).

A number of (home) countries exempt such income in the hands of the parent: Austria, Belgium, Denmark, Finland, Luxembourg, the Netherlands, Sweden and Canada. For parent companies in these countries, in the absence of personal taxes the cost of capital when funding the foreign subsidiary through new equity is the same as when the subsidiary retain earnings.

Other countries operate a limited credit system: essentially credit is given for the foreign tax paid (in the host country); the home country levies further tax only if the home country tax exceeds that of the host country. In many cases, then, financing the foreign subsidiary by new equity and paying tax on the dividend receipts generates a higher cost of capital of the transnational investments. As a result the average cost of capital for inward investments (the average of each column) is higher in Table 18 than in Table 17.

Table 19 presents the case in which the parent lends to the subsidiary and subsequently receives an interest payment (and the return of capital). All the host countries permit the interest paid to be deductible from corporation tax, although some charge a withholding tax as interest is paid; all home countries tax the interest income, with a credit for any foreign tax levied.

Relative to the case of financing by retained earnings, there are two main differences. First, the income of a marginal investment is primarily taxed in the home country rather than the host country; this can increase or reduce the cost of capital depending on which country has the higher rate. Second, however, it is generally the case that the parent is able to claim interest deductibility on its own borrowing if that borrowing is used to finance lending to the subsidiary; this gives an advantage over providing equity finance to the subsidiary for those countries where interest deductibility is not permitted if the loan is used to support equity investment in the subsidiary. In general these factors tend to reduce the dispersion of costs of capital across different possible locations of investments.

The figures shown in Tables 17, 18 and 19 illustrate a large variation in the way that each country treats other countries. Thus the return required by a subsidiary of a parent country in one country depends crucially on where that subsidiary is located. This suggests that there are considerable incentives for companies to choose tax-favoured locations for their investments which may not otherwise be the most favourable location. Similarly, subsidiaries operating in a given country face different required rates of return depending on where their parent company is located. Moreover, these figures show that, in line with the analysis of domestic investments, there are differences in taxation which arise solely because of the way which the investment is financed.

Mean

33.429.229.830.528.619.033.233.830.028.833.329.629.230.931.537.033.230.1

Kingdom

United

31.827.928.028.527.517.733.230.628.227.331.828.227.628.7./.

31.830.728.4

edenSw

26.022.222.222.721.711.030.324.822.521.626.022.521.8./.

24.828.427.622.9

Spain

35.231.231.431.930.821.733.534.031.530.735.231.5./.

32.131.641.235.031.6

Portugal

37.033.033.233.732.723.833.935.833.332.637.0./.

32.933.933.548.438.633.3

ndsrlaNethe

35.131.231.331.830.821.633.534.031.530.7./.

31.431.032.031.637.133.031.2

Luxembourg

36.632.632.833.332.323.433.435.432.9./.36.632.932.533.533.138.5

34.432.9

yItal

31.828.028.128.627.617.927.530.7./.27.531.828.327.828.828.338.329.928.0

Ireland

11.78.17.88.37.5-5.727.9./.8.47.1

11.78.37.48.522.111.725.09.9

i ngs.

i nanceGreece

earn34.430.430.631.030.025.4./.33.230.729.934.430.730.231.232.934.4

33.031.1

ne d

parent fGermany

46.142.142.442.941.8./.

41.545.042.441.846.142.342.143.142.750.443.543.0

retai

i th

wFrance

./.

Mean

33.029.429.330.128.520.135.833.430.232.832.929.828.830.531.939.6

Kingdom

United

31.828.628.028.527.818.638.030.628.831.831.828.827.628.7./.

31.8

edenSw

26.022.922.222.722.111.936.624.823.126.026.023.121.8./.

26.530.1

Spain

35.231.931.431.931.222.536.134.032.235.235.232.1./.

32.131.646.2

Portugal

37.033.733.233.733.024.735.335.834.037.037.0./.

32.933.933.558.4

ndsrlaNethe

35.131.931.331.831.122.436.134.032.135.1./.

32.131.032.031.638.8

Luxembourg

36.633.332.833.332.624.234.835.433.6./.36.633.532.533.533.140.2

yItal

31.828.728.128.627.918.727.830.7./.31.831.828.927.828.828.343.3

Ireland

i nance11.78.87.88.37.8-4.839.5./.9.0

11.711.79.07.48.529.811.7

Greece152

y parent f34.431.130.631.030.429.2./.33.231.434.434.431.330.231.234.134.4

quit

w eGermany

40.136.936.436.936.2./.

35.539.037.140.140.137.136.137.136.750.3

h ne

witFrance

./.

Mean

35.730.429.731.929.428.731.127.636.426.031.130.429.931.329.928.530.5

Kingdom

33.129.2United

29.131.528.727.831.026.938.023.830.829.929.330.929.327.4./.

edenSw

31.524.024.527.024.123.226.421.636.619.326.325.324.726.424.7./.

26.5

Spain

38.432.330.833.230.429.532.728.936.125.532.431.631.032.5./.

29.129.9

Portugal

43.633.931.934.231.430.633.730.135.328.333.432.632.1./.

32.130.231.0

ndsrlaNethe

34.732.030.833.230.329.432.628.936.125.532.431.5./.

32.531.029.129.9

Luxembourg

36.133.331.433.831.030.133.229.634.826.233.0./.31.633.131.629.830.6

yItal

31.628.633.535.833.132.235.331.935.228.3./.34.233.735.133.731.832.6

i nance

Ireland

34.711.9

bt15.918.615.414.517.911.739.5./.

17.816.716.117.916.114.229.8

parent f

153

h deGreece

34.631.9

wit28.330.727.927.030.231.9./.26.430.029.128.530.128.526.630.0

e d

Germany

39.537.535.037.334.633.736.8./.

36.329.936.535.735.236.635.233.434.1

y

is finance ighted average ofFrance

./.

36.634.730.2

31.433.129.6

29.731.525.5

40.936.231.0

49.339.731.9

35.733.630.9

36.334.231.4

43.136.233.3

18.334.718.7

154

30.631.928.9

44.437.635.0

Much of the discussion above regarding the cost of capital remains relevant for the case of the EATR. However, some additional factors are also now present.

In measuring the EATR it is assumed the investment is more profitable that is, extra-profit is generated by the subsidiary. This is paid to the parent in the form of dividend. Any tax liability associated with such a payment reduces its value to the parent and hence increases the EATR. In this case, the taxation of the dividend payments from the subsidiary to the parent does affect the EATR even when the subsidiary is financed by retained earnings. Further, since this tax rate varies according to the home country, the EATR on an investment in a specific host country may vary according to the home country of the parent, even when the subsidiary is financed by retained earnings. Of course, this is also true when the subsidiary is financed by new equity and debt.

Consequently, there is even more variation within each column in Table 20 than there is in the comparable Table 17 showing the cost of capital. When personal taxes are excluded, the main effect of the home country tax is through the tax rate; where the home country has a limited credit system, there is a positive tax on the dividend payment to the parent only if the home country tax exceeds the host country tax.

The variation within the rows of Table 20 also reflects home and host country taxation. Yet, the same host countries turn out to have high EATR as had high cost of capital: Germany has an average EATR of 43.0%, and France has an average EATR of 38.4%. At the other extreme, Sweden has an average of 22.9% and Italy an average of 28.0%. These would appear to be significant differences in the EATR facing companies in other countries choosing where to locate.

A similar pattern arises in Table 21, where the subsidiary is financed by new equity.

When the subsidiary is financed by debt, these differences are still considerable even if they are smaller.EATR ranges from 25.5% for Sweden as a host country to 36.0% for France as host country.

If the position of USA and Canadian investors is taken into account, the data tend to show that investments from these two countries - and from Canada in particular - into the EU are relatively more highly taxed than intra-EU investments. The worst situation for Canadian inbound investments is largely due to the weight of withholding taxes on dividends. Moreover, since the USA applies a foreign tax credit system, and Canada an exemption system, investments from the latter country retain the benefit of lower than Canadian rates in the host country. In general investments from countries which operate an exemption system benefit from a lower tax rate in the host country. Data (EMTR and EATR) related to foreign investment in Ireland from the USA and Canada clearly illustrates this.

Allocation effects of international taxation

As shown in the previous section, issues of international taxation are very complex to deal with since they involve the interaction between national tax systems. One commonly used criteria to assess the allocation effects of international taxation is to capture the extent to which the tax treatment of transnational investments gives an incentive to undertake transnational, as opposed to domestic, investments. This is done by evaluating the tax treatment of cross-border investment flows against the criteria of capital import and capital export neutrality.

Capital export and capital import neutrality

Capital export neutrality (CEN) occurs when the tax system is neutral towards the export of capital since the investors face the same effective tax burden on income from similar investments, whether they invest in the domestic economy or abroad. In such situations the tax systems provide no incentives to invest at home rather than abroad, and vice versa. A regime of capital export neutrality therefore tends to ensure an efficient allocation of resources across countries.

CEN is achieved when investors are taxed on accrued worldwide income and receive full credit against domestic tax liabilities for all taxes paid abroad. A pure credit system with no limitation on the foreign tax credit and no deferral of domestic taxes on profits retained abroad would ensure capital export neutrality. Under such a regime, the free mobility of capital would tend to equate the effective tax burden across borders, because each investor would then also obtain the same after-tax rate of return on domestic and foreign investments. A cross-country equalisation of the rates of return before tax implies that no output gain can be made by reallocating capital from one country to another. It is worth noting that a pure credit system does not exist in any EU tax regime.

The importance of attaining capital export neutrality is well assessed in terms of economic welfare. It seems clear that if taxation distorts the location of productive activity, then goods may be produced at higher cost, which is likely ultimately to be borne by the consumer in terms of higher price. Thus the absence of capital export neutrality creates an economic loss to the extent that differences in the cost of capital and effective tax rates result in changes in behaviour. Moreover, capital export neutrality maximises the volume of output obtainable from any given global stock of capital.

for savers in different countries, but would distort the pattern of international investment by causing the cost of capital to deviate from one country to another. Consequently the effect on welfare of attaining one neutrality or the other, depends on the relative sensitivity of investment demand to the cost of capital and of savings to the after-tax rate of return.

It is worth noting that capital export and import neutrality could be achieved simultaneously only in the hypothesis that the effective tax burden on profit is identical across all countries, that is in presence of a far-reaching tax harmonisation. This is clearly not the case today.

These concepts are both useful benchmarks by which to judge the efficiency effects of international tax arrangements and a useful starting point to analyse international tax arrangements. The use of these criteria may encourage policy makers to take a more global view of the benefits and costs of existing international tax arrangements and proposed changes thereto.

Relevant economic measures: average cost of capital and EATR by country

• A) 

  The case of a marginal investment

Tables 23 summarises the cost of capital shown in Tables 17 to 19 and compares these costs with those for domestic investments, in order to identify the impact of international tax regime on the incentive to undertake transnational, as opposed to domestic, investments.

Table 23 Average Cost of Capital by Country

• - 

  domestic, average inbound and outbound

• - 

  only corporation taxes

-average over sources of finance of subsidiary

EU AverageEU Standard

Cost of CapitalDeviation

%

D o mestic

InboundOutboundInboundOutbound

Austria6.36.57.10.20.6

Belgium6.46.76.30.30.6

Denmark6.46.66.30.30.6

Finland6.26.46.30.30.6

France7.57.76.20.30.5

Germany7.37.06.30.30.6

Greece6.16.46.60.30.6

Ireland5.75.96.40.40.6

Italy4.85.06.50.30.4

Luxembourg6.36.56.70.30.6

Netherlands6.56.67.10.20.6

Portugal6.56.76.30.30.6

Spain6.56.76.30.30.6

Sweden5.86.06.30.30.6

United Kingdom6.66.86.40.30.5

EU Mean6.36.56.50.30.6

outbound investments. The fourth and fifth columns present the dispersions respectively for inbound and outbound investment

44.

A comparison of the figures in the first three columns gives some indication of the effects of the tax systems on investment flows. Many factors will, of course, determine investment flows. Insofar as tax is of any importance, then if the required cost of capital when investing domestically is lower than when investing abroad, companies will prefer domestic operations (assuming, of course, that there are equal investment possibilities in each country). Whether this results in a net inflow of capital, depends on whether investment in that country is more attractive to foreign owned companies than investment in their own domestic economies or into other countries.

By computing the average required rate of return when investing in or from each country a lot of the information contained in Tables 17 to 19 is lost, in particular the variation between countries and source of finance. (Annex C presents detailed country tables showing the outbound and inbound cost of capital and EATR for the 15 investments analysed in this study). To compensate, the standard deviations (fourth and fifth columns) give some indication of the degree to which the international tax regime results in subsidiaries which operate in the same country face different effective tax rates according to the residence of the their parent company and the source of finance, and the extent to which the location of a subsidiary for a parent company and the source of finance may affect the effective tax rate. A low figure in the last column, for example, suggests that there is only a small dispersion across countries in the effective tax rate faced by a company when considering in which country to undertake an investment.

If the third column (outbound) were identical to the first column (domestic) and the fifth column (standard deviation of outbound) were full of zeros, this would imply that any company resident in a EU Member State would face (on average) the same cost of capital whether it invested at home or in any other Member State. If that applied generally in the EU, then location decisions of companies would not be affected by taxation. This situation represents capital export neutrality.

If the second column (inbound) were identical to the first column (domestic) and the fourth column (standard deviation of inbound) were full of zeros, this would imply that any company resident in a EU Member State or Canada or the USA would face (on average) the same cost of capital if it invested in a specific EU host country. This implies that companies choosing to locate in a specific location all face the same cost of capital, and hence, none has the benefit of a tax-induced competitive advantage over others. This situation represents capital import neutrality.

As far as the case of outbound investments is concerned, the companies located in countries that have the highest domestic cost of capital and apply an exemption system are less heavily taxed when investing abroad and, in particular can benefit from a lower effective tax burden if they invest in a foreign subsidiary either through new equity or profit retention (see Tables 17 to 19). Thus, from the point of view of high tax countries, investments in low tax countries are more advantageous than domestic investment, and debt financing is the least attractive way to finance the subsidiary.

On the other hand, if the investor is located in countries which have the lowest costs of capital, on average, domestic investment is more attractive than transnational investment. It is worth noting that, when different sources of finance are considered, it is not always true that high tax countries are always unattractive and that it depends on the source of finance of the subsidiary. If the subsidiary is financed by debt, as a consequence of interest deductions the foreign profits are taxed at the lower domestic and not the foreign corporate profit rate (see Tables 20 to 22). This could imply a lower effective tax burden.

When inbound investments are taken into account, all countries, except Germany, have average cost of capital for inbound investments higher than that of domestic investment. Therefore, domestic companies have on average a competitive advantage over subsidiaries located in their country.

• B) 

  The case of a highly profitable (infra-marginal) investment

As with the cost of capital, it is useful to summarise the information in Tables 20 to 22 in order to assess how far away the EU tax regimes are from either capital export neutrality or capital import neutrality. Table 24 presents results for the EATR comparable to that in Table 23 for the cost of capital.

Table 24Effective Average Tax Rate by Country

• - 

  domestic, average inbound and outbound

• - 

  only corporation taxes

EU AverageEU Standard

EATRDeviation

D o mestic

InboundOutboundInboundOutbound

Austria29.830.332.12.56.6

Belgium34.534.830.22.06.3

Denmark28.829.529.52.76.6

Finland25.526.529.73.46.5

France37.537.829.41.96.2

Germany39.138.522.20.97.4

Greece29.630.635.11.11.7

Ireland10.513.531.18.14.0

Italy29.830.030.42.26.5

Luxembourg32.232.530.72.16.5

Netherlands31.031.432.12.36.6

Portugal32.633.030.22.16.4

Spain31.031.629.32.46.6

Sweden22.924.129.93.96.3

United Kingdom28.229.131.33.03.5

The same is true for inbound investment, although in this case for every host country, except Germany,EATR on inbound investment is higher than that for domestic investments. Clearly, on average, on this measure the net impact of the taxation on international flows is to increase tax liabilities. There are again variations within each host country for the average EATR faced by companies resident in different home countries.

It is worth noting that the countries which show the highest differences between average domestic EATR and average outbound EATR are the countries which have the highest EATR and the highest profit tax rates on domestic investment (Germany, France, Belgium) and the countries which have the lowestEATR and the lowest profit tax rates on domestic investment (Ireland, Sweden and Finland). This seems to confirm that, even when the interactions of different tax regimes are taken into account, differences in national profit tax rates are of the utmost importance in determining the role of taxation in resource

allocation.

The tax minimisation approach: "tax efficient" average cost of capital and EATR by country

The previous section showed that, on average, the interaction of the EU Member States taxation systems implies differences in the tax treatment of domestic investment compared with outbound or inbound investment and, therefore, that capital export or capital import neutrality is never attained. But, as already explained, a lot of specific information for each Member State is lost when computing overall averages.

In particular, considering the different treatment across Member States of different sources of finance, it is realistic to suppose that parent companies would try to minimise their tax burden by choosing the most convenient source of finance of the subsidiary. If one particular source of finance is tax disadvantaged, then it would not be used.

This section considers how the international tax regime affects the effective tax burden faced by a company willing to invest abroad when it chooses the most tax-efficient means of financing the subsidiary

45.

• A) 

  The case of a marginal investment

Table 25"Tax Efficient" Average Cost of Capital by Country

• - 

  domestic, average inbound and outbound

• - 

  only corporation taxes

• - 

  only most favoured source of finance for the subsidiary

EU AverageEU Standard

Cost of CapitalDeviation

%

D o mestic

InboundOutboundInboundOutbound

Austria6.36.06.20.60.4

Belgium6.45.95.90.40.8

Denmark6.46.15.90.40.6

Finland6.25.95.70.40.6

France7.57.05.90.50.7

Germany7.35.75.50.60.8

Greece6.15.75.80.30.9

Ireland5.75.04.80.40.4

Italy4.84.36.20.60.5

Luxembourg6.35.96.00.50.8

Netherlands6.56.16.30.40.4

Portugal6.56.16.00.40.8

Spain6.56.16.00.40.7

Sweden5.85.55.70.40.5

United Kingdom6.66.35.80.40.6

EU Mean6.35.85.80.40.6

EU Standard Deviation0.60.60.3

This suggests also that, as already commented on in section 6.2, from the point of view of a foreign investor, so-called high tax countries are not always unattractive as a business location and that companies resident in low tax countries may take advantage to invest abroad instead of at home.

It is worth noting that it has been assumed here that the company is able to use the most "tax efficient" way of financing only in the case of transnational investment. This hypothesis is realistic considering that a multinational company has, in general, greater flexibility when financing its subsidiary than a domestic company (see footnote 44).

The relevant component that results in distortions with respect to cross border location and financing decisions is, above all, the profit tax rate. In fact, if there were only minor differences between the tax rates, there would be less incentives to use either debt (e.g. for investors located in low tax countries) or equity (investors located in high tax countries) for the financing of foreign investment. However, the tax base can have a greater impact on the cost of capital in particular situations (e.g. favourable depreciation regimes) as is the case, for example, in Belgium, Greece, Italy and Sweden.

It should be noted that Table 25 shows "extreme" situations and that national tax regimes may impose restrictions on the use of particular sources of finance.

Moreover, Table 25 shows higher dispersions than Table 23. This means that the averages shown in Tables 25 hide even greater variations across countries.

• B) 

  The case of a highly profitable (infra-marginal) investment

Table 26 shows the averages of the EATR across host countries and across home countries based on the most efficient way to finance the subsidiary, and is thus comparable to Table 24.

Table 26"Tax Efficient" Effective Average Tax Rate by Country

• - 

  domestic, average inbound and outbound

• - 

  only corporation taxes

• - 

  only most favoured source of finance for the subsidiary

EU AverageEU Standard

EATRDeviation

D o mestic

InboundOutboundInboundOutbound

Austria29.828.729.33.05.7

Belgium34.532.628.82.27.0

Denmark28.827.828.23.06.6

Finland25.524.827.93.66.1

France37.535.628.22.46.8

Germany39.134.919.01.78.3

Greece29.628.432.71.52.6

Ireland10.59.926.07.43.2

Italy29.827.929.43.17.0

Luxembourg32.230.728.22.77.1

Netherlands31.029.829.52.85.7

Portugal32.631.229.12.57.0

Spain31.029.828.32.86.9

Sweden22.922.327.84.05.9

United Kingdom28.227.429.43.33.3

EU Mean29.528.128.13.16.0

EU Standard Deviation6.55.92.8

Restrictions on imputation systems for transnational investments

The previous analysis was based on the assumption that the parent company has sufficient undistributed domestic profit that any additional dividend which it wishes to pay as a result of the additional investment in the subsidiary can, for tax purposes, be deemed to be a payment from domestic income. In this section the situation is considered in which dividend payments are paid from the foreign source of income (the subsidiary) to the shareholders. In this way another element is introduced: the domestic treatment of foreign source income.

In Finland and France, dividend payments deemed to be paid from foreign source income are subject to an equalisation tax. In 1999 in Germany, such dividend payments did not qualify for the imputation tax credit. In all three cases, then, there can be in effect an additional tax on outbound investments. In fact, compared to the situation shown in Table 23, in Finland and France, the average cost of capital for outbound investment increases substantially - from 6.3% to 7.8% for Finland and from 6.2% to 8% for France. There is a smaller rise for Germany: from 6.3% to 7%.

As far as the EATR is concerned, the rise is even more spectacular. Compared with the situation shown in Table 24, there is a substantial rise in the average EATR for outbound investment from Finland and France - from 29.7% to 46.7% in Finland and from 29.4% to 48.9% in France. Again, the rise for outbound investments from Germany is smaller - from 22.2% to 33.2%. In the case of Finland, this exacerbates the average discrimination against outbound investments. In Germany it reduces the average discrimination in favour of outbound investment. And in France it turns the discrimination from being in favour of outbound investment to being in favour of domestic investment.

Effects of the German tax reform on international investments

The purpose of this section is to examine the impact of the German tax reform on international investment. First, the case of inbound investment to Germany and outbound investment from Germany are considered. Then, the impact of the German tax reform on the overall means and standard deviations in the EU is analysed.

• A) 

  The case of a marginal investment

The average cost of capital for outbound investment, again averaged over the three sources of finance of the subsidiary, is virtually unchanged by the tax reform. This is partly because Germany largely exempts foreign source dividends. The lower tax rate on the receipt of interest from subsidiaries tends to reduce the cost of capital. However, this is offset by the reduction in the value of the deductibility of interest paid by the parent, and the abolition of the split rate system, which pre-reform gave an advantage to new equity financing of the parent.

However, this is only true on the assumption that all distributions are financed from the domestic income of the corporation. If we assume instead that distributions are financed from foreign source income, then the discrimination of foreign investment which was analysed in the section 6.3 above no longer exists - because under the reformed system, distributions are treated in the same way irrespective whether they are financed from domestic or foreign profits.

Table 27Average Cost of Capital for Germany and EU average

- domestic, average inbound and outbound

- only corporation taxes

EU AverageEU Standard Deviation

Cost of Capital

%m e stic

DoInboundO u t boundInbound

O u t bound

BEFORE REFORM

Average over sources of finance of subsidiary

Germany7.37.06.30.30.6

EU Mean6.36.56.50.30.6

EU Standard Deviation0.60.60.3

Tax Efficient source of finance of subsidiary

Germany7.35.75.50.60.8

EU Mean6.35.85.80.40.6

EU Standard Deviation0.60.60.3

AFTER REFORM

Average over sources of finance of subsidiary

Germany6.86.96.30.30.6

of capital rises, and the tax efficient form of finance may in any case also change. The average cost of capital for the tax efficient form of the outbound investment also rises.

However, these changes in Germany are not large on average. In no case does the average cost of capital change by more than one half of one percent. As a result, the position for the EU as a whole is virtually unchanged.

• B) 

  The case of a profitable (infra-marginal) investment

Table 28Average EATR for Germany and EU average

- domestic, average inbound and outbound

- only corporation taxes

EU AverageEU Standard Deviation

EATR

%m e stic

DoInboundO u t boundInbound

O u t bound

BEFORE REFORM

Average over sources of finance of subsidiary

Germany39.138.522.20.97.4

EU Mean29.530.230.22.75.8

EU Standard Deviation6.55.72.6

Tax Efficient source of finance of subsidiary

Germany39.134.919.01.78.3

EU Mean29.528.128.13.16.0

EU Standard Deviation6.55.92.8

AFTER REFORM

Average over sources of finance of subsidiary

Table 28 presents a similar analysis to the one in Table 27, this time for the EATR. For inbound investment the EATR tends to fall for investment financed by retained earnings and new equity, and to rise for investment financed by debt. Overall, the reduction in the tax rate causes the average EATR for inbound investment financed by an average of the three types of finance to fall, although by less than the fall in the average domestic cost of capital.

However, there is a significant rise in the EATR on outbound investment. Pre-reform the EATR on outbound investment was considerably lower than the EATR on domestic investment, thereby inducing German firms to invest abroad. This rise can be traced largely to the reduction in the value of the deductibility of interest paid by the parent and the abolition of the split rate system. However, as with the cost of capital, this is only true for the case in which dividends are paid from domestic source income.

A similar pattern is found for the tax efficient financing of the subsidiary, both for inbound and outbound investment. The larger changes in the EATR as a result of the reform create some impact on the overallEU averages, as shown in the Table.

Neutralities and distortions in transnational investments: concluding remarks from the

international analysis

The results for the international case, show that the co-existence of the Member States tax regimes may have an influence on transnational investment patterns and financing decisions.

The data arising from the quantitative analysis, and notably from Tables 17 to 22, illustrate a variation in the way in which each country treats other countries. Thus, the effective tax burden of subsidiaries of a parent company in one country depends crucially on where that subsidiary is located. The range of variation in the effective tax burdens of subsidiaries located in different host countries can rise above 30 points regardless of the method of financing of the subsidiary. This provides an incentive for companies to choose the most tax-favoured location for their investment, which may not be the most favourable location in absence of taxes. Similarly, subsidiaries operating in a given country face different effective tax burdens depending on where their parent company is located. Even in this case the range of variation can reach more than 30 points.

As said already in commenting on the results of the domestic analysis, neutrality of taxation systems is one of the goals of taxation policy and this has to be balanced against other legitimate goals. However, to the extent that the absence of neutrality determines movements of capital which are not justified by economic efficiency and that involve welfare losses, the picture presented in this section deserves attention. To what extent differences of the considerable size observed in this quantitative analysis and the consequent possible losses in welfare can be justified by the need to maintain national autonomy in view of national goals attached to taxation policy is, ultimately, a matter of political choice.

THE IMPACT OF HYPOTHETICAL POLICY SCENARIOS IN THE EU

Purpose of the simulations

The previous sections have clearly shown that the actual tax treatment of investment strongly differs across countries in the EU. In particular, the effective tax burden for cross-border investments in the EU considerably differs according to the home country of the parent companies and the location of their foreign subsidiaries.

This situation is not optimal with regard to the proper functioning of the Internal Market. Indeed, the absence of capital export neutrality may lead to distortions in the international allocation of investment as, ceteris paribus, investments may take place not in the lowest cost locations but in the lowest tax

locations. This in turn potentially limits growth in productivity and employment in the EU.

It is therefore particularly useful to consider how the measures of the cost of capital and effective tax rates presented above would be different in the event of various hypothetical tax policy scenarios.

In what follows, 15 hypothetical policy scenarios are considered. Each of the simulations is based on a particular element of the tax regimes being harmonised across the EU. This helps to identify the importance of specific features of tax regimes, notably for capital export and capital import neutrality. The simulations are divided into three groups (see Box 8). The first group of simulations examines the impact of elements of the "domestic" corporation tax regime, that is the statutory tax rate and the value of capital allowances. The second group turns to elements of the "international" corporation tax regime, such as the treatment of interest payments from one Member State to another, and the taxation of income received in the home country. The third group examines the relationship between corporation tax and personal taxes.

Box 8

Definition of simulations

A. Domestic Elements of Corporation Tax

1.Common corporation tax rate, incl. surcharges. Rate is EU average of 32.28%.

2.Common corporation tax rate, incl. surcharges and local taxes. Rate is EU average of 33.84%.

3.Common corporation tax rate, incl. surcharges and local taxes. Rate is 25%.

• 4. 

  Band of corporation tax rates, incl. surcharges and local taxes. Rates permitted between 25% and 35%. Countries outside

this band move to nearest limit.

5.Harmonisation of capital allowances. Common rules for depreciation, inventories and investment incentives.

6.Common tax base. Depreciation based on true economic depreciation.

B. International Elements of Corporation Tax

7.Abolition of withholding taxes on interest paid by subsidiary to parent within EU.

8.Limited credit system for foreign source dividends received by parent from EU subsidiary. No discrimination in imputation systems against foreign source income originating in the EU.

9.Full credit system for foreign source dividends and interest received by parent from EU subsidiary. No discrimination in imputation systems against foreign source income originating in the EU.

• 10. 

  Exemption for foreign source dividends received by parent from EU subsidiary. No discrimination in imputation systems

against foreign source income originating in the EU.

• 11. 

  Taxation according to parent country rules. Subsidiary taxed using the tax base of the home country of the parent but the

tax rate of the host country. No interest deductibility by the subsidiary. No taxation of foreign source income received by parent. No discrimination in imputation systems against foreign source income originating in the EU

C. Relationship of Personal and Corporate Taxes

• 12. 

  Classical system. Double taxation of dividends. Personal tax rates for taxpayers (not including zero-rated shareholders)

set to highest income tax rate on labour income. Equalisation taxes abolished. Capital gains taxes and personal property tax rates unchanged from base case.

Scenarios involving domestic elements of corporation tax

The Base Case

A number of simulations based on harmonization or approximation of the statutory corporate tax rate and the definition of the tax base are considered first. In the hypothetical policy scenarios listed in Part A of Box 6, the international elements of tax regimes - such as withholding taxes and the taxation of foreign source income - and the personal tax regimes are left unchanged.

The first row of Table 29 reproduces the average across the 15 EU Member States for the EATRs and the cost of capital as well as the standard deviation of the distribution across the 15 EU Member States in the current situation (including the German reform). The subsequent rows in the Table show how these summary measures change in the presence of various hypothetical policy scenarios.

As shown in section 6 above, the existing system (the "Base Case") clearly does not exhibit either capital export neutrality or capital import neutrality. For the measures based on the average source of financing of the subsidiary for outbound and inbound investments, the average cost of capital is only slightly higher than the average domestic cost of capital. The same is true of the average EATR. However, these averages hide considerable variations across countries, which are summarised in the figures for the standard deviations for both inbound and outbound investment. As explained in more detail in section

6.3, the figure for standard deviation for outbound investment expresses the average deviation in the cost of capital or EATR for an investment made in 14 potential host countries by a parent company based on their territory. The standard deviation summarises the variability in the tax treatment of foreign investments across the EU. It is therefore the focus of the analysis of the impact of hypothetical reforms presented below.

Comparing inbound and outbound investment, it is clear that there is significantly more variation across potential host countries for a company based in a specific home country compared to the variation across potential home countries of an investment taking place in a specific host country, i.e. standard deviation for outbound investment is higher than that for inbound investment. The Base Case therefore suggests that the EU tax regime is closer to exhibiting capital import neutrality than capital export neutrality. As underlined in section 6.3, the absence of capital export neutrality can give rise to economic inefficiencies.

at 7

Outbound

5.82.61.21.23.06.06.4

ight line

d .

Dev.Inbound

Stan

2 .00 .50 .30 .31 .12 .02 .0s stra

ible

u b s i d i a ryOutbound

ng

e Se28.729.830.122.628.028.730.6

Inbound) inta

. (c

EATR

Averag28.729.830.122.628.028.730.6

i n a n c i n g thOutbound at 4%

0.60.30.30.30.50.60.7

a y

of F

d .

Dev.Inboundight line

t W

Stan

0 .40 .10 .00 .00 .30 .40 .4 B

cien

s stra

i talOutboundndix

Effi

e5.96.36.46.06.15.96.5ppe

o

st TaxInbound

MCost of Cap

Averag5 .96 .36 .46 .06 .15 .96 .5

e a r

• s. 

  (b) buildinge t out in A

s s

o n taxOutbound

5.52.91.21.23.15.76.0e r

7 ya te

orati

d .

Dev. ov

Inbound

.

St an

1 .81 .00 .90 .71 .21 .81 .8e c i a tion r

ight linepr

ts of corp

enOutboundic de

em30.630.730.723.029.130.532.4 and 35%

e s stra

fin a n c e175

c elInbound

e stiu b s i d i a ry

30.630.730.723.029.130.532.4e e n 25%t

2 tim ec onom

o ma r n i n g sb s i d i a rye Setwe aue

tr

Domestics belanc

s of dt su

ont atese s t i c e

cienEATR

Averag29.229.930.022.428.129.231.1a te ba

domi n a n ce of thOutbound (3 )

r re g a r d ing

c lining

u l a tib er S effi

Approximation or harmonisation of tax rates

The first four simulations consider cases of approximation or harmonisation of the statutory tax rates. At one extreme, a full harmonisation of the rates is considered (scenarios 2 and 3). This implies setting a common statutory rate, including surcharges and local taxes, at the same level in all the Member States. However, an approximation of tax rates could exclude local taxes. This case of partial harmonisation is considered in scenario 1. At the other extreme, a limited approximation of the rates via the setting of a band of permitted tax rates is considered (scenario 4).

Three main conclusions can be drawn from this set of simulations :

· First, unsurprisingly, an approximation of the statutory rates leads to a reduction in the dispersion of

EATRs and the cost of capital across the Member States for all the types of investment considered in the four simulations. This result is mostly due to the fact that those countries with the more extreme average costs of capital move towards the middle of the distribution, with a consequent reduction in the standard deviation.

As the various simulations show, the changes in the EATR dispersion tend to be greater than changes in the dispersion of the cost of capital. The scale of the differences in the impact on the cost of capital and the EATR reflect the differences in these two measures, discussed above. In particular, since theEATR measures the effect of tax on a more profitable investment, the importance of allowances is rather less than it is for the cost of capital, and so the impact of the statutory tax rate is correspondingly greater.

This result is fundamental. Indeed, concretely it implies that a consequence of the approximation of the statutory rates in the EU is an increased capital import neutrality. In other words, there is less variation in the costs of capital and the EATRs faced by parents from alternative home countries choosing to locate a subsidiary in the same host country. Moreover, the differences between the potential locations faced by a parent in a specific country in simulations 1 to 4 are lower than under the current system, implying also a movement towards capital export neutrality.

To illustrate why this happens suppose there is complete harmonisation of the tax rate in the EU. If all countries have the same statutory tax rate, then the distinction between, say, taxing foreign source dividend income with a limited credit system as opposed to an exemption system would disappear. In the absence of personal taxes, equalising statutory corporate tax rates is therefore close to introducing source country taxation for equity financed investment - with each source (ie. host) country having the same rate of tax. This clearly corresponds to a move towards capital import neutrality. It is worth noting, however, that home country taxation still matters in some cases. For example, several countries disallow interest payments made by the parent if the loan is used to finance outbound investment. Also, some countries impose local taxes on interest receipts from the subsidiary. Finally, some countries do not operate a pure exemption system.

The comparison of scenario 1 with scenario 2 easily illustrates that result. Indeed, one can see that imposing identical local taxes in addition to a basic common corporate tax rate, moves the overall position within the EU still closer to source-based taxation, as there is then no difference in statutory tax rates between countries. As would be expected, the results are broadly similar in the two cases. However, the most striking difference between simulations 1 and 2 is that the standard deviation ofEATRs across host countries for outbound investment falls further. This again reflects the greater dependence of the EATR on the statutory tax rate. While significant differences in the cost of capital across potential host locations is still strongly influenced by the definition of allowances (which are still allowed to differ between countries), allowances are rather less important for the EATR.

· Third, in the partial analytical framework of this study, driving down an already common tax rate

would have little impact on economic efficiency. It is mostly by reducing the dispersion of effective tax rates that gains in economic efficiency can be realised.

This is illustrated by a comparison of scenarios 2 and 3. The only difference between the two simulations lies in the statutory rate which is imposed to Member States (25% instead of 33,84%). Unsurprisingly, choosing a lower rate reduces the average cost of capital and the average EATR for all forms of investment: domestic, inbound and outbound. For example, the average domestic cost of capital across all countries falls from 6.4 to 6.0, and the average domestic EATR falls from 30.0% to 22.4%. Similar falls occur for inbound and outbound investment. However, the dispersions of both measures is virtually unchanged from the previous case of a higher statutory tax rate. This is true of the dispersion of domestic measures across each countries, the dispersion across possible locations for outbound investment and the dispersion across possible home countries for inbound investment (although in this case there is a small reduction in the dispersion of the average EATR for the case of the average source of subsidiary financing).

Comparing simulation 3 with simulation 2 is nevertheless illuminating. Indeed, a large part of the discussion in this study is concerned with economic inefficiencies arising as a result of differences in taxation between types of investment, or between countries. A process of tax competition which simply drove down an already common tax rate (for example, comparing this simulation with the previous one) would have little impact on economic efficiency: it is only by reducing the dispersion of effective tax rates that gains in economic efficiency can be realised.

Harmonisation of capital allowances

A number of the remedies to tax obstacles suggest an approximation or even a harmonization of the tax bases across the Member States. This is in particular so in the case of comprehensive options, such as those presented in Part IV. It is therefore interesting to examine the impact of harmonising the main elements of the tax base taken into account in the model, namely the rules for depreciation of assets and for inventories (see Box 4 in section 3.2).

Simulation 5 presents the impact of a harmonization of capital allowances. Allowances are assumed to take the following values in every EU Member State :

· machinery - declining balance at 2 times straight line over 7 years

· buildings - straight line at 4%

· intangibles straight line at 7 years life

· inventories LIFO valuation allowed

· financial assets - zero.

All other aspects of EU tax regimes - including the statutory rate and other aspects of the tax base - are as in the existing systems discussed above. This falls well short of proposals to create a consolidated tax base throughout the EU, where any individual company would need to calculate its taxable profits only according to one set of rules.

The results of this simulation show that harmonising the capital allowances has almost no impact on the average cost of capital or the average EATR. The averages and standard deviations shown in Table 29 are almost identical to the Base Case. If anything there is a slightly greater dispersion of EATRs across both host and home countries. This may reflect that tax regimes are designed as a whole - a high tax rate tends to go with a more generous structure of allowances and vice versa, a conclusion already made in section 4. Harmonising only allowances may increase differences in effective tax rates between countries, unless such a reform is accompanied by an approximation in the tax rates. Furthermore, this reform has only small effects in any Member State. Part of the reason for this result is that allowance rates are already broadly similar throughout the EU.

average, EU tax regimes are relatively generous in their choice of capital allowances. The standard deviations also tend to be slightly higher. The conclusion from the previous simulation - no discernible gain in economic efficiency from harmonising the tax base - therefore would also apply even if capital allowances were set closer to true economic depreciation.

Box 10:

Domestic reforms with personal taxes

Simulations 1 to 6 have also been considered in the presence of personal taxes (see appendix F). As earlier in the report, it has been assumed that the company maximises the wealth of a resident, possibly tax-paying, shareholder, taking into account personal taxes on dividends, interest and capital gains.

The actual results presented below are based on an unweighted average over three types of shareholder (a zero-rated shareholder, a non-qualified top-rate shareholder, and a qualified top-rate shareholder) for the two following cases: (a) the parent distributes all dividend payments out of domestic source income, and (b) at the margin the parent distributes dividend payments from foreign source income.

As explained in section 4, compared to the Base Case, the costs of capital are lower on average, and the EATRs slightly higher, when personal taxes are taken into account. There also tend to be higher standard deviations especially for inbound investment. This is true for both cases (a) and (b). In fact, for case (a), unlike the case in Table 29, the standard deviation across potential home countries for inbound investment is now generally higher than that across potential host countries for outbound investment.

In general though, personal taxes have little effect on the impact of hypothetical policy scenarios to corporation tax. The most striking difference from Table 29 when personal taxes are introduced is that the standard deviation for inbound investment does not generally fall when corporation tax rates are harmonised within the EU. The reason for this result is clear: there is still substantial variation due to personal taxes being different across home countries. In other words, tax systems have to be considered as a whole. Simply harmonising one element of the tax system, without taking into account its relationship with the other elements thereof, notably personal taxes, could lead to inconsistencies.

Scenarios involving international elements of corporation tax

Outbound

5.85.82.72.76.37.5

d .

Dev.Inbound

Stan

2 .02 .12 .17 .11 .42 .7

u b s i d i a ryOutbound

e Se28.728.730.927.328.330.5

Inbound

EATR

Averag28.728.730.927.328.330.5

i n a n c i n g thOutbound

0.60.60.70.80.60.9

a y

of F

d .

Dev.Inbound

t W

Stan

0 .40 .40 .40 .50 .40 .8

cien

i talOutbound

Effi

e5.95.95.95.75.96.7

o

st TaxInbound

o n taxMCost of Cap

Averag5 .95 .95 .95 .75 .96 .7

Outbound

orati

5.55.52.21.96.17.5

d .

Dev.Inbound

ts of corp

enSt an

1 .81 .82 .06 .50 .92 .6

emOutbound

30.630.533.229.730.031.1

fin a n c e180

o n a l el

atiu b s i d i a ryInbound

30.630.533.229.730.031.1

n terna r n i n g sb s i d i a rye Se

Domestic

s of it su

ont atese s t i c e

cienEATR

Averag29.229.229.229.229.229.2

domi n a n

ce of thOutbound

u l a tib er S effi

Abolition of withholding taxes on interest

Abolishing withholding taxes on the payment of interest from a subsidiary to its EU parent has almost no effect on any of the measures presented in this report. As can be seen from Table 30 (simulation 7), there is virtually no change in any of the averages or standard deviations in this simulation.

The reason for this lies in the tax treatment of the parent in the home country. All EU Member States tax interest receipts from EU subsidiaries in the hands of the parent. All Member States use a limited credit system, so that taxes paid in the host country are credited against home country taxation. However, all Member States also permit interest payments to be deductible from corporation tax. So the only tax which the interest payments may face in the host country is a withholding tax when the interest is paid to the parent. In Germany, however, half of interest payments are subject to trade tax. This clearly limits the advantages of debt financing of a German subsidiary. But in virtually all cases, the rate of withholding tax is lower than the home country tax rate. As a result, the withholding tax does not increase the overall tax liability; it merely shifts revenue from the home country to the host country. But the measures presented here are not affected by which revenue authority receives the tax payment. As such, they are almost completely unaffected by the abolition of withholding taxes on interest payments from subsidiaries to parents.

There is one exception to this within the EU. Ireland has a 10% corporation tax rate, and three countries levy a withholding tax on payments to an Irish parent at a higher rate than this: Belgium (15%), Greece (20%) and Portugal (15%). In these cases, Ireland does not offer a full credit for the withholding taxes paid, and so these withholding taxes do affect the overall tax liability of the Irish parent. Analysis of the impact of this scenario on individual countries confirms that only these four countries are affected by the reform. An implication is that the same effect on effective tax rates in the EU could be generated by simply reducing these three withholding tax rates to 10% or less.

Harmonising the treatment of foreign dividends

Currently, only 3 Member States tax dividends received by a parent from an EU subsidiary on the basis of a limited credit system (Greece, Ireland and the UK). The other 12 Member States all use some form of an exemption system. Simulations 8 to 10 show the impact of an identical system for the 15 Member States, that is, either a credit system or an exemption system.

capital is independent of the taxation of the dividend paid by the subsidiary since the same tax rate is applied both to the cost and the return, it nets out of calculation.

This suggests that, for a marginal investment, reflected in the cost of capital, the international tax regime becomes a mixture of a residence-base and a source-base. Take the example of Ireland as a home country, for the average of possible sources of finance of the subsidiary. Where the subsidiary of an Irish parent is financed with retained earnings, the Irish tax rate is irrelevant, and so the cost of capital reflects only the host country tax system (which has not changed). But where the investment is financed by new equity provided by the parent, for example, the fact that there is a full credit system - and that the Irish tax rate is only 10% - implies that there is a significant reduction in the cost of capital.

These effects are partly reflected in results presented in Table 30. For outbound investment, there is a fall in the overall average cost of capital and of the average EATR across Member States in simulation 9 compared to the Base Case. At the same time a substantial fall in the standard deviation can be observed for the EATR. This is consistent with a move towards capital export neutrality. For inbound investment, the average cost of capital and EATR also diminish, but the standard deviation increases. Therefore, introducing a full credit system leads to less capital import neutrality.

Box 11:

Impact of a limited credit system (simulation 8)

Where the home country tax rate is lower (higher) than the host country tax rate, a limited credit system will not result in any further tax (reimbursement) in the home country. In this case, it therefore has the same impact as an exemption system. The main effects of the introduction of a limited credit system for the 15 Member States is therefore to increase the tax liability on the receipt of dividends from EU subsidiaries of parents in home countries with high statutory tax rates.

This scenario therefore pushes the overall system towards capital export neutrality, in that the gap in the impact of tax between domestic and outbound investment is narrowed. This does not show up clearly in the average results of Table 30, however. This is because cross-border investment (based on the average of the subsidiary sources of finance) already has, on average, a higher cost of capital and a higher EATR than domestic investment. Any increase in the average effective tax rate for cross border investment therefore increases this disparity.

definition of the home country tax base to its EU-wide profits. The tax base for each company would be allocated between different countries, which would apply their own tax rates to their allocated tax base.

In order to do this, it is necessary to make some assumptions about how the subsidiary is taxed. The first stage is that the profits of each subsidiary are consolidated in the whole groups' EU-wide profits. For this purpose, the relevant tax base rules are treated as those of the parent's home country. However, a further assumption is that there is no difficulty in allocating the EU-wide profit to individual countries. That is, it is assumed that the whole of the taxable profit earned by the subsidiary is allocated back to the host country, to be taxed at the host country corporation tax rate. There is then no further tax on repatriation of dividends or interest from subsidiary to parent; nor is there any further tax on the corporation on payment of dividends to the shareholders. Essentially, then, the host country tax rate is applied to the tax base as defined by the home country. Moreover, due to the consolidation process, interest payments from the subsidiary to the parent company are not deductible in the host country. This tends to increase the host country's tax base.

It is worth noting that two advantages of Home State Taxation are that it permits loss consolidation and that any parent would have to calculate its EU-wide profits according to only one set of rules. However, these advantages are not captured by the model used in this report, since it makes no allowance for the compliance and administrative costs of implementing taxes and considers only profitable situations.

To understand the impact of Home State Taxation, it is useful to consider the position if the proposal were taken one step further. Suppose instead that the home country tax rate was applied to the EU-wide profit, rather than the host country tax rate. In this case, the tax regime in the host country would be irrelevant; in effect there would be residence taxation at the corporate level; hence there would be capital export neutrality.

As Table 30 (simulation 11) makes clear, however, applying the tax rate of the host country makes a substantial difference. Instead of achieving capital export neutrality, the resulting tax regime is actually further from capital export neutrality than the existing regime. That is, the average dispersion of effective tax rates facing parent companies on outbound investment within the EU is higher than in the Base Case. This is true on every measure. Thus, the standard deviation of the costs of capital is higher - both for the tax efficient and average form of financing of the subsidiary. And the average dispersion of the EATR is also higher in both cases.

Scenarios involving the relationship between corporate and personal taxes

Much of the debate on the co-ordination of taxes on capital in the EU in the past has been concerned with the relationship between personal and corporate taxes. In a closed economy, this is a vital ingredient of the overall effective tax rate, since the overall level of saving and investment should depend on both forms of taxation.

But, as argued elsewhere in this study, the role of personal taxes for cross-border flows of capital is much less clear. Indeed

personal taxes levied on domestic shareholders may well affect the cost of capital and the EATR on domestic and outbound investment. But with an open capital market, there is no particular reason why the company may not be owned by foreign shareholders. Further, as already noted in Box 8, personal taxes in the home country do not play a very significant role in affecting the main conclusions of the analysis. This is mainly because personal taxes typically apply to all forms of profit, and hence do not discriminate between domestic and outbound investment.

This section nevertheless briefly reviews the impact on the measures described above of harmonising the relationship between corporate and personal taxes. Table 31 presents results for four hypothetical policy scenarios. It should be noted that the Base Case includes personal taxes, contrary to the Base Case in Tables 29 and 30.

The first hypothetical scenario is to introduce a classical relationship between the corporate and personal taxes (simulation 12). Under a pure classical system, company profits are taxed twice - once at the level of the firm, and once when distributed as dividends. Top-rated shareholders would pay the full top rate of income tax on dividend income. This is higher than the rate actually levied in several countries, since such countries reduce the double taxation of dividends by reducing the rate of personal income tax. However, it is assumed that zero-rated shareholders continue to pay no tax.

Given that many EU countries either allow a tax credit associated with a dividend payment, or levy a lower rate of income tax on dividends, harmonisation based on a classical system generally increases the size of the tax burden, unless tax rates are adjusted. This is reflected in the impact on the average EATRs shown in Table 31. Such a harmonisation tends to reduce the dispersion of the average EATRs across countries, and to reduce the average standard deviations for outbound and inbound investment. Thus, to the extent to which the overall tax systems become more similar to each other, there is a movement towards both capital import neutrality and capital export neutrality. Again, however, it should be noted that the importance of this result depends on the extent to which such personal tax rates are relevant in

an international context.

However, it should be noted that with a classical system, the average costs of capital tend to be lower than in the Base Case. This is because it is assumed within the simulation that interest receipts are also taxed at the top personal tax rate. Where this raises the existing personal tax rate on interest income, then because the post-tax return to lending, i.e. the alternative use of funds, falls, the required post-tax return to investment in equity also falls, and hence the cost of capital tends to fall.

Outbound

5.13.96.06.16.9

d .

Dev.Inbound

Stan

5 .42 .96 .82 .90 .0

u b s i d i a ryOutbound

e Se32.541.923.321.333.0

Inbound

EATR

Averag32.541.923.321.333.0

i n a n c i n g thOutbound

0.60.60.60.60.8

a y

of Fd .

Dev.

Inbound

t W

St an

0 .70 .50 .50 .50 .0

cien

i talOutbound

Effi

e4.84.44.04.07.4

o

st TaxInbound

MCost of Cap

Averag4 .84 .44 .04 .07 .4

Outbound

4.93.75.75.96.9

d .

Dev.Inbound

St an

5 .62 .77 .12 .80 .0

Outbound

a x e s34.043.125.023.133.1

fin a n c e186

T

u b s i d i a ryInbound

34.043.125.023.133.1

ersonala r n i n g s

b s i d i a rye Se

e s t i c et suDomestic

e and PcienEATR

Averag33.042.224.221.833.1

domi n a n

ce of thOutbound

of effi

This scenario involves a reduction of the overall tax liabilities, as measured by the EATR. The cost of capital is also reduced, even though this reform only affects the cost of capital when the parent finances the investment by new equity. However, the dispersion in either measure in the standard deviations on inbound and outbound investment, is generally at least as high as in the Base Case. This should not be surprising. Consider, for example, the variation in the EATR faced on inbound investment into the same host country by different parents. With a full imputation system in the home country, the effective tax rate in the home country more closely reflects the personal tax rate than the corporate tax rate. Since these tend to vary more between countries, so the average standard deviation of the EATR on inbound investment tends to increase relative to the Base Case.

A similar position is true of the next hypothetical scenario: a classical relationship between the corporate tax and the personal tax, but a personal tax rate on dividend income set at only 50% of the shareholder's normal income tax rate (simulation 14). Following OECD (1991), this is known as a shareholder relief system. On average, this generates similar costs of capital and EATRs as the full imputation system. However, compared to the full imputation system the overall effective tax rate depends less on the personal tax rate of the shareholder and more on the corporation tax rate. As a result, the dispersion of the EATR for inbound investment into a particular host country by different parents is lower than in the case of the full imputation system.

The final hypothetical policy scenario is based on a proposal considered by the US Treasury in 1992. This proposal - the comprehensive business income tax - was to abolish interest deductibility, and hence put equity and debt finance on the same footing (simulation 15). The reform analysed here extends this basic version to eliminate personal taxes entirely. In effect there is therefore only corporate level taxation, levied on profit before interest payments. Further, source country taxation is imposed by making each country exempt dividends and interest receipts from a subsidiary in the EU. Hence the single level of tax applies essentially in the host country.

As might be expected in such a scenario, capital import neutrality is completely achieved. All subsidiaries in a given host country face the same tax regime, irrespective of the nationality of the parent companies. Further, on average, the cost of capital for domestic investment is the same as the cost of capital for cross-border investment, and the same is true for the EATR. However, the reform also moves the EU tax regime away from capital export neutrality. That is, the dispersion in effective tax rates across possible locations for investment available to a parent company is greater than in the Base Case.

· By contrast, no other scenario would have such a significant effect. For example, introducing a

common tax base while leaving tax rates unchanged tends, if anything, to increase the dispersion in effective tax rates.

· Since withholding taxes on dividends between subsidiaries and their parents have been abolished

within the EU, the international features of corporation taxes do not play a significant role in increasing distortions. Introducing a common means of taxing foreign source income, for example, has little impact on the dispersion of effective tax rates.

· Similarly, introducing a common form of integration of corporate and personal taxes in each

Member State does not tend to reduce the dispersion of effective tax rates between Member States.

Box 5

Tax Analyser: Impact of hypothetical tax policy scenarios in the EU

As in the case of the hypothetical investment approach, the purpose of simulations is to identify the weight of the different tax drivers in the effective average tax burden, i.e. to compute the relative weight of the tax bases, the tax rates, the different types of taxes and the corporation tax system in the EATR separately. Therefore, the scenarios are divided into three categories: elements of the tax base, the corporation tax rate including local taxes and the corporation tax system. By setting just one element of the different tax regimes equal across countries, it is possible to identify the effect of this particular element on the level and also on the variation of the effective tax burdens and, hence on the possible distortion of competition within the countries under consideration.

Each simulation considers only the medium-sized company in the base case using typical data for the manufacturing sector. Except for Germany, the scenarios are based-as in the previous boxes Tax Analyser- on the tax regimes which were effective in the fiscal year 1999.

• 8. 

  Common book reserves for bad debts. Future guarantee payments within next 2 years, 2% of annual turnover.
• 9. 

  Common overall tax base (IAS). Depreciation methods according to (1)-(3) above, depreciation periods are 5

years for intangibles, 40 and 50 years for buildings, 5 to 10 years for tangible fixed assets, valuation of inventories according to (5), allocation of pension costs according to (7), no provisions for bad debts.

B. Corporation tax rate and local taxes

• 10. 

  Common corporation tax rate, including surcharges but excluding local taxes, at EU average of 32.28%.
• 11. 

  Common corporation tax rate, including surcharges and local profit taxes, at EU average of 33.84%.
• 12. 

  Common corporation tax rate of 25%, including surcharges and local profit taxes.
• 13. 

  Common corporation tax rate of 25%, including surcharges and all (profit and non-profit) local taxes.

C. Corporation tax system

• 14. 

  Common corporation tax system. Classical system (only corporate taxes).
• 15. 

  Common corporation tax system. Classical system (corporate and personal taxes).
• A) 

  Scenario involving the corporation tax base

The following table presents simulations referring to the corporation tax base. The first 8 of these simulations consider only a single element of the tax base. By contrast simulation 9 analyses the effects of a uniform tax base based on the "International Accounting Standards". All other elements of the tax regimes are unchanged.

TABLE FResults for simulations involving elements of the corporation tax base

-effective average tax rates

• - 

  only corporation taxes

EU-5

FDIRLNLUKEU-5

USA

AverageStand.

Dev.

Base case39.730.18.324.021.024.610.429.7

• 1. 

  Common depreciation on

intangibles39.730.18.324.021.024.610.429.7

• 2. 

  Common depreciation on

buildings40.229.38.323.621.024.510.428.8

• 3. 

  Common straight-line

depreciation on tangible fixed assets44.130.68.424.521.425.811.732.3

• 4. 

  Common declining

balance depreciation on tangible fixed assets41.028.47.922.319.723.910.929.8

• 5. 

  Common valuation of

inventories38.029.77.921.819.223.310.127.5

• 6. 

  Common pension scheme

(book reserve)44.530.19.227.124.027.011.334.1

• 7. 

  Common pension scheme

(pension fund)39.726.28.324.021.023.810.129.7

• 8. 

  Common book reserves

for bad debts36.226.67.921.418.822.29.326.6

standard deviation (by 0.3). Full cost tends to increase the tax burden and LIFO tends to decrease the tax burden. Obviously, for the base case model firm, the increasing effect of applying full cost is more than compensated for by the decreasing effect of applying LIFO allocation. Moreover, the calculation of production costs on a full cost basis tends to smooth the differences between different capital allowances and other elements of the production costs (such as pension costs) if the finished goods are stored for a certain period of time. This explains the decrease in the standard deviation.

The harmonisation of occupational pension schemes' allocation of pension costs considers two different

scenarios: (6) a common regime for building up a pension reserve (book reserve) as in Germany and (7)

a common regime for a funded system that allows the deduction of annual (periodical) payments to a pension fund as prevails in the Anglo-Saxon countries . In the case of book reserves , the national EATR, except in Germany, the EU-5 average EATR (by 2.4 percentage points) and the standard deviation (by 0.9) increase. The reason for this considerable increase is a broader tax base in each country except Germany and the high portion of pension costs in the total costs of the model firm. A common system for pension funds would only affect the position of Germany since it is already applied in the other countries. Therefore, the impact on both the EU-5 average EATR and the standard deviation is only minor.

It is always very arbitrary to specify conditions under which bad debts occur. Therefore, bad debts were not considered in the base case at all. The European Tax Analyzer model, however, can account for bad debts. To analyse the effect of such provisions on the tax burden, the build-up of a book reserve for bad debts is now allowed in all countries. In doing so it is assumed future warranties which amount to 2% of the annual turnover of the model firm have to be fulfilled within two years. Accounting for a book reserve for bad debts under these conditions decreases both the national EATR and the EU-5 averageEATR (by 2.4 percentage points). The decrease is highest in France and Germany (3.5 percentage points) and lowest in Ireland (0.4 percentage points). This result clearly reflects the (corporation) tax savings due to the deduction of the annual contributions to the book reserve from the tax base. The amount of tax saving increases with the statutory tax rate on profits, which is highest in France (40%) and Germany (39.3%) and lowest in Ireland (10%). The asymmetric decrease of the national EATR is also reflected by the lower standard deviation (9.3 compared to 10.4 in the base case). As a general conclusion from this example, it seems likely that both the national EATR level as well as the dispersion of the EATR across countries are lower compared to the base case if book reserves for bad debts are taken into account. However, with respect to the extreme assumptions in our example the bad debts amount to 2% of the annual turnover a considerable variation of the EATR still remains.

The results presented in Table F indicate that, with the exception of Germany, the national EATR would rise if the IAS were relevant for the determination of taxable profits. As a consequence, the EU-5 average EATR increases by 1.6 percentage points. Thus, by taking the IAS as a benchmark, we can conclude that the national accounting provisions are currently more generous. However, since the changes in the EATR in Germany, Ireland, the Netherlands and the UK are only minor, it is only France that seems to be in a relatively favourable position with respect to the tax base. Therefore, the change in the EU-5 average EATR can be above all attributed to the large change in France. As far as the Anglo- Saxon countries and the Netherlands are concerned, the minor changes of the EATR can be attributed to the fact that the tax bases in these countries already correspond to the IAS to a greater extent than the tax bases in European continental states (e.g. France), except in the USA. Germany's result also suggests that the tax base corresponds with the IAS. Since the German EATR decreases, the national tax accounting rules are even a bit less generous. Although the broadening of the tax base according to international standards was one of the aims of the German tax reform, a lot of differences, which would compensate for each other considerably if the IAS became relevant still exist. However, with respect to these compensating effects, it is also clear that it is not sufficient just to compare the individual elements of the tax base (e.g. depreciation rules or provisions) when analysing the impact of the tax base on the effective tax burden.

The most striking result of this harmonisation scenario is, however, the increasing standard deviation (by 2.0) which is greater than in all previous simulations. This indicates that with a harmonised overall corporation tax base the differences between the national EATR would, ceteris paribus, not only remain, but that these differences would even increase since the variation in the EATR across countries increases. The remaining variations in the EATR can be attributed to the different statutory tax rates, local taxes and the corporation tax systems. In the case of a uniform tax base, the real effects resulting solely from these differences are clearly demonstrated. As pointed out earlier, higher tax rates tend to be combined with lower tax bases and vice versa (see the first box Tax Analyser). This correlation might explain the increasing dispersion of the EATR when tax bases are harmonised: in such instances the compensating element (i.e. a more generous tax base) disappears.

• B) 

  Scenarios involving the corporation tax rates and local taxes

TABLE GResults for simulations involving tax rates and local taxes

• - 

  effective average tax rates
• - 

  only corporation taxes

FDIRLNLUKEU-5EU-5

USA

AverageStand.

Dev.

Base case39.730.18.324.021.024.610.429.7

• 10. 

  Common CT rate, at EU

mean36.734.123.621.822.727.86,328.4

• 11. 

  Common CT rate incl.

local profit taxes, at EU mean37.124.024.523.124.026.55.325.3

• 12. 

  Common CT rate of 25%

incl. local profit taxes33.617.018.316.617.520.66.519.7

• 13. 

  Common CT rate of 25%

incl. all local taxes9.617.016.216.214.814.82.716.2

Simulation 10 introduces a common corporation tax rate of 32.28% including all surcharges levied in France and Germany. The rate of 32.28% is the average rate across all 15 EU Member States for the fiscal year 1999 - except for Germany where the rates applying after the tax reform in 2001 have been used. However, this simulation does not imply that in all Member States all profits will be taxed at the same rates. Since local profit taxes are still levied, differences will remain.

Introducing such a common corporation tax rate increases the EU-5 average EATR significantly, by 3.2 percentage points. At the same time, however, there is also a considerable reduction of the standard deviation from 10.4 to 6.3 indicating less variation of the EATR across EU Member States. In fact, since the EATR does not change significantly in France, the Netherlands and the UK, these changes are mainly caused by Germany and Ireland. The EATR in Germany increases by 4.0 percentage points. Since there is an extra burden imposed by the trade tax on income in Germany, the statutory tax rate on profits would rise from 39.3% to 44.2%. However, the most significant change is in Ireland with an increase of the EATR by 15.3 percentage points from 8.3% to 23.6%. This is because the advantage of the 10% corporation tax rate would now disappear.

corporation tax. The uniform tax rate chosen is 33.84%, which represents the average statutory tax rate across the 15 EU Member States.

Since only Germany levies a local profit tax within the EU-5 Member States, this harmonisation scenario would significantly reduce the EATR in Germany. By contrast, there would only be moderate increases of the EATR in the other countries. Germany would improve two positions to second place, bringing it to a level equal to that of the UK. Altogether, the EU-5 average EATR would be reduced moderately and the standard deviation of the EATR across countries would fall further. The new value of 5.3 is almost 50% lower than the standard deviation of the EATR for the existing tax regimes (base case). Compared to the existing tax regimes, introducing a harmonised statutory tax rate on profits would therefore substantially reduce distortions resulting from differences between the EATR for domestic investment in Europe.

However, not all distortions would disappear. The remaining differences can be attributed to the tax bases, the local non-profit taxes and the corporation tax systems. France in particular, would still be in a very disadvantageous position with respect to the levy of non-profit taxes.

Compared to the previous case, simulation 12 simply reduces the common statutory tax rate on profits from the EU average rate of 33.84% to 25%. As a consequence of the lower tax rate all of the nationalEATRs as well as the EU-5 average EATR would be further reduced. However, with respect to the increasing standard deviation, the more striking result of this simulation is that the dispersion of theEATR rises. This happens because the tax saving due to the deductibility of local non-profit taxes becomes smaller as the statutory tax rate on profit decreases. In other words, the effects of the different non-profit tax levels emerge more obviously if the statutory tax rate on profits is reduced.

The result of this simulation is also important since it reveals the effects of the tax rate on the variation in the effective tax burdens. The comparison with the previous simulation makes clear that a further reduction of a statutory tax rate on profits, which is already harmonised, to a level significantly lower than the average across countries, will only reduce the average effective tax burden. However, the economic distortion will increase, since the standard deviation rises.

In addition to measures considered in the previous case, simulation number 13 abolishes non-profit taxes (e.g. real property tax). This implies that the only tax that exists in each country is a tax on profits which is levied at a uniform rate of 25%. Since there are no other taxes, the remaining differences between theEATR are the result of the different tax bases and the corporation tax systems. As a result of this simulation, all of the national EATRs and the EU-5 average EATR would be further reduced. Moreover, the standard deviation of the EATR falls to 2.7, by far the lowest value obtained in any of the 15 simulations.

been introduced in a separate step. In the case of personal taxes, the amount of dividends is the same as in the base case and there is no "shareholder relief" on dividend income.

TABLE HResults for corporation tax system scenarios

• - 

  effective average tax rates

• - 

  corporate and personal taxes

EU-5

FDIRLNLUKEU-5

USA

AverageStand.

Dev.

Base case (only corporate taxes)

39.730.18.324.021.024.610.429.7

Base case (corporate and personal taxes)

48.831.017.232.025.630.910.432.0

• 14. 

  Common CT system,

classical system (only corporate taxes)48.030.18.324.021.026.313.029.7

• 15. 

  Common CT system,

classical system (corpo- rate and personal taxes)55.732.017.237.426.633.812.832.0

At the corporate level the introduction of a common classical system would only affect the EATR in France. This is because all of the countries considered here except France already apply a kind of classical system at the corporate level. The abolition of the imputation system, in general, would increase the EATR. The reason is that if the tax credit is denied, a relatively higher cash distribution will be required to pay the same amount of dividends as before. As a result of the increasing French EATR, both the average EU-5 EATR and the standard deviation of the EATR across countries rises.

• D) 

  Concluding remarks

A major finding of the analysis of the effective tax burdens by means of the Tax Analyser model is that all tax regimes seem to be designed as more or less integrated systems. This means that there is a particular relationship between the tax rate, the tax base and the corporation tax system. As a broad conclusion, it is possible to say that, in general, a higher statutory tax rate on profits correlates with a lower taxable base and vice versa.

The simulations presented in this box help to understand the weights of each of the most important elements of a tax regime in the effective tax burden.

Introducing a common corporate tax base (simulation 9) clearly helps to achieve more transparency in the calculation of effective tax burdens. However, the outcome of this simulation indicates that such a change would result in increasing values both for the average effective tax burden and what seems more important the dispersion of the EATR across EU Member States.

Similarly, a common classical corporation tax system (simulation 14) tends to increase both the average effective tax burden in Europe and the dispersion of the EATR across EU Member States.

By contrast, the introduction of a common statutory tax rate on profits (simulation 11) would significantly reduce the dispersion of the EATR across the EU Member States. None of the other hypothetical scenarios considered here reduced the variation of the effective tax burdens in a comparable way. The effects on the EATR of a common statutory tax rate on profits depends on level of the tax rate.

In our example, the average EATR increases. Although a lower common tax rate would reduce the average EATR, the dispersion of the EATR across countries tends to rise again. Therefore, reducing statutory tax rates on profits significantly does not simultaneously ensure greater neutrality towards taxation.

SOME EFFECTS OF TAX OPTIMISATION BY MEANS OF FINANCIAL INTERMEDIARIES ON THE

EFFECTIVE TAX RATES ON TRANSNATIONAL INVESTMENTS BY GERMAN AND UK COMPANIES

Introductory remarks

The potential distortions highlighted in the analysis of cross-border investments indicate that there can be considerable incentive for companies to alter their behaviour in order to minimise their global tax burden.

Section 6.3.3 showed that companies can considerably minimise their effective tax burden if they choose selected forms of finance for the subsidiaries, the most tax convenient form of finance depending on the result of the interaction of the taxation regimes of the home/host countries. The results of the analysis indicate that this situation corresponds to a degradation of capital export and capital import neutralities.

Companies may also use more complex financial arrangements and group structures in order to minimise their effective tax burdens. The area of financing offers many possibilities, in particular for multinational companies. In general, the implementation of an intermediary financial company is advantageous if the relevant income bears a lower tax burden compared with the direct financing of a foreign subsidiary by the parent company.

Therefore, the purpose of this section is to understand to what extent the tax optimisation strategy of companies -by means of an intermediary financial company- affects their effective tax burden and, in general, the results arising from section 6, in terms of tax induced distortions in the allocation of resources.

The most tax efficient strategy for cross border financial arrangements through financial intermediaries depends on the tax burden in all countries involved and on the provision of the tax treaties between these countries. It is therefore obvious that there is no universally valid tax optimisation strategy for international financing (see Box 10).

In order to work out the most relevant tax driven factors influencing this strategy and keeping the analysis still manageable, Germany and the UK only are taken as examples for the location of the parent company (the subsidiary being located in all Member States). For the intermediary financial companies, the case of a Belgian co-ordination centre and a Dutch finance company are considered.

conditions as a place of location for foreign investors. It can, therefore help to understand whether removing these forms of financial intermediaries, for example in the context of the Code of Conduct for business taxation, helps to solve the problem of tax-induced resources misallocation.

Box 12:

Possible financial arrangements

Transfer of money from parent to financial intermediary / from financial intermediary to subsidiary

Taxation at the level of the

SubsidiaryFinancial intermediaryParent

Symmetric financing

New equity /Taxation of profitsTaxation of dividends, international double taxation is avoided by exemption or credit with limitationTaxation of dividends, international double taxation is avoided by exemption or credit with limitation

New equity

Debt /No taxation of profits as interest on loan is deductibleIn principle no taxable income as the interest received from the subsidiary is decreased by the interest paid to the parentTaxation of interest received from the financial intermediary

Debt

Asymmetric financing

New equity /Taxation of profitsTaxation of dividends, international double taxation is avoided by exemption or credit with limitationTaxation of interest received from the financial intermediary

Debt

• - 

  Debt / Debt: A financial intermediary that is used as a conduit company for interest payments typically gains an

advantage from the reduction of withholding taxes on interest (treaty shopping). This can be achieved if the tax treaty between the state of residence of the subsidiary and the intermediary company allows to collect lower or even no withholding taxes on interest compared to the tax treaty between the state of residence of the subsidiary an the parent company.

• - 

  New equity / Debt: Where debt capital is transformed into new equity capital there is an advantage to an intermediary

company if the tax saving from the interest deduction at the level of the intermediary is higher than the tax burden on the interest receipts at the level of the parent.

• - 

  Debt / New equity: The transformation of equity into debt financing offers advantages if the financial intermediary

pays taxes at a lower rate than the subsidiary would have to pay on its profits and the parent company would have to pay on receipt of the interest. Moreover, the ultimate repatriation of the equity funds to the parent should bear no further tax liability (e.g. dividends are exempt from tax at the level of the parent).

The German parent's approach for optimising international financial arrangements

(1999)

The legal framework of the analysis

In 1999 Germany has a comparably high national tax rate on profits. International double taxation of dividends is eliminated by the exemption method. One might reasonably argue therefore that the principle aim of German multinationals is to establish financial intermediaries as base companies (asymmetric financing) in order to shelter the low taxed income from taxation in Germany. The predominant way of financing is that the German parent contributes new equity funds to the intermediary company and the intermediary company transforms these equity funds into debt financing of the subsidiary (i.e. new equity / debt from box 12). Figure 3 summarises the sources of finance and the resulting cash flows.

From the perspective of a German parent company Belgium, Ireland and the Netherlands are the most important countries for the location of financial intermediaries. The following analysis concentrates on Belgium and the Netherlands which allows the fundamental principles of the German taxation of foreign financial intermediaries in 1999 to be highlighted.

Figure 3: German perspective of international financial arrangements

EU Subsidiary (M em ber States)

Debt

Interest

Financial Interm ediary

(Belgium C o-ordination Centre BCC

Dutch Financing Com pany DFC)

Dividends

Equity

EU Parent Com pany (G erm any)

D ebtIS ales

nterestE quityD iv.R etained

E arnings

of shares

Individual Shareholder (G erm any)

• A) 

  The case of a Belgian co-ordination centre

As a general rule, retained profits of the BCC are only taxable in Belgium (deferral principle) and, according to the Belgian-German tax treaty, distributed profits are exempt from corporation tax and trade tax at the level of the German parent.

The deferral principle is violated to a certain extent by the German Controlled Foreign Companies (CFC) legislation including Passive Foreign Investment Company (PFIC) rules. Accordingly, following the German "deemed dividend approach", retained profits of a foreign company resulting from passive income are attributed to the German parent company and are subject to corporation tax and trade tax if the shareholding in the financial intermediary is at least 50% (controlled foreign company) and the total (effective) tax burden on the profits is less than 30% (low tax jurisdiction).

As passive income is defined, inter alia, as income realised by the foreign financial intermediary from holding of liquid funds or the lending of equity capital received by a parent company, a BCC earns passive investment income. The German legislation explicitly provides, however, that the tax treaty provisions applicable to distributed dividends also apply to deemed dividends under CFC legislation. As the Belgian-German tax treaty does not contain an "activity proviso" the passive income of a BCC would be as a first step exempt from German taxation.

With regard to capital investment income, however, the German PFIC rules set out of force treaty provisions which provide the exemption method. To the extent that the capital investment income is stemming from the financing of a foreign based company which is engaged in active business, 60% of the investment income is treated as passive income and attributed to the German shareholder (i.e. the parent company). Although the amount of 60% is entirely subject to corporation tax it is exempt from trade tax. To mitigate double taxation an indirect foreign tax credit is granted to taxes paid by the BCC and to taxes withheld on investment income received by the BCC. The remaining 40% of the capital investment income are treated as ordinary passive income so that the participation exemption according to the Belgium-German tax treaty applies. Moreover, a genuine distribution of the profits (part of the 60% profits) of the financial intermediary company which have been already attributed to the German shareholder according to the PFIC rules are explicitly exempt from corporation tax.

Depending on the location of the subsidiary that is financed by the BCC the financial operation bears different tax burdens: The (minimum) statutory corporation tax rate including the solidarity levy is 42.2% (40% * 1.055) plus trade tax in the case of a German subsidiary and 25.32% (60% * 40% * 1.055) in the case of a foreign based subsidiary. Therefore a German parent does not benefit from a BCC with regard to the financing of its domestic subsidiaries. The tax burden will be the same. With regard to the financing of foreign subsidiary, however, the tax burden is significantly lower compared to both domestic subsidiaries and direct financing by the German parent. In the latter case the interest receipts from the foreign subsidiaries would also be subject to the statutory tax rate of 42.2% plus trade tax.

With regard to the taxation of profits of a DFC the general rule is that they are not taxable in the hands of the German shareholder as long as they are retained (deferral principle). In case of a distribution the dividends of a DFC are exempt form corporation tax and trade tax according to the Dutch-German tax treaty.

As the DFC in principle is always resident in the Netherlands and the effective tax burden usually is above 30%, there is no taxation in Germany either due to a German place of residence or according to the German PFIC rules. The latter rules are likely to apply, however, if the DFC makes contributions to a risk reserve resulting in an effective statutory corporation tax rate below 30%. The critical contribution is around 14.3% (35% * 85.7% = 29.995%). As any contribution above this amount would result in an even higher tax burden due to the PFIC rules compared to the situation in which no contributions to the risk reserve are made at all, the most tax efficient strategy would be to contribute such an amount so that the effective corporation tax rate is just above 30%.

Although this strategy bears certain risks as the total profits of the DFC could be treated as capital investment income it is assumed to prevail in the case where the DFC is granting loans to its subsidiaries. Moreover, it is assumed that the reserve can be released tax free (e.g. a purchase of participation is assumed) so that the Dutch corporation tax rate will be fixed at 30%.

Relevant economic measures: cost of capital and EATR of a German parent and its EU

subsidiaries

Tables 32, 33 and 34 compare the cost of capital and the EATR of a German parent and its EU subsidiaries both for the most tax efficient way of financing the subsidiaries by the parent and for the case of a financial intermediary. The financial intermediary can be either a BCC or a DFC, financed by the German parent with equity capital, and which transfer the money as debt capital to other group members located in the Member States

• 47. 

  For each way of financing averages and standard deviations are

calculated.

As a consequence of the high German statutory tax rate the most tax efficient direct way of financing anEU subsidiary from the perspective of a German parent is always profit retention of the subsidiary (see section 6).

applied (as shown above the effective tax rate is around 25%). All Member States except Ireland and Italy impose corporation tax rates above 25%.

Compared to a BCC a DFC might be expected to offer a less favourable tax effect as the relevant tax rate on the interest income is higher (as shown above, the effective tax rate is around 30%). For four Member States, however, the opposite turns out to be true. This result is explained by the levy of withholding taxes on interest payments in the host country of the subsidiary to the financial intermediary. In case of aDFC these withholding taxes are not relevant as they can be credited against Dutch corporation tax on the interest income. In case of a BCC, however, the withholding tax becomes relevant for that part of the underlying income that is not covered by the PFIC rules (i.e. 40%).

As only 60% of the income attributed to a BCC is taxable and 40% exempt, in Germany 40% of the withholding tax deducted cannot be credited against German corporation tax and thus becomes final. In the case of France, Greece, Portugal and Spain the 15% withholding tax on interest payments to a BCC overcompensates for the advantage of a lower effective statutory tax rate on profits as compared with aDFC.

The ranking of the host countries for the subsidiaries from the perspective of a German parent company differs between the cost of capital and the EATR.

In the case of a financial intermediary the cost of capital of a cross-border investment is determined by the (individual) tax base in the host country of the subsidiary and the (uniform) effective tax rate imposed on profits of the financial intermediary. Therefore, in the case of a marginal investment, Member States only compete as places of location in their rules for determine corporate profits (i.e. their tax bases). This is demonstrated by the BCC case where Belgium (4.5%) has a lower cost of capital than Luxembourg (4.6%) and Denmark (5.2%).

For a more profitable investment the EATR is relevant. Although with respect to the EATR the changes in the country ranking are only minor, one important aspect has to be considered. As the economic rent cannot be shifted from the subsidiary to the financial intermediary by means of financial arrangements based on the arm's-length-principle (the possible shifting by transfer prices is not considered here), the economic rent is always taxed at the statutory rate in the host country of the subsidiary. As the national tax rates on profits are lower in Denmark (EATR of 16.8%) and Luxembourg (EATR of 19.2%) both countries can improve their position in the ranking compared with Belgium (EATR of 21.2%).

ev.Stand. D

0 .88 .30 .55 .70 .45 .40.67 .20 .77 .90.65 .50 .55 .20.66 .7

Mean

5.519.05.318.45.518.95 .116.73.919.03 .718.23.918.73 .416.7

Kingdom

United

5 .917.76 .118.86 .118.65.917.74.317.74.518.54 .518.34.317.7

Sweden

5.011.05.011.25.513.35 .011.03.611.63.511.54.013.53 .511.5

Spain

5 .821.75 .620.85 .520.65.520.64 .221.33 .920.33 .820.13.820.1

ugalPort

5.923.85.422.05.321.85 .321.84.223.43.621.63.621.43 .621.4

e stic earningsdsNetherlan

5 .821.64 .918.45 .520.54.918.44 .221.33 .218.13 .820.13.218.1

domLuxembourg

5.723.44.619.25.221.34 .619.24.123.12.919.13.521.02 .919.1

Italy

3 .417.95.223.95 .123.73.417.91 .818.33.523.73 .423.51.818.3

Ireland

4.5-5.76 .33 .46.23.24 .5- 5.73.2-4.34 .94 .04.93.73 .2- 4.3204

Greece

5.525.45.023.74.817.64.817.63.925.43.323.23.117.63.117.6

Germany

. / .. / .. / .. / ../../.. / .. / .. / .. / .. / .. / ../../.. / .. / .

ev.Stand. D

0 .87 .00 .65 .00.44 .60.65 .90 .43 .90 .93 .50.43 .10.93 .8

Mean

6.028.95.326.76 .028.95 .125.84.642.63.440.05 .544.53 .439.8

Kingdom

United

6 .327.96 .227.56.528.66.227.55 .042.04 .741.26.044.34.741.2

Sweden

5.422.24.920.36 .024.14 .920.34.639.12.935.15 .541.42 .935.1

Spain

6 .331.35 .729.26.030.35.729.25 .044.44 .042.35.545.44.042.3

ugalPort

6.433.15.530.25 .831.45 .530.25.245.93.843.15 .346.13 .843.1

dsNetherlan

6 .331.24 .826.46.030.34.826.44 .443.12 .539.15.545.42.539.1

fo reign earnings

Luxembourg

6.332.74.427.15 .730.94 .427.14.143.62.239.75 .245.92 .239.7

Italy

3 .928.05 .331.95.733.03.928.03 .644.23 .744.35.247.43.644.2

Ireland

4.98.06.314.46 .615.64 .98 .04.530.75.032.46 .135.64 .530.7205

Greece

5.933.05.230.95.327.85.227.84.744.13.741.94.943.93.741.9

Germany

. / .. / ../../.. / .. / .. / .. / .. / .. / ../../.. / .. / .. / .. / .

If we now consider that distributions to the ultimate shareholder are also financed from foreign profits and ignore personal taxes for the moment, we can see from the results shown in Table 33 that a BCC does improve its relative position in comparison with both the most tax efficient direct way of financing and the use of a DFC. This can be explained by the German PFIC rules that add 60% of the profits of aBCC to the German corporation tax base. As these profits are subject to corporation tax they qualify for the reduction of the corporation tax rate for distributed profits. On the other hand, both profits resulting from the most tax efficient direct way of financing and distributed dividends from a DFC are exempt from corporation tax in Germany and therefore cannot take advantage from the split-rate corporation tax system. As a consequence, only in the case of Greece, Ireland and Italy will a BCC offer fewer advantages. (Compared with the other scenario that assumes that distributions are taken from domestic earnings, a BCC now is also more attractive for subsidiaries in France, Portugal, Spain, Sweden and the United Kingdom).

Moreover, if we take personal taxes into account, a BCC would even become more attractive as only certain dividends taken from profits of a BCC (i.e. those 60% of the profits covered by the PFIC rules) qualify for the domestic German corporation tax credit.

Finally, Table 34 compares the cost of capital and the EATR on domestic investment and on outbound investment depending on the assumptions of the financing of profit distributions. The results are overall averages taken from the section on domestic investment and the previous Tables in this section.

Table 34Tax Optimisation in the case of Germany: comparison of domestic and outbound investment

• - 

  cost of capital and EATR

• - 

  most tax efficient way, BCC and DFC

• - 

  corporate and personal taxes

Corporate taxesPersonal taxes

Distributions by

Cost of Capital/parent out ofDistributions byDistributions by

parent out ofDistributions by

EATRdomesticparent out of

domesticparent out of

earningsforeign earnings

earningsforeign earnings

increases the advantage of outbound investment. In addition, if personal taxes become relevant, outbound investment can now be in a better position with respect to the cost of capital if a BCC is used.

Tax optimisation can therefore overcompensates for the disadvantages of foreign source income that does not carry a tax credit within the imputation system. However, with respect to the EATR, domestic investment remains in a better tax situation. The reason is that the economic rent distributed from the subsidiary to the parent does not allow for a tax credit at all.

The UK parent's approach for optimising international financial arrangements (1999)

The legal framework of the analysis

In contrast to Germany the effective statutory tax rate on profits in the UK (currently 30%) is below theEU-average. Another important difference that has an impact on any tax planning strategy is the method for eliminating international double taxation of foreign dividends. Instead of the exemption method the limited foreign tax credit method applies. Having regard to the comparably low corporate tax rate in theUK it is therefore less attractive to establish a foreign intermediary as a base company. Sometimes, however, an intermediary is used as a group financing company following in principle the same structure as was suggested for a German multinational (i.e. asymmetric financing of new equity / debt - see figure 3). Having regard to the application of the tax credit method the principle aim of a UK multinational ought to be to optimise its position towards foreign tax credits. This can be achieved by an intermediary holding company that serves as a so-called "Mixer Company". Recalling the principle forms of financial arrangements set out in box 10 this structure belongs in the category of symmetric financing (new equity / new equity). Figure 4 summarises the sources of finance and the resulting cash flows.

Figure 4:UK perspective of international financial arrangements

EU Subsidiary (Member States)

DebtEquity

(DFC)(Mixer)

InterestDividends

(DFC)(Mixer)

• A) 

  The case of a Dutch Finance Companies (DFC)

The interposition of a Dutch Finance Company (DFC) between a UK parent and another EU subsidiary follows the same structure as set out in section 8.2 on Germany. Moreover, the Dutch taxation of UK owned Dutch finance companies is exactly the same. Contributions to a risk reserve are also the same and it is assumed that the risk reserve can be released tax free so that the effective corporation tax rate of the DFC is 30%.

Under the UK controlled foreign companies legislation, the profits of the DFC would not be imputed back to the UK if it was a "holding company" within the meaning of the relevant provisions. This could normally be achieved relatively easily. If and when the finance company declares a dividend to the UK parent, that dividend will be subject to UK corporation tax at 30% less a credit for the underlying tax borne by the finance company. Assuming a fixed Dutch corporation tax rate of 30% there will be no further UK tax liability.

• B) 

  The case of a Dutch Mixer Company

Against the background of tax optimisation for foreign tax credits under the UK method for eliminating double taxation of foreign dividends, the use of an overseas holding company was a relatively common tax planning strategy for UK multinationals. Although this structure has been legislated against in theUK Finance Act 2000, it was the standard tax planning for UK based multinationals up to 1999 and is therefore considered here.

When a UK company receives a dividend from an overseas subsidiary, that dividend is subject to UK corporation tax at 30%. The UK eliminates double taxation of foreign dividends by granting a tax credit for the underlying foreign tax. However, this credit is calculated on a per-source basis and limited to theUK corporation tax on the grossed-up amount (i.e. dividend plus underlying tax). Therefore, optimum efficiency arises where the overseas subsidiary's underlying rate of tax is equal to the UK tax rate of 30%. To the extent that the underlying rate is greater than 30%, some of the credit is effectively wasted since the first 30% of the underlying tax is sufficient to prevent any further UK tax liability on the dividend arising. Correspondingly, to the extent that the underlying tax rate is less than 30%, UK tax is payable. UK law did not permit the averaging out of underlying tax rates at the UK company level.

The use of mixers to smooth the rate of tax on overseas subsidiaries could be extended by the introduction of subsidiaries with artificially low rates of tax, effectively money-box companies, where an appropriate amount of zero-taxed income could be mixed with `real' foreign income to achieve a 30% rate.

Although in an optimum situation there is an effective tax rate of 30% on all overseas subsidiaries it is not clear to which subsidiaries the benefits should be attributed. They could be attributed to those subsidiaries located in high tax jurisdictions (i.e. with statutory tax rates above the UK rate of 30%) as their tax burden is effectively reduced to 30%. However, the benefits could also be attributed to subsidiaries in low tax jurisdictions (i.e. with statutory tax rates below the UK rate of 30%) as a furtherUK tax on their dividends is avoided due to mixing. Economically, the latter dividends are thus exempt from UK tax. In the following, however, the first interpretation is used.

8.3.2 Relevant economic measures: cost of capital and EATR of a UK parent and its EU

subsidiaries

Tables 35 and 36 compare the cost of capital and the EATR of a UK parent and its EU subsidiaries both for the most tax efficient way of financing the subsidiaries by the parent and for the case of a financial intermediary. The financial intermediary can be either a DFC or a Dutch mixer company that are both financed by the UK parent with equity capital. However, different forms of finance are used by the

intermediaries: whereas the DFC transfers the money as debt capital to other group members the mixer company injects new equity capital into the subsidiaries. In case of a mixer company the subsidiaries can also retain their earnings. In addition to the country results, averages and standard deviations for each way of financing are calculated.

Table 35 shows that, as a consequence of the comparably low UK statutory tax rate and the limited tax credit on foreign source dividends the most tax efficient way of directly financing an EU subsidiary from the perspective of an UK parent is, on average, debt financing. Only in source countries with an even lower statutory tax rate does profit retention in the subsidiary turn out to be more tax efficient. This is the case for Finland, Ireland, Italy and Sweden. The advantage of profit retention over new equity financing and debt financing can be attributed to the fact that the foreign subsidiaries do not pay the higher UK tax rate on profits.

the case of a Belgian subsidiary (corporation tax rate of 40.17%) compared with an Danish subsidiary (corporation tax rate of 32%)).

Looking at the results for a German and Italian subsidiary, specific features of these two tax systems come up. In the case of Germany the cost of capital for the most efficient way of direct financing (i.e. debt financing) are increased by 50% of the local trade tax (because 50% of the interest payments have to be added back to the base of the trade tax). With the interposition of a mixer company, however, new equity financing becomes advantageous as all German taxes including trade tax can be credited againstUK corporation tax. In the case of Italy the result can be attributed to the dual income tax (DIT) which lowers the cost of capital of an Italian subsidiary. In case of a mixer company it is assumed that Italian corporation tax can be credited against UK corporation tax at the statutory rate leaving the DIT advantage unaffected.

In contrast to the tax optimising strategy of a German parent company (where debt it used as the only source of finance of a foreign subsidiary), the interposition of a mixer company also reduces the effective average tax rate by a considerable amount. This can be attributed to the fact that not only the foreign corporation tax on the marginal return but also the foreign corporation tax on the economic rent can be credited against UK corporation tax. The ranking of the host countries for the subsidiaries from the perspective of a UK parent company, do not differ between the cost of capital and the EATR. With respect to the cost of capital and the EATR among Member States, however, differences remain. These are explained by the differing domestic tax bases (rules for computing taxable income) which are still relevant. This also explains the greater standard deviation.

The introduction of personal taxes has no effect on the country ranking and the relative advantages of different ways of financing. This is because the shareholder relief system means all distributions are treated in the same way in the hands of UK shareholders.

ev.Stand. D

0.63 .30.93 .20.42 .40.93 .20.62 .70.92 .60.42 .10.92 .6

Mean

5 .829.45 .524.16 .130.55 .524.14 .834.44 .629.95 .135.24 .629.9

Kingdom

United./../../../../.. / ../.. / ../../../../../.. / ../.. / .

Sweden

5.724.85 .724.86 .126.55 .724.84.930.64 .930.65 .231.84 .930.6

Spain

6.129.96.026.06.129.96.026.05.134.65.131.35.134.65.131.3

ugalPort

5 .931.05 .725.05 .931.05 .725.04 .935.64 .830.64 .935.64 .830.6

dsNetherlan

6.129.96.025.96.129.96.025.95.134.65.131.35.134.65.131.3

Luxembourg

5 .830.65 .624.55 .830.65 .624.54 .835.34 .730.24 .835.34 .730.2

Italy

4.328.32.814.85.832.62.814.83.333.62.022.24.837.22.022.2

Ireland

C4.922.14 .922.17 .129.84 .922.14.328.84 .328.86 .134.44 .328.8211

Greece

5.530.05.223.45.530.05.223.44.534.84.429.24.534.84.429.2

pany and DFGermany

5 .534.14 .420.35 .534.14 .420.34 .538.53 .627.04 .538.53 .627.0

Table 36 compares the cost of capital and the EATR on domestic investment and on outbound investment

Table 36Tax Optimisation in the case of UK: comparison of domestic and outbound investment

• - 

  cost of capital and EATR

• - 

  most tax efficient way, Dutch mixer company and DFC

• - 

  corporate and personal taxes

Cost of Capital/

EATRCorporate taxes Personal taxes

Domestic investment6.65.7

28.233.2

Outbound investment

• - 

  Financing in most tax efficient way5.84.8

29.434.4

• - 

  Mixer Company (Netherlands)5.54.6

24.129.9

• - 

  Financing Comp. (Netherlands)6.15.1

30.535.2

• - 

  Minimum(1)5.54.6

24.129.9

Final remarks

The strategy of tax optimisation depends on the tax system in the home country (in particular on the method for eliminating international double taxation) and the relative ranking vis-à-vis the statutory tax rate. For German and UK parents is has been shown that tax optimisation in the field of financing does reduce both the cost of capital and the effective average tax rate on outbound investment. However, there are still differences between the host countries, which can be attributed to the domestic tax bases in these countries and their statutory tax rates. The tax base always has an impact. However, this impact is only minor. The tax rate becomes important for profitable investments, as pure profits cannot be shifted from one country to another by simple financial arrangements. Therefore, tax optimisation cannot remove all tax obstacles for cross-border investment caused by different tax rates and different tax bases. Moreover, under specific circumstances there is a significant impact from withholding taxes on interest payments.

Compared to domestic investments the cost of capital and the EATR on outbound investments can be significantly reduced by the use of financial intermediaries. The advantage is more substantial in the case of a German parent which is explained by the higher tax burden on domestic investments. Given the fact that mixer companies will not be allowed in the UK in the future, only minor advantages from the use of financial intermediaries will remain. It seems likely that a UK multinational bears the same cost of capital and effective average tax rate on investment whether these investment are financed directly (by a loan) or a financial intermediary is interposed.

The effects of abolishing tax-reducing financing structures in Europe and elsewhere will depend on the location of the parent company. From the above results it seems reasonable to conclude that the tax burden for multinationals will be increased in those countries imposing high taxes on domestic investment. In contrast, there will be only minor effects for multinationals resident in low tax countries.

Therefore, removing these financial intermediaries will not contribute, per se, to solve the problem of tax-induced resources misallocation. Taking into account that the main tax driver for effective tax rates differentials is the overall tax rate, companies located in "high tax" countries will have the possibility to compensate for the removal of financial intermediaries by exploiting the possibility of tax arbitrage arising from those differentials.

EFFECTIVE TAX RATES FOR SMALL AND MEDIUM-SIZED ENTERPRISES IN

GERMANY, ITALY AND THE UK IN 1999

This section presents estimates of the costs of capital and effective average tax rates which apply to small and medium sized enterprises in Germany, Italy and the UK. First the tax regimes which apply in each country is outlined. Then the position of small and medium sized enterprises is compared with that

of large corporations.

Tax Regimes in Germany, Italy and the UK

• a) 

  Germany

Entrepreneurs in Germany may set up the legal form of their firm as a corporation or partnership (or as a sole trader). For the purposes of the analysis it is assumed that there are restrictions on the sale of the shares and participations. Hence the entrepreneur is assumed not to sell the firm, with the consequence that the effective capital gains tax rate is zero.

Small and medium-sized corporations are taxed in the same way as large corporations. However, depending on the book value of their assets, small corporations are permitted a 20% accelerated capital allowance for machinery. In addition, they can establish a tax-free reserve of 50% of the initial cost two years in advance of the purchase of the asset. This reserve is reversed when the first depreciation allowance is granted.

Partnerships are not subject to an income tax directly; instead the income is imputed to the partners. However, for the purposes of the trade tax and the real estate tax, the partnership itself effectively is subject to tax. Both taxes are deductible from the income tax base of the partners. The income tax base of the partners is determined in the same way as the corporate profit tax for a corporation. The accelerated capital allowance also applies. The tax base of the trade tax and the real estate tax are also determined in the same way as they are for corporations. However, the trade tax has a tax exempt amount, which means that for small partnerships the marginal rate can be zero. To allow for this, results below are presented for both a zero-rate and top-rate trade tax.

overall CT rate on retained earningsoverall CT rate on distributed profitsoverall income tax ratesoverall income tax rates on interest

corporation52.3543.70 ; 35.4 ; 54.30 ; 35.4 ; 54.3

partnerships:

no trade tax--0 ; 37 ; 47.50 ; 37 ; 55.9

with trade tax--17.6 ; 48.1 ; 56.70 ; 37 ; 55.9

The income tax rates shown correspond to each of three cases (a), (b) and (c) described above.

• b) 

  Italy

As in Germany, small and medium sized enterprises can take the form of a corporation or a partnership/sole trader. The tax regime for corporations is the same as for larger corporations.

Also as in Germany, the profits of a partnership are imputed to the partner and taxed according to the personal progressive income tax with rates ranging from 19% to 46%. Partnerships pay the regional tax (IRAP) and the tax on immovable properties (ICI). Neither is deductible from the personal income tax.

To be eligible for the Italian DIT (dual income tax) relief, partnerships must comply with two rules. First, partnerships must use ordinary (not simplified) accounting. Second, the "ordinary return" must be within the first bracket of tax - and hence taxed at 19%. It is assumed that, in the case of debt finance, partnerships borrow from banks, implying that the tax rate charged is 12.5%, as for corporations.

As with the German case, we analyse shareholders with three different personal income tax rates (zero; marginal rate on taxable income of 30,000 Euro; top rate). The results are as follows:

overall CT rateoverall income tax rates on profitsimputation rateoverall income tax rates on interest

• c) 

  The United Kingdom

In the UK, most small and medium sized enterprises take the legal form of a corporation. Corporation tax differs in two respects from that levied on larger companies. First, the tax rate applied is lower: here it is assumed to be 20%. Second, plant and machinery receives capital allowances at the rate of 40%, rather than 25%.

The personal tax rates for the three types of shareholders (zero; marginal rate on taxable income of 30,000 Euro; top rate). Considered here are:

overall CT rate on retained earningspersonal income tax rates on distributionsimputation rateoverall income tax rates on interest

corporation200 ; 10 ; 32.50 ; 10 ; 100 ; 20 ; 40

Cost of Capital and Effective Average Tax Rate

This section presents measures of the cost of capital and the effective average tax rate for each of the three countries considered. There are three tables, corresponding to three possible income tax positions of the owner of the firm: a zero rate, a top rate, or a medium rate. In the first two cases, the results for small and medium sized enterprises can be compared with the results for large companies. The types of assets and sources of finance as well as their weights are the same as for the base case. This results in 15 types of investment.

Given the different legal forms available to small enterprises and the differences in the tax treatment described above, tax measures for a number of cases are presented. Thus, in Germany, a company is considered, and the company may or may not benefit from the accelerated depreciation allowances and the tax-free reserve. Both positions for a partnership are also considered. In addition, partnerships which are liable, and those which are not liable, to trade tax are considered. This gives six separate categories for German enterprises.

Table 37Cost of Capital and Effective Average Tax Rate

- zero-rate personal taxpayer

- average over 15 types of investment

Large CorporationSmall and Medium-sized Enterprise

Cost of Capital (upper line)

EATR (lower line)

Base Casecorporationpartnership

no trade taxwith trade tax

incentivenoincentivenoincentiveno

incentiveincentiveincentive

Germany6.96.46.95.15.15.65.7

22.320.422.30.40.415.716.1

Italy4.04.0-4.64.6--

27.127.1

UK6.66.1-----

28.219.3

Note. Only Germany levies a trade tax. The "incentives" are as follows.

Germany: 20% accelerated capital allowance for machinery; tax-free reserve of 50% of investment two years in advance of the purchase of the asset, which is reversed when the first capital depreciation allowance is granted.

Italy: the "incentive" refers to when the DIT relief is available.

UK: corporation tax rate of 20% and 40% capital allowances for plant and machinery.

Top rate personal shareholder

Table 38 analyses the situation of entrepreneurs who pay income tax at the highest rate. In general, the costs of capital are lower in Table 38 than in Table 37 while the EATRs are higher. This reflects the discussion in Section 4. The shareholder is assumed to compare the return which can be earned on the hypothetical investment analysed here, with some other financial investment, such as a bank account which pays interest. Introducing tax on the interest receipt from the bank implies that the post-tax rate of return from the alternative asset is now much lower. Other things being equal, this reduces the required return - ie. the cost of capital - from the hypothetical investment. On the other hand, an extra layer of tax must raise the EATR.

The relative position of large companies, small companies and small partnerships in Germany and in theUK is similar to that shown in Table 37. The small company is assumed to have non-qualified shareholders, and so the cost of capital and the EATR in the absence of the incentives are the same as for the large company. The incentives marginally reduce the cost of capital and the EATR. The figures for the partnership are lower than those for the companies. Similarly, in the UK, the small company again benefits from the lower corporation tax rate and the higher allowance.

Table 38Cost of Capital and Effective Average Tax Rate

- top-rate personal taxpayer

- average over 15 types of investment

Cost of Capital (upper line)Large CorporationSmall and Medium-sized Enterprise

EATR (lower line)

Base Casecorporationpartnership

qualifiednon- qualifiedno trade taxwith trade tax

incentivenoincentivenoincentiveno

partnerships is that the cost of capital is lower for the partnerships. The same is true of the EATR. Part of the difference between large and small companies is that the owners of the latter are assumed not be liable for capital gains tax.

"Medium" rate shareholder

Table 39 presents the results for "medium" rate (the average of the zero and top rate) entrepreneurs. The position is similar to that discussed above.

Table 39Cost of Capital and Effective Average Tax Rate

• - 

  medium-rate personal taxpayer

• - 

  average over 15 types of investment

Cost of Capital (upper line)Small and Medium-sized Enterprise

EATR (lower line)

corporationpartnership

no trade taxwith trade tax

incentivenoincentivenoincentiveno

incentiveincentiveincentive

Germany4.34.53.73.94.24.5

33.033.822.523.133.033.6

Italy4.0-4.25.4--

24.725.228.9

Concluding remarks

To conclude, the results of this section show that the specific tax rules applied to SMEs in the countries analysed have the effect of lowering the effective tax burden. In Germany (1999) and in Italy the parternships bear a lower tax burden in comparison to companies whatever the position of the shareholders.

But, when comparing the results of this section with those shown in section 6.3.3, which examined the tax minimisation approach, it is worth noting that small and medium sized enterprises in Germany, Italy and the UK bear a higher tax burden than multinationals investing abroad.

MEMBER STATES' EFFECTIVE TAX RATES IN 2001

This section presents estimates of the cost of capital and of the marginal and average effective tax rates for the EU Member States in 2001. The purpose of this section is to give a summary update of the 1999 picture in order to highlight the principal changes in individual positions and of the overall picture during the last two years. The case presented is that of a domestic investment taking into account corporate taxes but not personal taxes. In fact, the detailed analysis presented for 1999 has shown that the general conclusions arising from the domestic and international analysis are broadly similar in terms of the range of variation, the ranking of Member States and the principal tax driver.

The hypotheses underlying the definition of the investment and of the economic framework in which the investment takes place are the same as in the 1999 analysis. The same tax parameters are also applied, updated to take into account the 2001 tax regimes. Annex B details the tax parameters applied in the present computation.

Tables 40 and 41 present 2001 country data and correspond to Tables 7 and 8 shown in section 4.3.

Table 40 shows the cost of capital and the EMTR for the level of corporation and compares these data with the overall country data presented in Table 7.

Table 40Cost of Capital and EMTR by Country -2001

-by asset, source of finance and overall

-only taxes on corporations

Overall MeanOverall MeanCost of Capital

199920012001

Country

lldy

p itapitataine

Cost ofCaEMTRCost ofCaEMTR

IntangiblesIndustrialBuildingsMachineryFinancialAssets

InventoriesReEarningsNewE q uitDebt

Austria6.320.95.712.65.35.85.26.65.66.66.64.1

Belgium6.422.46.422.45.27.05.38.06.78.08.03.5

Denmark6.421.96.421.64.38.15.66.96.97.37.34.6

Finland6.219.96.421.36.16.45.66.86.87.37.34.6

France7.533.27.331.85.28.48.57.67.08.78.74.8

Germany7.331.06.826.15.47.16.18.26.98.08.04.4

Greece6.118.26.016.96.75.16.05.27.17.47.43.5

Ireland5.711.75.711.75.36.85.25.55.55.95.95.2

Italy4.8- 4.14.3-15.92.44.03.27.54.44.74.73.6

Luxembourg6.320.76.320.75.26.85.37.76.57.77.73.7

Netherlands6.522.66.522.75.17.05.97.46.97.87.84.1

Portugal6.522.56.321.06.56.15.17.56.47.67.64.0

Spain6.522.86.522.86.56.75.47.46.47.77.74.1

Sweden5.814.35.814.35.06.05.06.66.66.76.74.3

UK6.624.76.724.85.58.35.66.96.97.77.74.8

Table 40 shows that 7 Member States (Belgium, Denmark, Ireland, Luxembourg, Netherlands, Spain and Sweden) have the same cost of capital in 2001 as in 1999. Five Member States have minor changes. Finland and the UK show a slight increase in the cost of capital and in the EMTR due to an increase in the corporation tax rate from 28% to 29% and a minor increase in real estate tax (Finland) and a minor increase in real estate tax (UK). France, Greece and Portugal have a slight decrease in the cost of capital and in the EMTR. In the case of France this is due to a reduction in the corporation tax rate from 40% to 36.43% which is compensated to a small extent by a reduction in the coefficient of declining-balance depreciation by 0.25%. The reduction for Greece and Portugal is due to a reduction in the corporation tax rate from 40% to 37.5% and from 34% to 32% respectively. It is worth noting that Denmark has the same cost of capital in 2001 as in 1999 and a very minor reduction in the EMTR, which is almost unchanged. This is due to the combination of a reduction of the corporation tax rate from 32% to 30% almost entirely compensated for by lower allowances for machinery.

decreased. The aim of the introduction of a dual system was to ensure a more homogenous treatment between debt and equity financing as is the case for the Italian dual income system. When comparing theEMTR for the different forms of finance in 1999 (see Table 7) and in 2001 it can be observed that the reduction in the Austrian EMTR is the result of the reduction of the EMTR on retained earnings and new equity whereas the EMTR on debt financing is practically unchanged. Therefore, the reform does seem to result in a more homogeneous treatment of the different sources of finance.

The reduction in the Italian cost of capital and EMTR in 2001 is the result of a reduction of the corporation tax rate from 37% to 36% and, more substantially, of an increase in the equity base concerning the calculation of the allowance for the dual income tax from 100% to 120% in 2000 and 140% in 2001. It is worth noting that, following the 2001 spring elections, the new Italian government has "frozen" the dual system as of the 30 June 2001 and has introduced other forms of tax incentives. It is therefore rather difficult, at this stage, to define precisely the tax code applied to Italian investments in 2001.

Overall, the picture for 2001 shows that a number of Member States have a lower cost of capital andEMTR as a result of changes in the tax codes aimed at reducing the corporate tax rate or introducing or modifying the dual income system. These changes have not fundamentally affected the ranking of Member States nor reduced the dispersions of the EMTRs inside the EU. The most evident change is the remarkable reduction in the Austrian effective tax burden, which now places Austria at the lower end of the ranking together with Italy, Ireland and Sweden.

As far as the situation of the different taxation of specific forms of investment by assets and sources of finance is concerned, the picture is not very different from that of 1999. Only Austria and Italy, due to the implementation or modification of the dual system, seem to have in 2001 a considerably more balanced taxation of the different sources of finance. However, due to the reduction of the statutory tax rates in a number of Member States, the benefit of debt finance on equity finance decrease as the value of the tax deductions decrease.

Table 41 presents a summary of the effective average tax rate for each Member State in 2001 for investment where the pre-tax real rate of return is 20% and compares these data with the overall country data presented in Table 8.

Table 41Effective Average Tax Rate by Country - 2001

-by asset, source of finance and overall

-only taxes on corporations

d

r a llr a lln ery

Countryine

Ove1999

Ove2001ta

IntangiblesIndustrialBuildingsMachiFinancialAssets

InventoriesReEarningsNewEquityDebt

Austria29.827.926.428.226.230.927.630.730.722.6

Belgium34.534.530.736.131.039.235.339.139.125.8

Denmark28.827.319.933.324.729.329.330.730.721.0

Finland25.526.625.726.623.928.328.330.030.020.2

France37.534.727.838.238.435.633.839.039.026.8

Germany39.134.930.836.033.039.235.438.738.727.7

Greece29.628.033.328.531.311.934.832.432.419.7

Ireland10.510.58.915.88.29.89.811.711.78.2

Italy29.827.622.527.124.935.128.428.728.725.5

Luxembourg32.232.228.633.729.236.632.936.636.624.0

Netherlands31.031.026.732.629.234.232.535.235.223.3

Portugal32.630.731.330.126.934.430.934.834.823.0

Spain31.031.031.131.827.434.230.735.235.223.3

Sweden22.922.919.623.419.725.725.726.026.017.1

UK28.228.324.234.024.729.329.331.831.821.7

Almost all the changes to corporation taxes in the EU between 1999 and 2001 have resulted in a reduction of tax liabilities. In two of these cases, Austria and Italy, reforms have been directed towards reducing the tax burden on marginal investments. Consequently, the reforms in these two countries have a greater effect on the EMTR than the EATR.

However, six other countries have reduced their statutory tax rates, albeit with relatively small reductions (the main exception being Germany which reduced its rate from 40% to 25%). In some of these countries, governments have offset the reductions by also reducing allowances (again to the greatest extent in Germany). In general, this has resulted in relatively small reductions in the EMTR, but rather larger reduction in the EATR.

COMPARISON OF THE RESULTS WITH THOSE OF THE RUDING AND BAKER

&MCKENZIE REPORTS

During the last decade a number of studies have presented international comparisons of effective corporate tax burdens. For the reasons outlined in Section 3, the most commonly used approaches for international comparisons have been based on general forward-looking frameworks partly similar to the approach taken in this study. At the European level, the most well-known studies are the Report of the Ruding Committee of 1992 (to which the mandate given to the Commission from the Council for the current study explicitly refers) and the Baker & McKenzie study commissioned by the Dutch Ministry of Finance in 1999.

The purpose of this section is to compare the principal results arising from these two studies with those presented in the above analysis.

As already explained in Section 3, the application of the forward-looking approach, common to all these studies, gives synthetic measures of the effective tax burdens based on hypothetical situations. In practice, it consists in defining a hypothetical investment identical in all countries and then applying to this identical hypothetical investment the different national tax codes. This "isolates" the taxation elements among all the other factors influencing the effective tax burdens and can thus help identify the most important tax drivers influencing the effective tax burdens and their differences.

The results obtained by the application of this approach depend, therefore, on the hypotheses and assumptions underlying the definition of the hypothetical investment considered and the economic framework.

Consequently, when comparing different studies using this methodology, it should always be borne in mind that the aim of these approaches is not to give universally valid values of the effective tax burden for each country considered, but to allow comparisons between countries on the basis of the same investment. In fact, the individual measures of effective tax burdens rely on the specific assumptions used when applying the model. Therefore, when comparing the results of these different studies, the differences in the underlying hypothesis and assumptions must always be borne in mind.

The present report has gone beyond this approach to also consider effective average tax rates both on hypothetical investments and hypothetical model firm behaviour.

As far as the hypotheses and assumptions are concerned, the Ruding report presented results related to a simple investment in the manufacturing sector taking into account the effects of corporate taxes and ignoring all personal taxes, but considering the effects of imputation credits. The Ruding report considered three forms of assets (machinery, buildings and inventories) whereas the present study takes into account two more forms of assets: intangible and financial assets. For the simulations, the personal taxation of the suppliers of fund was included. The present report has considered the effects of overall corporate taxation (statutory tax rates, surcharges and local taxes) on a simple investment in the manufacturing sector and has also included, separately, the effects of personal income taxation of dividends, interest and capital gains.

Concerning the other tax parameters used in the calculation, the Ruding report and the present report rely on the same framework.

• B) 

  Baker and McKenzie study

As with the current study, the purpose of the Baker & McKenzie study was to compute the differences of the EU countries corporate effective tax burden in order to underline the principal tax drivers for these differences.

The Baker & McKenzie study presented an analysis based on the King and Fullerton methodology related to the taxation of a hypothetical investment. Only the case of a marginal investment was covered and, thus the study presented evidence of effective marginal tax rates for the 15 EU countries using the tax codes in force in 1998. Effective marginal tax rates were computed for all the 15 countries in the domestic case. For the transnational investment only the cases of Germany, the United Kingdom and the Netherlands were analysed. The report did not assess the contributions of specific features of taxation to the lack of neutrality by means of simulations.

As far as the hypothesis and assumptions are concerned, the hypotheses related to the hypothetical investment considered are very similar to those used in the present study. The main difference is that the Baker & McKenzie study considered a pre-tax real rate of return of 10% against the 5% post-tax rate of return considered in this study and in the Ruding report and an inflation rate of 1.1% against the 2% used here.

Box 13Comparisons of the hypothesis and assumptions between the Commission study (2001) and the Ruding (1992) and the Baker & McKenzie (1999) reports

Commission study(2001)Ruding report (1992)Baker & McKenzie study (1999)

MethodologyDevereux & Griffith: Marginal and average indicatorsKing and Fullerton: Marginal indicatorsKing and Fullerton: Marginal indicators

Type of analysisDomestic analysis Cross-border analysis (all EU countries)Domestic analysis Cross-border analysis (all 1992 Community countries plus selected partners)Domestic analysis Cross border analysis (only UK, D and NL)

Year (tax code)1999 and 200119901998 and 2001

Simulations of hypothetical reformsyesyesNo

SectorManufacturing Sensitivity analysis for servicesManufacturingManufacturing Sensitivity analysis for services

Types of assetsIntangibles, machinery, buildings, financial assets, inventoriesMachinery, buildings, inventoriesIntangibles, machinery, buildings, financial assets, inventories

Weight of assets1/5 eachMachinery: 50%,

Buildings: 28%

Inventories: 22%Machinery: 70.49%

Buildings: 12.99%

Inventories: 29.84%

Intangibles: 1.43% Financial assets: 38.25%

Source of financeNew equity, retained earnings, debtNew equity, retainedNew equity, retained earnings, debt

earnings, debt

Weight of sources of financeNew Equity: 10% Retained earnings 55%

Debt: 35%New Equity: 10% Retained earnings 55%

Debt: 35%New Equity: 10% Retained earnings 55%

Debt: 35%

Inflation rate2% for all countries3.1% for all countries1.1% for all countries

Rate of returnPost-tax rate:5%Post-tax rate:5%Pre-tax rate:10%

Tax parametersOverall corporation tax rates including surcharges and local taxes; Corporate real estate taxes, net wealth taxes and other non-profit taxes on wealth;

Tax credit associated with dividend and equalisation tax; Personal income tax rates, including withholding taxes on dividend, interest and capital gain; Individual net wealth taxes on shareholdings and lending Withholding taxes on dividends and interest Treatment of foreign source dividends and interest received by parent companies Capital allowances for industrial buildings, machinery and intangibles. Tax treatment of financial assets and inventoriesStatutory corporation tax;Overall corporation tax rates including surcharges and local taxes; Corporate real estate taxes, net wealth taxes and other non-profit taxes on wealth;

Tax credit associated with dividend and equalisation tax; Personal income tax rates, including withholding taxes on dividend, interest and capital gain; Individual net wealth taxes on shareholdings and lending Withholding taxes on dividends and interest Treatment of foreign source dividends and interest received by parent companies Capital allowances for industrial buildings, machinery and intangibles. Tax treatment of financial assets and inventories

Comparisons of the results in the domestic case

Tables 42 and 43 present a comparison of effective tax burdens across the three studies considered. The Ruding report presented only cost of capital indicators. Therefore Table 42 compares the cost of capital presented in Section 4 of this study with those computed in the framework of the Ruding report. The Baker & McKenzie study presented only effective marginal tax rates. Table 43 thus compares the EMTR presented in Section 4 of this report with those published in the Baker & McKenzie study.

Debt

2.62.83.2-3.51.42.74.62.63.43.22.93.53.13.7

quityNew e

6.86.97.2-3.52.22.75.02.67.87.07.37.54.34.7

earnings

Retained6.86.97.2-7.08.87.15.48.87.87.07.37.56.47.4

(1992)AL

I T

Inventories C AP

8.38.36.8-

7.36.95.95.56.38.46.26.47.96.37.4

REPORT

T OF

C OSMachinery

4.04.25.3-4.65.24.85.05.54.95.25.25.54.55.2

RUDING

Buildings5.45.46.0-5.45.15.04.96.76.96.06.15.75.15.8

Overall average5.35.45.8-5.45.65.15.16.06.25.75.76.15.05.9

Debt4.03.54.44.54.63.23.45.23.63.74.13.94.14.34.8

quityNew e

7.58.07.57.29.07.67.65.95.57.77.77.97.76.77.7

:earnings228

e ntRetained7.58.07.57.29.09.77.65.95.57.77.77.97.76.77.7

stm

mi ssion study (2001)

Inventories

ic Inve(2001)6.36.77.16.87.47.97.45.55.06.56.96.56.46.66.9

AL

I T

Table 42 shows a rather different picture across the two studies. In general, the cost of capital is lower in the figures arising from the Ruding report and the ranking of the Member States is different in the two computations. The differences in the tax parameters considered may have a great influence on these results, notably on the observed levels of the cost of capital. As was shown in Box 11 the present study considers a broad range of taxes levied on companies, whereas the Ruding report considered just the statutory corporate tax rate. Section 4 of this study underlined the weight that taxes other than the corporate tax may have on the level of the effective tax burden, in certain countries in particular. Therefore, it cannot be concluded, simply by comparing the results of Table 42, that the cost of capital is generally higher now than ten years ago.

It is interesting to observe that differences due to different assets and weight of assets seem to play a lesser role in determining differences in the results. In fact, Section 5 (sensitivity analysis) showed that the results of the present study would not be fundamentally affected by using the same weights as those used in the Ruding report. By contrast, when using the "Ruding report" weights for the sources of finance the cost of capital of the present study would be slightly lower. (See sensitivity analysis number

5 in section 5.1).

The fact that the ranking of the Member States is also different may be due, on the one hand, to the differences on the tax parameters used and, on the other hand, to the number of tax reforms undertaken by Member States this last decade. It is difficult, just on the basis of the results shown in Table 42, to infer which is the relative weight of these two factors in determining differences in the picture presented now and the situation of ten years ago. It is evident that national tax reforms may have had a considerable effect in determining ranking differences.

One useful way to compare the results presented in Table 42 is to analyse whether the global picture in terms of influence of the tax systems on the incentives to invest and financing decisions has evolved since early 1990s.

The analysis of Section 4 suggested that the tax systems tend to create a misallocation of resources by favouring certain forms of assets and certain sources of finance. The same patterns are observable in the figures of 1992. Today, however, the differences in the effective tax burdens held by different forms of investment or financing are smaller then ten years ago, although still considerable. This seems to suggest that the tax reforms undertaken during the last decade may have in part contributed to make more neutral the influence of taxation on the organisation of domestic investments. Nevertheless, the persistence of high differences reopens the debate on the possible distortive effects of taxation on the allocation of resources within the EU and on the possible frustration of equity goals of policymakers by tax arbitrage arrangements.

 

mi ssion study (2001)

Debt

6.1-3.90.7-0.318.44.4-11.91.7-0.5-4.20.3-2.612.8-1.42.5

quityNew e

38.137.934.527.852.454.227.333.127.338.135.235.843.327.032.8

earnings

Retained

38.137.934.527.852.454.227.333.127.338.135.235.843.327.032.8

(1999)

Inventories

28.226.324.521.842.340.425.225.122.826.625.224.534.521.724.3

REPORT

I

E year 1998)R

ssetsFinancial A

28.226.322.318.439.440.45.221.017.726.622.924.534.518.420.3

KENZEMT

C code of

• & 

  MMachinery

(Tax23.78.414.912.037.326.718.015.54.916.118.016.523.77.614.6

ER Buildings

BAKIndustrial

25.430.235.917.347.234.75.929.124.518.326.019.935.617.535.2

Intangibles

28.410.9-17.818.425.920.823.121.18.912.523.124.835.06.214.5

Overall average

27.023.522.818.140.737.013.722.317.723.523.222.532.817.222.3

Debt

-25.0-42.9-13.6-11.1-08.7-56.2-47.13.8-38.9-35.1-21.9-28.2-21.9-39.5-25.0

232

quityNew e

33.337.533.330.544.435.534.215.210.035.135.136.735.125.435.1

earnings

Retained

33.337.533.330.544.448.434.215.210.035.135.136.735.125.435.1

The main difference between the Commission report and the Baker and McKenzie report is that the first considers a fixed post-tax return (the real interest rate) and the second a fixed pre-tax rate of return. It is noteworthy that the results of the quantitative approaches depend on the interest rate assumption adopted as the real interest rate corresponding to the post-tax rate of return. This study uses a post-tax rate of return of 5% as the base reference case for real investment decisions and computes the range of effective tax rates applicable at the fixed post-tax rate of return. The Baker and McKenzie study, contrary to the prevailing practice, use a fixed pre-tax rate of return typically set at the higher rate of 10% and compute the range of effective tax rates applicable at the fixed pre-tax rate of return. This explain to a large extent the different numerical results of the two studies.

The differences of the values arising from the two computations are also partly due to the different hypothesis in terms of the considered investment and tax parameters.

Notwithstanding these differences, there is a remarkable similarity in the general conclusions arising from the two pictures.

First of all, the ranking of countries is consistent in the sense that, even if the individual country positions can vary (for instance France has the highest EMTR in the Baker & McKenzie study and the second highest EMTR in the present study), we find the same group of countries at the top and at the bottom of the scale (apart from Ireland). It should be underlined that when countries have a very similar cost of capital (see previous Table) even a small change in the assumptions and parameters is likely to affect the ranking of these countries.

Notwithstanding the difference in the relative position of Ireland due to the important difference in the tax rate considered, the lowest EMTR countries in both studies are Italy, Sweden, Greece and Finland and the highest EMTR countries are France, Germany and Spain. The results of both studies reveal a considerable range of around 30 points between the highest and lowest values. Moreover, the Baker & McKenzie study stressed that the principal tax driver for these differences is, above all, the overall corporate tax rate and that differences in the tax base can play an important role in specific individual situation. The present study comes to at the same conclusions.

The two studies also show similar results for the differences in the effective tax burden on different assets and sources of finance, and therefore similar conclusions can be drawn from the two studies. It is worth noting that the Baker and McKenzie study considered an inflation rate of 1.1%, against 2% in the present study. A higher rate of inflation exacerbates the differences between debt and equity finance. This is due to the fact that nominal interest rates are assumed to rise in line with inflation; since these are deductible in the case of debt financing, the EMTR falls for debt financing.

The following table shows the Member States' effective marginal tax rates computed by Baker and McKenzie for 2001 where the pre-tax rate of return is fixed at 10% and 6% respectively. These data are compared with the effective marginal tax rates computed in this study where the post-tax rate of return is fixed at 5%.

Table 44EMTR for domestic investment: Baker and McKenzie report (2001) and Commission Study (2001)

• - 

  only corporation taxes

BAKER and McKENZIE report (2001)COMMISSION Study (2001)

Tax code of year 2001Tax code of year 2001

PRE-TAX RATE OF RETURN (p)POST-TAX RATE OF

RETURN (r)

10%6%5% (base case)

COUNTRY

Austria18.2520.4212.6

Belgium18.8917.2222.4

Denmark18.8119.7921.6

Finland18.0918.5821.3

France30.1136.8431.8

Germany25.2023.8026.1

Greece6.764.8916.9

Ireland9.4310.6411.7

case in which the pre-tax rate of return is fixed at 6% is the closest to the case considered in this study where the post-tax of return is fixed at 5%, accordingly these two situations forms the basis of the present comparison.

The differences in individual results of the two studies are due not only to the different fixed pre- tax/fixed post-tax assumptions but also due to differences in a range of other assumptions.

The Baker and McKenzie study for 2001 considers a broader range of non-profit taxes including the payroll taxes and the rate applied to Ireland is 12,5% as against 10% rate in force in 2001 for the manufacturing sector. Moreover, the "dual income" system in force in Italy has been applied in a different way in the two studies. The Italian "dual income" tax systems splits the tax base for profits into two components, taxed at different tax rates. Very broadly, the ordinary return, calculated as the interest rate multiplied by equity invested into the company, is taxed at 19%, while the residual profit is taxed at 37%. However a rule applied in the past and abolished in 2000, stated also that, whatever the result of the application of the dual system, at the end the average rate applied (resulting in the application of these two different rates according to the method of financing and the amount of residual profit) should not be less than 27%. This study has not taken into consideration this "minimum" global corporate rate of 27% and in the marginal case (no extra-profits) the rate of 19% has been applied. Instead, the Baker and McKenzie study applies this "minimum" rate of 27%. As a result the EMTR in this study is lower than the EMTR in the Baker and McKenzie study. And, finally, the weights of the five assets composing the hypothetical investment considered is different in the two studies.

Notwithstanding these differences, there is a remarkable similarity in the general conclusions arising from the two pictures for 2001, as it was the case for 1999. The same conclusions drawn from the comparison with the 1999 results remain valid for the 2001 results.

Comparisons of the results in the transnational case

As, in the transnational case, the Baker & McKenzie report computed effective marginal tax rates only for three countries (D, NL and UK), this section does not make a comparison between the results of the present study and those arising from the Baker & McKenzie study. The comparison is thus limited to the results of the Ruding report.

UK debt,

6.45.96.97.06.76.47.39.06.2e of

e n t i n aerag

Portugal

7.25.78.18.46.06.514.07.67.7t ;

i n v e s t mh t

ed av

e ig

w

lands t h

e paren

Nether-6.15.97.46.46.36.26.88.36.3 fr

om fi n a n

ce in

es

t rais

bourgthi r

d debten

Luxem-6.56.46.96.56.76.57.08.06.2n

e-

d o

t an

yItal

7.26.79.57.47.06.76.6.07.1a rene s are zero; par

e p

) thal tax

on

48

fr o m

e ntse

nt toIreland

5.65.36.19.16.75.45.19.65.5u ity

eq

w h e re; pers

( investmFrance ne w

ery

rce 7.06.57.710.37.25.47.39.57.0third

n

e- 5% ev

So u, o

Spainiaryt rate of

7.55.67.28.86.17.214.28.57.0b s idteres

o r Transnational Investme su

y th

6.27.05.9

8.45.76.6

5.711.16.2

6.79.56.5

6.37.16.1

5.58.47.0

7.07.67.0237

7.07.06.9

Tables 17 to 19 and Table 45 are not directly comparable, for three main reasons. First, as was the case for a domestic investment, differences in the parameters used, and notably the number of taxes levied on companies, can explain some differences in the results between the two studies.

Second, the tax reforms of the last decade and notably the reforms related to the tax regimes can explain the different results in individual Member States in 1999 compared with those of ten years ago.

Third, the fact that these matrices summarise all the domestic and international features of taxation makes it very difficult, when comparing the figures from the two studies, to attribute to one factor or another the origin of differences in the results.

All that said, two observations can be made. Tables 17 to 19 show that subsidiaries located in some host countries, and notably Ireland and Italy, always have a lower cost of capital than subsidiary located in other host countries. On the other side, subsidiaries located in Germany and France always have a higher cost of capital. Table 45 presents a more mixed picture and it is impossible to say that some host countries are generally more or less attractive.

As far as the size of variation is concerned, the picture arising from Tables 17 to 19 shows considerable variations within columns and rows. Therefore, on the one hand, there is considerable incentive for companies to choose tax favoured locations and, on the other, the effective tax burden of a subsidiary depends heavily on the home country of the parent company. When looking at Table 42, it does not seem that such incentives have diminished compared to the situation of ten years ago.

Table 46 compares the figures in the Ruding report and the present study which assess the extent to which the tax treatment of transnational investments gives incentives to undertake transnational, as opposed to domestic investments.

Outbound

0.50.70.4-0.51.11.32.91.10.50.81.60.30.50.4

Inbound

EU StandardDeviation

1.10.50.9-1.10.42.11.50.80.91.52.12.11.50.8

Outbound

6.76.56.1-6.27.37.98.68.06.66.58.06.66.56.4

Ruding Report (1992)Inbound

6.76.66.9-7.66.17.06.77.17.07.07.98.06.46.8

EU Average

Domestic

5.35.45.8-

5.45.65.15.16.06.25.75.76.15.05.9

Outbound

0.60.60.60.60.50.60.60.60.40.60.60.60.60.60.5

i a ryInbound

EU StandardDeviation

0.20.30.30.30.30.30.30.40.30.30.20.30.30.30.3

s i d

As is the case now, both capital export and capital import neutrality were absent ten years ago. The Ruding report showed differences between domestic, outbound and inbound cost of capital that were larger than the differences found in the present study. Moreover, the standard deviations for inbound and outbound costs of capital were also higher. It seems, therefore, that the reforms undertaken during the last decade had a positive influence in reducing, on average, the incentives to undertake domestic investments as opposed to transnational investments.

But, as pointed out above, the averages presented in the current study mask large differences across the possible home/host countries. A lot of information is lost, in particular the variation between countries and sources of finance. There is, in fact, a remarkable range of variation in the effective tax burden of subsidiaries located in different host countries or for subsidiaries operating in a given country.

Moreover, it has to be stressed, once again, that the different picture arising from the two studies is at the same time the result of evolving tax systems and of different economic assumptions and parameters underlying the computations.

One important difference in the results of the two studies is that, while ten years ago all Member States had a cost of capital for domestic investment lower than the outbound cost of capital, now, there are some Member States for which outbound investment is less heavily taxed than domestic investment. These are the Member States with the highest national profit rate. On the other side, Member States with the lowest national profit tax give a marked advantage to domestic investment.

Comparison of the results of the simulations of hypothetical tax reforms

The results of the simulation presented in section 7 of this study are generally in line with the impact of hypothetical reforms simulated in the Ruding report.

In particular, the importance of the harmonisation of statutory corporate tax rates to increase the degree of locational tax neutrality is one of the main conclusions of the Ruding report. In addition, that report similarly demonstrated that "harmonisation of certain aspects of the tax base in isolation does not always increase neutrality".

Lastly, the conclusion, made in the Ruding report, that the adoption of a common classical corporation tax or a common imputation system would not improve the overall degree of locational tax neutrality slightly differs from the results presented in the current study. Indeed, simulation 12 in section 7 suggests that a common classical system could improve the overall allocational neutrality in the EU.

Results from other studies

Measuring the effects of corporate income taxes on the cost of capital is a standard way of assessing the potential impact of corporate taxation on investments.

During the last few years a number of studies have presented estimates of effective corporate tax burdens at national and international level by means of the application of approaches similar to that used in the present report.

In particular, the French Sénat (1999) published a report on tax competition in Europe which includes a comparison of domestic and international effective corporate tax burdens for the EU Member States in 1998. More recently Bond and Chennels (2000) published a comparative study of effective corporate tax burdens for seven selected world economies (five EU Member States plus USA and Japan) with the purpose of studying the trends in effective corporate taxation over the last thirty years.

As already underlined above, the numerical results arising from the application of forward-looking methodologies and therefore also the country ranking can differ due to different hypotheses concerning the economic environment in which the investment takes places and the specific investment considered.

A detailed analysis of the different numerical results of studies published over the last few years would demand a detailed assessment of the different hypotheses and goes beyond the scope and purpose of the present report. It is nevertheless useful to compare the general statements arising from these different studies in order to understand whether the different applications give coherent answers to the question of the potential role of taxation on investments.

hypotheses related to the underlying methodology, as well as to the domestic or international localisation of the investment, the profitability of the investment or of the firm considered, and the size and behaviour of the companies. The computations have been supplemented by "sensitivity analysis" which tests the impact of different hypotheses on the results.

The broad range of data computed is not intended to give "universally valid values" for the effective tax burden in different countries, but rather to give indicators, or illustrate interrelations, in a series of relevant situations. In fact, effective tax rates may vary depending on the characteristics of the specific investment project concerned.

A number of general conclusions regarding both the differences in the effective tax burdens and the identification of the most relevant tax drivers which influence these tax burdens, can nevertheless be formulated on the basis of the results. Therefore, coherent explanations can be given of how Member States' tax regimes create incentives to allocate resources. The most striking feature of the quantitative analysis in this study is that, across the range of different situations, the relevant conclusions and interpretations remain relatively constant.

This study does not aim to estimate quantitatively the impact of differences in effective tax rates on the actual location of investments. The data arising from the application of the principles and assumptions underlying this study give summary measures of the incentives (or disincentives) to undertake different investments but do not provide empirical evidence of the impact of taxation on actual economic decisions. The empirical studies which have attempted to study the relevance of tax considerations in investment decisions show that there is, to differing degrees, a negative correlation between the size of taxation and location decisions. Nevertheless, certain methodological weaknesses and data shortages which affect these studies make it difficult to define "the" quantitative indicator which summarises this relation.

When domestic investments are considered, (see section 4) the analysis for 1999 suggests that there is a variation in the effective tax burden faced by investors resident in the different EU Member States, depending on the type of investment and its financing. However, the Member States tax codes tend to favour the same forms of investment by assets and sources of finance.

effectively subsidises investment. But, when the profitability of the investment rises the effective tax burden rises and for an intra-marginal investment the Italian effective tax burden is in the middle of the EU range. In general, there are not substantial differences in the ranking of the Member States when comparing marginal and infra-marginal indicators.

The analysis suggests also that, in practically every situation analysed, on the one hand, the tax systems tend to favour investment in intangibles and machinery and, on the other hand, debt is, by far, the most tax-efficient source of finance for all Member States.

The introduction of personal taxation substantially increases the effective tax burdens and the observed differences. In this situation the ranking of Member States by effective tax burden is very different from the case in which only corporate taxation is considered. Moreover, it is no longer true that debt is always the form of finance which minimise the effective tax rate, even if for a majority of Member States debt is still the most favoured form of finance.

It is worth noting that the values of the effective tax burden for each Member State can vary according to the definition of the economic variables and parameters underlying the application of the methodology and, as mentioned above, there is no universally valid value in one country. However, the sensitivity analysis in section 5, suggests that the ranking of Member States is largely unaffected by changes in the assumptions used in the base scenario.

Differences between the effective tax burden in the EU Member States may be important for two reasons. First, differences in effective tax rates faced by companies located in different Member States, but competing in the same market, may affect their international competitiveness: two different companies, competing in the same market, may face two different tax rates. Second, when multinational companies face only the tax rate of the country where the activity takes place then differences in the effective tax rates between countries could also affect the location choice of individual activities. This can occur either as a result of the provisions of international tax codes, for example when the repatriation of profits by way of dividend from a subsidiary to a parent results in no further taxation because the dividend is exempt, or as a result of tax planning. A multinational company may therefore face different tax rates, depending on where its activities are located.

and that, consequently, the compensatory effects of a lower tax base in countries with high tax rates on effective tax rates tend to disappear when the profitability rises.

When transnational investments are considered, (see section 6) the data for 1999 arising from the quantitative analysis illustrates a variation in the way each Member State treats investments in or from other countries. Thus, the effective tax burden of a subsidiary of a parent company in one country depends crucially on where that subsidiary is located. The range of variations of the effective tax burdens of subsidiaries located in different host countries can reach even more than 30 points regardless of the method of financing of the subsidiary. This suggests that there is considerable incentive for companies to choose the most tax-favoured locations for their investment, which may not be the most favourable location in the absence of taxes. Similarly, subsidiaries operating in a given country face different effective tax burdens depending on where their parent company is located. Even in this case the range of variation can reach more than 30 points.

The analysis of the effective tax burden of transnational investment also allows an assessment of the allocation effects of international taxation by capturing the extent to which the tax treatment of transnational investments gives incentives to undertake transnational, as opposed to domestic, investment. The data show that, on average in theEU, outbound and inbound investments are more heavily taxed than otherwise identical domestic investments and, therefore, the additional components of the transnational system add somewhat to the effective tax rates on investment.

But, to the extent that companies are free to choose the most tax-favoured form of finance, then the international tax system works such that foreign multinationals operating in a host country are likely to face a lower effective tax burden than domestic companies. This is true even when the treatment of multinationals is compared with the more favourable domestic treatment allowed for small and medium sized companies (see sections 6 and 9).

The spreads observed between the effective rates of taxation in the international analysis are the result of complex interactions between different tax regimes and cannot be explained by just one feature of taxation. However, as was the case for the domestic investment, the analysis tends to show that the relevant component that results in distortions with respect to cross border location and financing decisions is, above all, the overall national tax rate. This is, in general, the most important tax driver when the incentives of taxation to use particular sources of finance and specific locations are considered. The tax base does however have a greater impact in specific situations - for example when a Member State applies a particularly favourable depreciation regime as is the case in Greece, and to a lesser extent in Finland and Sweden.

cost of capital and effective marginal tax rate as a result of changes in the tax codes aimed at reducing the corporate tax rate or introducing of modifying the dual income system. These changes have not fundamentally affected the ranking of Member States nor reduced the dispertion inside the EU. The most evident change is the remarkable reduction in the Austrian effective tax burden, due to the introduction of the dual income system, which now places Austria at the lower end of the ranking together with Italy, Ireland and Sweden.

In the case of a more profitable investment, the EU differential is reduced mainly due to the effect of the reduction of the German and French effective tax burden, whereas the lowest effective average tax rate remains the Irish rate. Apart from this reduction in the range of rates of about 4 percentage points, the global picture arising from 2001 in terms of countries at the highest or lowest range of the ranking and principal tax drivers is fundamentally unchanged in comparison to the 1999 picture.

The German tax reform that entered into force on 1.1.2001 is undoubtedly a significant reform which implies a substantial cut in the corporation tax rate and in income tax rates, partly financed by the broadening of the tax base, including the abolition of the split rate system and the imputation system. However, despite these changes and the considerable reduction of the average effective tax burden at the corporate level, the German tax reform has only minor effects on the relative position of Germany in the EU country ranking. In fact, the overall national corporate rate in Germany remains high by the standards of the EU. Consequently, the effective tax burden remains among the highest in the EU.

When the role of specific features of the tax regimes is examined by simulating the impact of hypothetical policy scenarios on the measure of effective tax rates by means of harmonisation of particular features of taxation in isolation (see section 7), the results of the application of the two methodologies underlying this study show that:

· Introducing a common statutory tax rate in the EU would have a significant impact

by decreasing the dispersion - both between parent companies and between subsidiaries - of effective tax rates across the Member States. To the extent that taxation matters, such reform would be likely to go some way in reducing locational inefficiencies within the EU.

· Introducing a common form of integration of corporate and personal taxes, other

than a pure classical system, would not tend to reduce the dispersions of effective tax rates between Member States.

It should be noted that these conclusions are the result of a static analysis. They therefore do not assess the dynamic effects and possible reactions induced by the harmonisation of particular features of taxation in isolation.

The potential distortions in the allocation of resources found in the analysis of transnational investments indicate that there can be a considerable incentive for companies to alter their behaviour in order to minimise their global tax burden. Therefore this study considered some stylised examples of tax optimisation strategies of companies by means of an intermediary financial company, focusing on the likely effects of an abolition of these tax reducing financing structures (see section 8). The study shows that preventing such tax optimisation strategies will not contribute, per se, to solving the problem of tax-induced resources misallocations. Taking into account that the main tax driver for effective tax rates differentials is the overall tax rate, companies located in "high tax" countries will have the possibility to compensate the removal of financial intermediaries by exploiting the possibility of tax arbitrages arising from those differentials.

The size of tax differentials and dispersions in the EU measured in this study deserves attention. The principal tax driver for these differences is, above all, the overall national corporation tax rate. Although the existence of market failures can justify a certain degree of tax differentials in order to offset these externalities, the size of the differences observed in this study is likely to impact on economic efficiency. This study has not attempted to quantify the size of any efficiency loss that might be associated to existing differences in effective corporation tax rates in the European Union.

But taxation ultimately involves a political choice and may entail a trade-off between pure economic efficiency and other legitimate national policy goals and preferences. The objective of neutrality of taxation systems within the Internal Market has therefore to be seen in the context of national autonomy in the field of taxation..

PART III:

COMPANY TAX OBSTACLES TO CROSS-BORDER

ECONOMIC ACTIVITY IN THE INTERNAL MARKET

2. INTRODUCTION

The Council mandate given to the Commission defines "the remaining tax obstacles to cross-border economic activity in the Internal Market" as "tax provisions that may hamper cross-border economic activity in the Internal Market". This description of tax obstacles is rather broad. Hence, the present study focuses on additional tax or compliance burdens which companies incur as a result of doing business in more than one Member State and which therefore represent a barrier to cross-border trade, establishment and investment.

The underlying cause of those additional tax and compliance burdens is the existence within the Internal Market of 15 separate tax systems. First, the fact that each Member State is a separate tax jurisdiction has a number of consequences. In particular:

· companies are obliged to allocate profits to each jurisdiction on an arm's length

basis by separate accounting, i.e. on a transaction by transaction basis;

· Member States are reluctant to allow relief for losses incurred by associated

companies whose profits fall outside the scope of their taxing rights;

· cross-border reorganisations entailing a loss of taxing rights for a Member State are

liable to give rise to capital gains taxation and other charges;

· double taxation may occur as a result of conflicting taxing rights.

Moreover, each Member State has its own sets of rules, in particular laws and conventions on financial accounting, rules for determining taxable profit, arrangements for collection and administration of tax and its own network of tax treaties. The imminent enlargement of the EU makes it all the more urgent to address the underlying problems.

cope with 15 different tax systems in the EU as well and pleads, for instance in the Trans-Atlantic Business Dialogue, for a simpler EU system.

49

In a broader perspective, the tax obstacles to the Internal Market may also result in economic terms in a loss of potential EU welfare. In setting a tax incentive towards domestic economic activity the obstacles violate the basic neutrality criterion explained in Part I of this study. Moreover, they may result in an overall economic situation of theEU and its citizens that is less efficient, equitable and effective than it could be and thus reduces the general well-being. However, in order to determine the size of these welfare effects, it would in particular be necessary to quantify the compliance cost inherent in the existence of 15 different tax systems within the Internal Market or, in other words, the fiscal surcharge for international activities.

Currently, such calculations are not available for the EU and there is certainly scope for future economic research in this area. The present study could not embark on such a demanding exercise. The available studies

50 on compliance cost mostly concern the

USA, Canada and Australia, i.e. countries which are economically broadly comparable to the EU economies. It is difficult to draw firm detailed conclusions from these studies. Nevertheless, they show that tax compliance costs for international and cross-border activities are substantial. Moreover, such costs are regressive to size, which means that

ceteris paribus they hit small and medium-sized enterprises relatively harder than bigger companies.

Against this background, the following analysis of company tax obstacles in the Internal Market focuses on the company tax issues encountered by groups of companies active in the Internal Market. This includes the tax rules governing mergers and acquisitions, capital gains taxation, transfer pricing, the cross-border off-setting of losses as well as the taxation of all forms of cross-border flows of income, notably dividends. Company tax obstacles in the Internal Market also concern possible hindrances to cross-border economic activity resulting from the taxation of specific forms of remuneration (notably stock options, etc.) and of posted and migrant workers in the EU (notably concerning supplementary pensions). Finally, the specific situation of small and medium-sized enterprises and partnerships is considered and a section on value-added tax completes the overall picture of tax obstacles hampering cross-border economic activity. It is not attempted to provide a detailed classification of any of the obstacles that are presented. Where appropriate, however, indications are made of whether a given obstacle is related to one-off or to ongoing measures and an attempt is made to elaborate on concrete welfare effects.

DIVIDEND TAXATION

A basic concern of companies operating within the Internal Market is the flow of (correctly taxed) income between associated companies free of (additional) tax.

Payments of interest and royalties between associated companies of different Member States are often still subject to withholding taxes that effectively create situations of double taxation. The Commission has already presented a proposal for a directive on this subject [COM(1998)67], and it is expected that this proposal will be adopted in the context of the "tax package". The usefulness of the Directive will however be undermined by its relatively narrow scope. While the Commission proposal would apply to direct or indirect holdings of 25% or more, the Council has decided to limit the scope of the directive to direct holdings.

Cross-border dividend payments still cause problems which are considered in detail below. Industry sees these, despite certain achievements at the EU level, as a major impediment to cross-border activities in the EU.

Double taxation of profits and dividends distributed to corporate and

individual shareholders

There is a general risk of (economic) double taxation inherent in dividend payments. The dividends are paid out of profits which have usually already been subject to corporation tax. At the shareholder level, the dividends are then liable for income or corporation tax. Unless some form of relief applies, this means that the profits are taxed

twice: at company and at shareholder level.

Dividends paid to Member State companies by other Member State companies are in principle covered by Council Directive 90/435/EEC (the "Parent-Subsidiary Directive"), which requires double taxation to be avoided either by exempting the dividends or granting a tax credit equivalent to the tax already paid on the distributed profits. The directive covers dividends paid between associated companies where the recipient company has at least a 25% holding in the company paying out the dividends. However, where dividends are not covered by the Directive, for example where the holding is less than 25%, there is no obligation for Member States to avoid double taxation. Such dividends may still be subject to withholding taxes which often give rise to interest costs (for the period before the other state gives relief) or in certain circumstances even become definitive. In any event, they create additional compliance costs and distort allocational efficiency.

Box 14:

The main provisions of the Parent-Subsidiary Directive

Scope 1. The Directive applies to distributions of profits.

• 2. 

  Profit distributions must be effected between associated companies from different Member States.
• a) 

  Distributions covered by the Directive Profit distributions must meet all the following conditions:

-

they must be between companies from different Member States;

• - 

  they must be effected by companies subject to corporation tax and made to companies also subject to corporation tax;
• - 

  they must be effected between companies with a legal form listed in the Annex to the Directive;
• - 

  they must be made to associated companies with a minimum direct holding of 25% in the capital of the companies paying the dividends. Member States have the option of not applying the Directive if this minimum 25% holding is not maintained for a period of at least two years.
• b) 

  Profit distributions not covered by the Directive The Directive's tax rules do not apply:

-

to distributions of dividends between companies of the same Member State;

• - 

  to distributions of dividends paid to partnerships not subject to corporation tax;
• - 

  to dividends paid between companies subject to corporation tax where one of the companies concerned does not take one of the legal forms listed in the Annex to the Directive;
• - 

  if the company receiving the dividends has a direct holding of less than 25% in the company paying the dividends. Indirect holdings by other companies in the same group are not taken into account;
• - 

  if the 25% holding is not maintained for an uninterrupted period of at least two years. However, although Member States have the option of not applying the Directive if the minimum holding has been maintained for less than two years on the date of payment of the dividends, they must apply it retroactively if the holding is still maintained when the two-year period expires (Court of Justice ruling, 17 October 1996, Joined Cases C-283, 291 and 292/84 Denkavit, Vitic and Voormeer ).

Member States also have the option of not applying the Directive in the case of fraud or abuse.

The Directive's tax rules 1. For the Member State of the company paying the dividends

The Member State of the subsidiary paying the dividends may not charge withholding tax on the dividends paid. Transitional measures were provided for Greece, Portugal and Germany but they are no longer being applied.

2.For the Member State of the company receiving the dividends There are two options:

(a) the dividends received may be exempted from corporation tax;

(b) the dividends received are taxed but a tax credit equivalent to the corporation tax paid by the

Dividends currently not covered by the Parent-Subsidiary Directive

The taxation of dividend payments has not yet been harmonised at Community level. Traditionally there are two systems of taxation, the so-called "classical" system and the "imputation" method. There are also a number of varieties of the two systems but the basic problems can be usefully discussed by examining the effects of the two fundamental methods.

Different methods for taxing dividend payments 51

Under the imputation system the shareholder includes the dividends he receives in the tax base for income or corporation tax but is granted a tax credit equivalent to all or part of the corporation tax paid by the company distributing the dividend on the profits from which the dividends derive.

The classical system, on the other hand, does not neutralise double taxation as shareholders receiving the dividends are taxed on them without being granted a tax credit to offset the corporation tax already levied on the profits of which the dividends form a part. However, in some Member States modified systems are applied and double taxation is mitigated by applying a lower rate of tax to the dividends, by taxing only a proportion of the value of the dividends or by granting a specified imputation tax credit in the amount of a percentage of the dividend received - so called `shareholder relief'.

Moreover, the Scandinavian countries apply a 'dual income tax system' which generally provides for a progressive tax rate for employment income and taxes capital income separately, at a lower proportional rate.

The various systems exist side by side in the Community. However the current trend seems to be towards a partial or total switchover from the imputation to the (modified) classical system or shareholder relief system. Nevertheless, irrespective of the particular system used by a Member State dividends paid to shareholders are often taxed differently depending on whether they are domestic or cross border, i.e. foreign dividends. Two examples illustrate this: first a shareholder in two companies, one domestic, one foreign, receiving dividends may receive on the domestic dividend a tax credit (imputation system) or some form of shareholder relief (modified classical system), but on the foreign dividend an unusable or only partially repayable tax credit (imputation system) or no or a reduced form of shareholder relief ((modified) classical system)). Second, two shareholders resident in two different states who own shares in the same company may be taxed differently, one making use of the tax credit, the other unable to and/or receiving a partial repayment (imputation system) or one receiving shareholder relief, the other not or only a reduced form of relief ((modified) classical system)). In principle these differences could be considered discriminatory, and hence an obstacle to the Internal Market.

Obstacles relating to the imputation system

Usually the tax credit is granted to shareholders receiving the dividends only if they are established in the same Member State or there is a permanent establishment there. The shareholder established or resident in a different Member State is not usually able to make use of it against his own `local' tax. Where this is the case Member States argue that this is because the tax has been paid to another Member State. Neither is the shareholder often able to obtain a repayment from the State levying the initial tax. Member States argue that this is justified on the grounds that residents are liable to tax on dividends and benefit from a tax credit, whereas non-residents are not liable to tax on the dividends and hence not able to benefit from the tax credit.

52

Thus - depending on their precise design - imputation systems often create inequalities of treatment between resident and non-resident shareholders that many commentators in the literature consider discriminatory.

53

Box 15:

Example of how the imputation system works

A French company holds shares in a French company and in a Danish company which both pay a dividend of 100. Corporation tax in France in 1999 was 40% (including surcharges) for large companies. No withholding tax is levied on the dividends paid. The tax credit granted by France in 1999 is equal to 45%

of the dividend received where the shareholder is a legal person.

The French company pays the following corporation tax on dividends it receives:

-

On the dividend of 100 received from the French company: 100 + 45 (tax credit) = 145. Gross tax = 145x 40% (rate of corporation tax) = 58.Net tax to be paid = 58 45 (tax credit) = 13.

• - 

  On the dividend of 100 received from the Danish company: 100 x 40% = 40.

It should be noted that France has meanwhile changed this system and notably reduced the tax credit (25% in 2001; 15% in 2002) as well as additional social security contributions.

It is true that some double-taxation treaties provide in certain cases, generally on the basis of reciprocity, for a tax credit or repayment for foreign shareholders. However, this is not the general rule and often the tax credit or repayment that is granted is smaller than the one received by shareholders who are resident in the same country as the company concerned. Also, where the double-taxation treaties of two Member States effectively extend this tax credit cross-border, this is usually subject to an extremely cumbersome and slow procedure which is, in itself, a disincentive for and obstacle to foreign investment. Moreover, this system does not provide for an equal treatment within the Internal Market and even creates inequalities in the sense that some foreign shareholders of a given company receive a tax credit or repayment whereas others do not.

Clearly, such divergence of treatment and its economic effects are difficult to explain to the shareholders who do not receive a tax credit or repayment and which therefore have a clear incentive to invest in Member States which do give some form of relief.

By contrast, where the situation of non-resident taxpayers is the same as that of resident taxpayers, the tax credit must be granted under the same conditions to resident and non-resident taxpayers (as confirmed by the Court of Justice judgement of 28 January 1986 on case 270/83).

Box 16:

Court of Justice judgement of 28 January 1986 in Case 270/83

Summary of the issue

Insurance companies with registered offices in France, including French subsidiaries of foreign companies, receive a tax credit for dividends paid by French companies. French permanent establishments of insurance companies with registered offices in another Member State are not granted a tax credit.

The Court's ruling

French tax law does not distinguish between companies which have their registered offices in France and French-based permanent establishments or branches of companies whose registered offices are in other countries for the purposes of establishing the corporation tax base. By treating the two forms of establishment in the same way for the purposes of taxing profits, the legislator has acknowledged that there is no objective difference between their positions which could justify different treatment. It is hence not possible to treat companies and permanent establishments differently as regards the granting of a tax credit without creating discrimination contrary to the principle of the freedom of establishment.

This difference in treatment can also seriously affect cross-border mergers and exchanges of shares as the shareholders of a company resident in one Member State can be expected to be reluctant to accept a merger under which the dividends on their new shares will arise in another Member State and therefore without (or with a smaller) tax credit attached. This is a very important practical problem that in economic terms leads to sub-optimal business decisions that imply overall efficiency losses for the Community.

Obstacles relating to modified classical systems or shareholder relief systems

Under this system dividends are included in the taxable profits and no tax credit is granted. There are, however, ways of mitigating double taxation such as taxing dividends at a lower rate than that normally applying or granting other forms of tax relief on dividend payments so called `shareholder relief'.

However such preferential tax arrangements sometimes apply only to domestic shareholdings and not to dividends from foreign shares. Where this is the case, such discrimination usually concerns individuals but may also apply to company dividend payments. As confirmed by the European Court of Justice in its judgement of 6 June 2000 in Case C-35/98 (Staatssecretaires van Financien v. Verkooijen) , these arrangements are not compatible with the free movement of capital.

Box 18:

The Verkooijen judgement

Summary of the issue

Dutch legislation exempts income tax on the first NLG 1 000 of dividends paid by Dutch companies (NLG 2 000 for married couples). This relief does not apply to dividends paid by companies from other Member States.

The Court's ruling

Such a provision is a restriction on the free movement of capital.

Dividends covered by the Parent-Subsidiary Directive

Assessment of the functioning of the directive in practice

The narrow scope of the Directive

national legislation and bilateral double taxation treaties treat profits distributed by such companies to their associate companies as dividends.

To overcome this problem the Commission presented a proposal for a directive on 26 July 1993 (COM(93) 293 final) amending the 1990 Parent-Subsidiary Directive to extend it to all companies subject to corporation tax irrespective of their legal form. The stalemate reached in discussions within the Council in mid-1997 means some companies are still unjustifiably excluded from its scope.

Triangular cases and the calculation of the threshold

Moreover, the Parent-Subsidiary Directive does not cover shares held through permanent establishments. However, such cases are dealt with to a large extent by the above-mentioned interest and royalty payments proposal as permanent establishments often receive or pay interest or royalties.

Although it is rare for permanent establishments to have holdings covered by the Directive, this can occur (see Case C-307/99 Compagnie de Saint-Gobain v. Finanzamt Aachen-Innenstadt ). Views are often divided on how to apply the Directive in the case of permanent establishments and different approaches are taken by national law. It would therefore be useful to clarify this matter.

At present the Parent-Subsidiary Directive applies where the parent company has a direct holding of 25% in the subsidiary located in another Member State. This can create cross-border problems, particularly in the case of restructuring, because indirect holdings are not taken into account to calculate the Directive's threshold. As explained in more detail below, this can have undesirable implications for the internal organisation of groups of companies and hamper restructuring operations. Some Member States do apply thresholds of 5% or 10%, which are lower than the Directive's 25%, but not all.

Credit versus exemption method

If a subsidiary of a parent-company is in turn parent to another (sub-) subsidiary, the question arises whether under the credit method the parent company can credit the tax paid by the sub-subsidiary or whether it has to limit the credit, as currently provided by the directive, to the tax paid by its immediate subsidiary.

flat-rate amount may not exceed 5% of the dividends (article 4 (2) of the directive). The fact that management costs are deductible from the related income is generally accepted. However, when the flat-rate method of calculation is adopted it is sometimes criticised inasmuch as companies who incur management costs under 5% are unable to deduct their true costs.

54

Problems related to the implementation of the directive by Member States

Some national legislation seems to contain provisions whose compatibility with the Parent-Subsidiary Directive is doubtful. This applies in particular to national legislation applying the "anti-abuse" clause contained in the Directive. Article 1(2) of the Directive does not preclude the application of domestic or agreement-based provisions to prevent abuse or fraud, but the implementation of this is not always consistent.

Some national legislation appears to make the application of the Parent-Subsidiary Directive subject to conditions which are not contained in the Directive. For example the provisions of the Directive may not be applied when the capital of a Community company is held by non-Community residents. This is considered to be a presumption of tax evasion or avoidance and the provisions of the Directive only apply when proof is provided to the contrary. The existence of such restrictive national legislation reduces the intended effect of the Directive.

Conclusion

The taxation of dividends in the EU is still not completely in line with Internal Market requirements. There are clear examples of both economic and legal double taxation at the level of both the corporate and individual shareholder. The Parent-Subsidiary Directive seems to work reasonably well but only in a limited sense. It does not cover all the companies it could and it does not address all the situations it could. Its implementation at Member State level also raises doubts, in particular in relation to specific anti-abuse rules. In addition to the obstacle of "double taxation", obtaining relief where it is available involves an unnecessarily high compliance cost. Economic decisions such as mergers or investments are distorted and efficiency at the EU level is therefore potentially reduced.

representatives strongly deplore the fact that company tax arrangements often create obstacles to a commercially straightforward organisation of their business.

55

Generally, cross-border restructuring is one of the few areas of direct taxation where there has been harmonisation at Community level. The relevant Community instrument is Council Directive 90/434/EEC on the common system of taxation applicable to mergers, divisions, transfers of assets and exchanges of shares concerning companies of different Member States, more commonly known under the name "Merger Directive". Other important tax aspects of cross-border business restructuring are not regulated in Community law. The following section considers both aspects.

Box 19:

The main provisions of the Merger Directive

Scope

Companies concerned: Companies from different Member States must be involved in the restructuring operations. These companies must

-

be subject to corporation tax;

• - 

  take one of the legal forms listed in the Annex to the Directive;
• - 

  be registered for tax purposes in a Member State.

The following are excluded from the Directive:

-

restructuring operations involving companies from the same Member State;

• - 

  operations involving companies which are not subject to corporation tax (partnerships, natural persons, etc);
• - 

  companies which are subject to corporation tax but take a legal form not listed in the Annex to the Directive.

Restructuring operations concerned:

-

mergers of companies;

• - 

  divisions of companies;
• - 

  transfers of assets whereby a company transfers one or more branches of its activity to another company in exchange for the transfer of securities representing the capital of the company receiving the transfer;
• - 

  exchange of shares whereby a company acquires a majority of the voting rights in a company in exchange for the issue to the shareholders of the company acquired of securities representing the capital of that company in exchange for their securities.

The Directive's tax rules

Mergers, divisions and transfer of assets

No capital gains tax charged on assets received

Member States cannot charge capital gains tax on assets received. However, the receiving company must compute depreciation and capital gains according to the rules that would have applied to the transferring company if the merger, division or transfer of assets had not taken place. Special rules apply to triangular operations (mergers, divisions or transfer of assets by a company with a permanent establishment in another Member State).

No taxation of tax-exempt provisions or reserves

These provisions or reserves are not taxed at the time a merger, division or transfer of assets takes place. However, the receiving company must assume the rights and obligations of the transferring company as regards the restatement of these provisions or reserves in its taxable profit.

Transfer of losses

Member States must ensure that cross-border operations covered by the Directive are subject to the same national rules relating to the transfer of losses from the transferring company to the receiving company that they apply to domestic mergers, divisions or transfers of assets.

Exchanges of shares

No tax is charged at the time shares are exchanged but Member States may tax the profit generated by the subsequent disposal of shares received in exchange in the same way as they would tax the profit from the disposal of the shares exchanged.

Anti-abuse clause Member States have the option of not applying the Directive or excluding profit:

-

if the merger, division, transfer of assets or exchange of shares has as its principal or one of its principal objectives fraud or tax evasion;

• - 

  if the operation is intended to prevent employees from being represented on the company's management bodies.

Limits to the tax solution regulated in the Merger-Directive

Although the Merger Directive has improved the situation, it is not wholly satisfactory and does not enable companies to undertake cross-border restructuring operations in the way they would wish. Despite the Directive, cross-border restructuring operations can still involve significant tax costs.

companies who are unable to improve their organisation. Currently only cross-border transfers of assets or exchange of shares can be undertaken in all Member States.

The adoption of the European Company Statute will, however, change this situation and clearly calls for adaptations of the existing body of tax law, both at EU level and in Member States, in the area of company taxation. Therefore, the are good reasons to believe that the lack of EU company law on mergers will soon no longer raise problems:

as from 2004 cross-border mergers will be possible under the form of European

Companies.

The narrow scope of the Directive

At present the Merger Directive does not cover certain company restructuring operations even though the companies in question may be wholly subject to corporation tax. Companies which have adopted a corporate form created after 1990 (e.g. simplified joint stock company in France) or corporate forms which for various reasons were omitted from the 1990 list annexed to the Directive (Belgium cooperative societies, certain Irish banking companies, etc.) are excluded. Similarly partnerships, which in some Member States are or may be subject to corporation tax, are excluded from the Directive's scope.

The Commission presented a proposal for a Directive [COM(93)293] to remedy this situation. It would amend the 1990 Merger Directive to extend it to all businesses subject to corporation tax irrespective of their legal form. However, a stalemate was reached in discussions within the Council in mid-1997. The main problem concerns the extension of the Directive's scope to partnerships. Consequently businesses subject to corporation tax are still excluded for no good reason from the Directive's scope. The conclusions of the ECOFIN Council of 26 and 27 November 2000 state that priority should be given to updating the list. It goes without saying that the new list will have to include the European Company statute.

Box 20:

Insufficient coverage of restructuring operations by the Directive

There are doubts about whether the "subsidiarisation" of companies' branches is covered by the Merger Directive. This involves the transfer by means of a transfer of assets within the meaning of the Directive of a permanent establishment located in a Member State to a new company established in the same Member State. For example, a company headquartered in Member State A has a permanent establishment in Member State B and wishes to transform this permanent establishment into a company of country

B. If Article 4 of the Directive which applies to transfers of assets under Article 9 is applied literally the Directive would not cover such operations. Article 4 of the Directive makes the deferral of capital gains tax conditional on the assets continuing to be effectively connected with a permanent establishment situated in the Member State of the transferring company, i.e. Member State A in the above example.

The Commission takes the view that, given the Directive's purpose, Article 4, which was designed to cover mergers and divisions of companies, cannot be intended to exclude the conversion of permanent establishments into subsidiaries from the Directive's scope. The conversion of a branch into a subsidiary does not affect the taxation right of the State where the former permanent establishment was located and cannot be interpreted as excluding it from the Directive's tax neutrality principle.

Generally, the objective of the 1990 Directive is to guarantee the tax neutrality of restructuring operations covered by the Directive and, at the same time, safeguard Member States' financial interests. Consequently - for mergers, divisions or transfers of assets - the assets transferred by the receiving, divided or transferring company must remain effectively connected with a permanent establishment of the receiving company located in the Member State of the acquired, divided or transferring company. The Directive does not apply when there is no permanent establishment.

Box 21:

Example of cross-border restructuring not covered by the Directive

of capital gains taxation whereas it is common that in the framework of group taxation schemes within Member States such transfers do not give rise to capital gains taxation.

Unsatisfactory outcome of the application of the Merger Directive

Even where restructuring operations can be undertaken the results are not always satisfactory. A brief analysis of national measures transposing the Merger Directive has revealed important divergences. This is due to two different reasons.

In some cases national legislation seems to have adopted transposition measures which raise doubts concerning their compatibility with the Directive. In others, divergent legislation has been adopted because of the Directive's lack of clarity on some - particularly important - points. The Directive has not necessarily been misinterpreted but the lack of uniformity in national legislation is unhelpful and undermines the practical usefulness of the directive.

Doubts concerning the incompatibility of some national legislation with the

Directive

The present study does not make an exhaustive analysis of national transposition measures but the following impediments for cross-border economic activity are very

apparent.

The anti-abuse clause

Some national legislation makes the application of the Merger Directive subject to conditions which are not laid down in the Directive. According to the Member States concerned these are based on Article 11(1)(a) of the Directive which allows them not to apply the Directive or to deny its benefit where the merger, division, transfer of assets or exchange of shares has as its principal objective or as one of its principal objectives tax evasion or tax avoidance ("anti-abuse-clause").

The case most often cited is where a number of Member States require that shares received under a transfer of assets or an exchange of shares be kept for a certain period which varies from three to seven years. The rapid disposal of shares received as a result of a transfer of assets or exchange of shares could be an abuse within the meaning of Article 11 of the Directive. However, in its judgement in Case C-28/95 Leur Bloem

(1997), the European Court of Justice ruled that such abuse had to be assessed on a case-by-case basis. A blanket refusal to apply the Directive where shares received are disposed of before a particular deadline without giving taxpayers an opportunity to prove that such disposals are not of an abusive nature is therefore unlikely to be be consistent with the Directive. Moreover, minimum holding periods that are particularly long - up to five or seven years after the initial operation - appear to be difficult to justify on the grounds of preventing abuse.

may subsequently tax the gain arising out of the subsequent disposal of shares received in the same way as the gain arising out of the transfer of shares existing before the acquisition.

In some Member States shareholders exchanging shares in the acquired company for shares in the acquiring company are taxed before disposing of the shares received in the acquiring company, especially if shares in the acquired company are transferred by the acquiring company before shareholders dispose of shares in the acquiring company. Such an approach does not seem to be fully in line with Article 8 of the Merger Directive which does not provide for any form of taxation before a shareholder sells shares in the acquiring company received in exchange for shares in the acquired

company.

Box 22:

Example on the taxation of exchanges of shares before the disposal of shares received in exchange

Company A of Member State A holds shares in company X. Company A exchanges its shares in company

X with company B situated in Member State B for shares in company B. Article 8(1) of the Directive provides that the exchange of X shares for B shares may not give rise to taxation of associate A which relinquishes its shares. However, Article 8(2) stipulates that the Member State of company A may still apply capital gains tax on the disposal of B shares by A in the same way as the gain arising out of the disposal of X shares existing before the exchange of shares.

The imposition by Member State A of capital gains tax on X shares which were not taxed in the exchange of shares before company A had disposed of B shares received in exchange is questionable. This would be, for example, where company A is taxed by Member State A when company B disposes of X shares even if it had not disposed of B shares received in the exchange of shares.

Problems not resolved by the Directive

Double-taxation of capital gains

• - 

  Exchange of shares: the shareholder is taxed on the capital gains on which taxation

had been deferred when shares in the acquiring company received in exchange for shares in the acquired company are disposed of; the acquiring company is taxed on the same capital gains when shares in the acquired company are sold.

Box23:

Example of double taxation in the case of transfers of assets

Company X which has its activities in Member State A transfers assets to company Y situated in Member State B. Following this transfer of assets the activities are transferred to the permanent establishment of company Y situated in Member State A. Company X receives in exchange for the transfer of assets shares in company Y. The assets transferred by X to Y have a net book value of 100. At the date of the transfer these assets are worth 300. The capital gains on the assets transferred are hence 300 - 100 = 200. Company X receives shares in company Y which have a value corresponding to that of the assets transferred, i.e. 300.

Article 4(1) of the Directive prohibits Member State A from taxing company X on capital gains of 200 realised on the transfer of assets. However on the disposal of the assets transferred by company X, the permanent establishment of company Y situated in Member State A may tax the capital gains in the same way as company X would have done if the transfer of assets had not gone ahead, including on the capital gains of 200 which had not been taxed on the transfer of assets.

If Member State A taxes the capital gains on the disposal of Y shares received by company X on the basis of the book value of the assets transferred (100) and not their acquisition value of 300, Member State A will tax the capital gains of 200 which had not been taxed on the transfer of assets twice: 1. in the hands of company Y, when this company disposes of the assets transferred, and 2. in the hands of company X, when this company disposes of Y shares received in exchange for the transfer of assets.

Such double taxation would not appear to conflict directly with the wording of the Directive as it is not very specific about the value for tax purposes which must be attributed to shares received by the acquiring company in exchange for the transfer of assets or, in the case of exchange of shares, for shares in the company acquired received by the acquiring company. Nevertheless in the cases described above it is clear that the same capital gain is taxed twice, albeit for two different shareholders. This is not acceptable from an Internal Market perspective and runs against the spirit of the Directive.

Other transfer taxes

Directive 69/335/EEC of 17 July 1969 (as amended) concerning indirect taxes on the raising of capital has not allowed capital duty to be levied on mergers since 1986. However Article 12(1)(b) authorises, by way of derogation, Member States to charge transfer taxes, including land registration taxes, on the transfer of immovable property or businesses to companies. In its judgment of 11 December 1997 in the SIF Case C- 42/96 the European Court of Justice acknowledged that this provision allows registration, mortgage and land registration fees to be charged in connection with the capital increase of a company brought about by the contribution of immovable property provided that such taxes do not exceed those applicable to similar transactions in the Member States charging them.

Such taxes which can often account for 10% of the value of the immovable property contributed in company mergers or transfers of assets may well increase the cost of restructuring operations covered by the Merger Directive or be an impediment to them. The panel of experts assisting the Commission with this part of the study strongly underlined the importance and practical relevance of this problem.

Box 24:

Example of the cost of a cross-border restructuring operation

The following recent real-life example of a large European multinational group has been presented by business representatives to the Commission. It can only be re-produced in an anonymous form. This group calculated the tax costs it would incur to restructure its European activities within one single company with permanent establishments in the Member States. The cost would be even higher if the Merger Directive had not been adopted.

The starting point is a Dutch holding company with shares in five national sub-holding companies situated in different Member States which themselves have holdings in three operating companies in the same Member State. The objective is to create a structure under which the parent holding company of the group would remain unchanged but would hold shares in only one single "European" company with five permanent establishments. The five national sub-holding companies and 15 operating companies would cease to exist. The means of arriving at this reorganisation would be as follows:

1.

to the difficulties created in restructuring operations by the fact that only direct holdings are taken into account when calculating whether the Directive's threshold is met. The sole inclusion of direct holdings may hamper restructuring operations.

A company may, for instance, decide not to undertake a cross-border transfer of assets to a company whose capital is held by another company in the same group if the shares to be received in exchange account for less than 25% of the capital of the beneficiary company. The dividends that it will receive from this company are not covered by the Parent-Subsidiary Directive even though the subsidiary is wholly owned by other companies in the same group.

Similarly, a transfer of assets may dilute the shareholdings in the company receiving assets from other companies in the group and reduce their holding below the 25% threshold. This would mean they were no longer covered by the Parent-Subsidiary Directive. This also forms an obstacle to restructuring.

Box 25:

Example on restructuring operations and dividend taxation

Company A in Member State A belonging to a group of companies A holds 26% of company B's capital in Member State B. The dividends paid by company B to company A are covered by the Parent- Subsidiary Directive. Company A holds 100% of the capital of company C situated in Member State C. Company C transfers assets to company B and receives in exchange 10% of company B's original capital, i.e. 9.09% of the capital after a capital increase (10% x 100/110). Following this capital increase company A's holdings in company B is reduced to 26% x 100/110 = 23.63%.

The transfer of assets has the following consequences:

1.

The dividends paid by company B to company C are not covered by the Parent-Subsidiary Directive although company C holds only 9.09% of company B's capital. The holding of other companies of the group in company B, here the participation of A in B, are not taken into account.

• 2. 

  Company A's holding in company B is reduced from 26 to 23.63%. Consequently the dividends paid by company B to company A are longer covered by the Parent-Subsidiary Directive. The holding of C in B is not taken into account for calculating the threshold of 25% although company

A holds 100% of C.

requirement and results in inefficient structures and welfare losses for the Community as

a whole.

CROSS-BORDER LOSS-COMPENSATION

Loss-compensation in general and the difficulties encountered by businesses with loss- offset are a key element in the analysis of tax obstacles to cross-border economic activity in the Internal Market. The panel assisting the Commission services with this part of the study has identified the absence of cross-border loss-relief or full consolidation at EU level as one of the major obstacles that requires action as a matter

of priority

• 56. 

  The relevant issues are therefore presented in some detail. Given the

considerable complexity of the issues, first some explanations about the technical possibilities for cross-border loss-offset and the current situation in Member States are given.

The tax treatment of domestic losses and foreign losses generated by

permanent establishments and subsidiaries

Domestic losses

The rules on loss-compensation differ substantially for companies within a group (i.e. incorporated companies with a proper legal personality) and for unincorporated separate units of one company (i.e. branches). Broadly speaking, holdings in other companies constitute subsidiaries whereas business units of a single company abroad form permanent establishments.

The possibility to set off losses against profits for assessing the tax liability of a single domestic company is a basic feature of any company tax system. On the domestic level, it is available in all Member States and by definition includes losses from domestic branches. The detailed conditions, however, differ substantially. All Member States' tax legislation allow, for varying periods, the carry-forward of losses. Only a few Member States allow for loss carry-back. The basic functioning of the arrangements is illustrated in the following box:

Box 26:

Loss-compensation on the domestic level

In a given period, a single domestic company establishes the taxable income by taking into account all profits and losses of the company headquarter and all branches in the national territory. The entity is taxed as one company and thus by definition full loss-offset is ensured.

Company X

Member State A

X1X2

If the overall result of the company is negative, the loss can be carried forward to future tax periods (or carried back to previous ones) and thus reduces the taxable profits in other years.

Most Member States permit domestic group taxation (profit consolidation) but the details differ substantially, for example concerning ownership thresholds and whether both direct and indirect holdings can be amalgamated. More fundamentally, three Member States (Belgium, Greece, Italy) do not have such schemes at all and within these Member States a branch structure would have to be used for a 'group' to achieve consolidation. The basic functioning of the arrangements are illustrated in the following

box:

Box 27:

Profit consolidation on the domestic level

Domestic group or consolidation taxation schemes provide full effective consolidation of profits and losses. Tax is assessed for the group and not for the individual corporations forming the group. Therefore, losses of the parent can be offset against profits of the subsidiaries (downstream vertical) and the other way round (upstream vertical) and if, say, the parent is making neither a loss nor a profit or when its profits are not sufficient for full loss-absorption, losses in one subsidiary can be offset against profits in another subsidiary (horizontal).

Losses in Permanent Establishments

As regards cross border situations most Member States permit, subject to certain conditions, losses from Permanent Establishments to be relieved. Immediate relief can be obtained under two methods

Credit/Imputation method

The Permanent Establishment losses are included in the Head Office's results and the net profits taxed. If the Permanent Establishment makes profits any tax it has paid is creditable against the Head Office's tax liability. The basic functioning of the arrangements is illustrated in the following box:

Box 28:

Treatment of losses of permanent establishments (foreign branches) - credit method

A single domestic company has foreign branch(es)/permanent establishment(s).

The parent company includes the income (positive or negative) of its permanent establishments and receives a tax credit equal to the tax paid by the permanent establishment.

Member State A Company Xprofits/losses X +/- profits/losses X1 = taxable income

Þ tax in A calculated

but reduced by tax in B

Member State B

PE

X1

profits/losses X1

• = 

  taxable income

Þ tax in B

Box 29:

Treatment of losses of permanent establishments (foreign branches) - deduction/reintegration method

The losses incurred in a certain period by a permanent establishment (or a subsidiary) of a company are credited. In the subsequent years, the profits of the permanent establishment (or subsidiary) are also included in the taxable base of the company.

year 1year 2

Member State A Company X

profits X profits X

- losses X1 + profits X1

= taxable income = taxable inc.

Þ tax in A Þ- tax in A

Member State B

PE

X1

losses X1 profits X1 taxable income

Þ tax in B

Therefore in both cases the losses of a Permanent Establishment are effectively recognised at the Head Office level as incurred, as are any subsequent profits.

The rules are often complex and differ between Member States. For example not all Member States permit unused tax credits to be carried forward and therefore where a Permanent Establishment is profitable and pays tax, but the combined Head Office and Permanent Establishment tax liability is less than the local tax suffered by the Permanent Establishment part of the tax credit may be lost. Many Member States limit the tax credit to the hypothetical domestic tax which would have been due on the Permanent Establishment profits had it been a domestic branch rather than a foreign Permanent Establishment.

Box 30:

Treatment of losses of permanent establishments (foreign branches) - exemption method

Member State A Company Xprofits/losses X

Þtax or loss-carry-forward

Member State B PE X1

profits/losses X1

Þtax or loss-carry-forward

Losses in subsidiaries

Only two Member States (DK, F) extend immediate cross border loss relief to subsidiaries (again this is subject to certain conditions) both under the deduction/reintegration method.

A measure of relief is available in some Member States who permit the parent company a tax deduction for a write down in the carrying value of its investment in a subsidiary when it has made losses. However, such a deduction is limited to the carrying value in the balance sheet, and would only be available after the losses had been incurred (and the loss of value was considered permanent).

In most cases there is no provision for recognising losses incurred in a foreign subsidiary and the situation can arise where the subsidiary makes losses in excess of the parent company's profits, ie the group is in an overall loss position, but tax is still due because the parent is unable to relieve the losses against its profits and reports a taxable profit.

Box 31:

The definition of "losses"

Logically, the issue of loss compensation cannot be separated from the general determination of the taxable base and taxable income leading to a loss. Rules differ significantly between Member States in this respect. There are various aspects to this: the definition of various categories of income, the recognition of business expenses for tax purposes, the interrelation between specific (positive or negative) elements of the tax base and the deductibility of the overall loss. Against this background, it is also important to make sure that losses are not deducted (or offset) twice, in two different countries.

Some examples highlighting the difficulties and differences in this area are illustrated in the following

questions: Are capital gains and losses included in the definition of "loss"? What about imputed costs (e.g. the theoretical income from letting a house occupied by the owner)?

Are representation costs deductible? To what extent? Can revenue in all income categories be offset with expenses from all income categories or are there

limits (e.g. between "active" and "passive" categories of income)?

Which is the tax period for assessing the revenue and for assessing the deductible expenses? When are revenues and expenses (gains and losses) realized, i.e. accounted for, for tax purposes? What is the territorial scope for revenues and expenses (gains and losses) to be taken into account?

Notwithstanding the different treatment in Member States, it should be noted that for many of the above aspects a certain de facto approximation has taken place in Member States over the last few years. But the inclusion of some losses seems to remain highly controversial (e.g. capital losses).

Consolidation of profits and losses

Finally, even within Member States there are a number of varying group taxation or consolidation schemes. In some of them, a parent-company will include the income of its subsidiaries in its profit-determination and pay tax accordingly, in others losses may be `surrendered' to other group companies. As mentioned above, in at least two Member States group taxation provisions are extended to foreign subsidiaries, thus extending the potential loss-relief cross-border (Denmark, France). In Denmark this applies to subsidiaries and in France this applies because of a different notion of the territoriality principle.

company returns to profit. Similarly if the amount of losses which a subsidiary can surrender to its parent is restricted to the current profits of the parent the routes permitted become potentially important.

The same principles apply to cross border operations, and when assessing the effectiveness of the various techniques currently available for cross border loss compensation these factors must be taken into account.

The basic functioning of the consolidation is illustrated in the following box:

Box 32

Cross-border consolidation of profits and losses

The existing consolidation schemes work vertical upwards, i.e. losses of the subsidiaries are taken into account on parent level (but not the other way round).

Member State A

Parent company

X

Member State B

sub. X1sub. X2

Problems created by the absence of cross-border loss-compensation

Differing loss-compensation arrangements as factor for localisation decisions

As regards the domestic loss-compensation arrangements, carry-forward and carry-back are important criteria in deciding whether to take an economic risk in a country. Small businesses are particularly hit by not being able to carry back losses, and small start-up companies in particular risk losing the benefit of losses which they are not able to carry- forward long enough for offsetting (or only when their value has effectively diminished). According to the indications made by members of the expert panel assisting the Commission services with this part of the study, the increasing importance of cost centres (implying either cost-sharing/pooling arrangements or royalty payments)

in modern EU-wide branch structures tend to exacerbate these problems. Cost centres for R&D are a good example. In this context, however, it should not be overlooked that cost centres, whether organised as a subsidiary or permanent establishment, often also give rise to transfer pricing problems which are considered below.

Bias towards domestic investment and investment in larger Member States

The current situation also implies a clear incentive for companies to invest in those countries where sufficient taxable profits are available against which future losses can be set off. In the absence of cross-border relief, investments in a location with an existing tax base, mostly in bigger Member States, will tend to be favoured. Moreover, the present arrangements put foreign investment at a disadvantageous position compared to domestic investment. Loss-compensation is generally available domestically but not always in cross-border situations and even when cross-border loss compensation is available, the general conditions (timing, availability of profits etc.) are more generous in the domestic context. Whenever, due to the above arrangements and their interplay, losses that cannot be absorbed locally (either immediately or in later periods via carry- forward) cannot be offset against profits within the same group of companies, (economic) double taxation occurs. When loss-relief can only be obtained in the future when an activity becomes profitable, there is no double-taxation but in cross-border situations foreign losses may be offset later than domestic ones. The resulting interest cost for the companies concerned are considerable and of high practical relevance.

Box 33:

Group consolidation and loss-compensation:

Acquisition of a start-up company at home and abroad

Nb: The following arrangements are originally based on the tax law of a specific Member State but similar examples can be constructed for other Member States as well.

Company A possesses all shares of company B, purchased at a value of 500.000 . A generates a profit of 1.000.000 a given year while B, a start-up company, suffers a loss of 600.000 . If company B is resident in the country in which A is registered, under a Consolidated Tax Regime both companies are considered as a single taxpayer (subject to certain conditions, e.g. if A holds at least a certain percentage of the B shares). The consolidated group tax base is the result of the sum of the components' tax bases, i.e. 400.000 in this case. The losses of the domestic subsidiary are thus entirely taken into account for determining the parent company's profits. A possible excess loss can be carried forward up to ten taxable periods.

This group taxation is only available under the condition that all companies are tax residents in the same country. If B is not tax resident, its tax base of - 600.000 has to be taken into account under B's domestic tax law. However, A will, in certain Member States, be allowed to build up a provision reflecting the lower value of B's shares. Generally, such a tax effective provision (book-reserve) for shares quoted on the stock market is possible when the market value is lower than the acquisition value. For non-quoted shares the provision is deductible for the difference between the theoretical value at the beginning and the end of any given year. In both cases, the maximum provision possible is normally limited by the share acquisition value (i.e. the loss in value is restricted to cost). Consequently, A's tax base amounts to 500.000 , resulting from the profits of 1.000.000 , reduced by the provision of 500.000 . If the subsidiary is not tax resident, losses are taken into account through tax deductible provisions subject to the shares' acquisition value. The excess cannot be carried forward.

Even when the losses of foreign branches (permanent establishments) are immediately transferable to the head office, the overall tax situation still works in favour of local branches. For instance, in countries operating the tax credit method for permanent establishments, the losses of foreign branches (permanent establishments) are an integral part of the head office determination of the taxable base as are those from any domestic branch. When, however, losses are made in the country of the head office and taxable profits arise in the foreign branch country, the profits of the permanent establishment often reduce the overall tax loss available for carry-forward in the home country, even though tax has already been paid in the country of the permanent establishment. This boils down to effective double-taxation of identical income in the two Member States involved, which can only be avoided if the tax credit granted in the headquarter state for compensating the foreign tax can be carried forward to subsequent tax periods. This is currently not a common feature of Member States' tax systems.

Box 34:

Example on loss-compensation: Losses in a domestic branch vs. losses in a foreign branch

A Belgian company sets up a branch in Rotterdam. The losses which the company suffers in Belgium are available for carry-forward and set-off against future profits, but only after being reduced by the profits made in the foreign branch (even when these profits have been taxed there). If the branch is set up in Antwerp, the amount of losses of the head office that may be set off against future profits is the same as for a branch in Rotterdam but the profits of the branch are not taxed. This situation may also illustrate why industry representatives constantly suggest that the permanent establishment should be able to offset the head office losses against its profits.

Relation to double-taxation conventions

Generally, the varying availability of loss-compensation is often influenced by the arrangements in double-taxation treaties and the double taxation relief method (exemption vs. credit method) applied. It may therefore change even for identical transactions within one country, thus creating significantly different competitive conditions for competing EU businesses. The respective advantages and disadvantages of the exemption method and the foreign tax credit system are not discussed here. However, some points for discussion should be noted. The above example comparing domestic and foreign branches illustrates a peculiarity of the functioning of credit methods. Moreover, it is logical to argue that under the exemption method foreign branch profits are not subject to tax in the head office country and that consequently the losses should not be taken into account either. In any case, this situation provides an obstacle to cross-border expansion inasmuch as the benefit of start-up losses is lost (or the losses can only be offset after a long period when they have lost their value).

It follows from the foregoing that the differing loss-compensation arrangements do not only impact on the decision on where to locate but also on how to carry out an investment. This is, for instance, an effect of the different treatment of cross-border losses of permanent establishments (which can immediately be transferred to the parent- company) and subsidiaries (which can generally not be offset against parent profits). Where operations are initiated abroad with foreseeable substantial start-up losses, the possibility of cross-border loss compensation offered by branches (forming permanent establishments) will induce companies to opt for this legal form rather than for immediate incorporation of the foreign operation (as a subsidiary), even though the latter may well be the preferred structure for other reasons. It is therefore not uncommon to see permanent establishments being transformed into subsidiaries when they become profitable (an activity is initially operated via a permanent establishment so that its losses can be offset in the head office but when it moves into profitability it is converted into a subsidiary). It should however not be overlooked that in many situations companies effectively do not have a choice between the two legal forms of running foreign operations, notably in specific sectors (e.g. banking, retail trade etc.) or for other external reasons.

Box 35:

Cross-border compensation of losses in permanent establishments:

the Futura Participations case and the AMID case

The issue of cross-border losses was first considered by the European Court of Justice in Case C-250/95 Futura Participations, where it held that Luxembourg was entitled to demand that losses of a French company with a permanent establishment in Luxembourg should have an economic link with its territory in order to be taken into account. The limitation was said to be justified because Luxembourg only exercised source taxation. The Court considered that, provided residents who were liable for tax were not given more favourable treatment, this condition was compatible with Article 52 of the Treaty because it was in line with the principle of tax territoriality.

The second time that the European Court of Justice dealt with the issue was in its judgment of 14 December 2000 in the case C-141/99 Algemene Maatschappij voor Investering en Dienstverlening NV (AMID) v Belgische Staat. In this case, the Court had to consider the following situation: The Belgian tax legislation establishes an order of set off for losses which in the AMID case had the effect of requiringAMID's Belgian Head Office losses be set off against its Luxembourg Permanent Establishment's profits, even though these profits were exempt from Belgian tax under the Double Tax Agreement between Belgium and Luxembourg. Therefore these losses were not available for offset against subsequent Belgian Head Office profits, and in effect were never relieved against taxable Belgian profits. In contrast, hadAMID's permanent establishment been Belgian then the Head Office losses would have been relieved against the permanent establishment profits. Despite the fact that the legislation could also produce a situation where a company benefited (a profitable Belgian Head Office with a lossmaking Luxembourg permanent establishment could relieve the Luxembourg losses against both the Belgian Head Office profits for Belgian tax purposes and also relieve them against subsequent Luxembourg permanent establishment profits for Luxembourg tax purposes) the Court found that AMID suffered an inequality of treatment in relation to companies without establishments outside Belgium. Since there was no objective difference between such companies and no justification for this discriminatory treatment it created a hindrance to the freedom of establishment guaranteed by Article 52 of the Treaty and such legislation is therefore precluded by the Treaty. The AMID case illustrates the application of the non-discrimination principle in relation to the taxation of permanent establishments. It may have wider implications for the situation where losses incurred by foreign subsidiaries are currently not relievable against domestic profits, but losses incurred either by foreign permanent establishments or by domestic subsidiaries are relievable.

The European Court of Justice has not yet to rule on other cases in this area. In particular, it did not have occasion to consider whether a residence state applying the credit method (and hence claiming taxing rights over foreign profits) must take account of foreign losses incurred in another Member State.

substantial amounts of corporate taxes in specific Member States while the overall EU group result was negative. UNICE gives the example of a company having in 1993-1995 overall losses of 880 m. ECU in various EU Member States whereas it had taxable profits in other Member States amounting to 870 m. ECU, resulting in the payment of corporate taxes in the latter countries of 320 m. ECU. Clearly, full cross-border loss- compensation would have prevented this tax payment in this period. It also mentions one particular company with overall profitability in Europe but losses in some Member States and profits and others. This overall company loss carry-forward is reported to be 820 m. ECU (1996) which gives an idea about the possible cost-savings that would be achievable via immediate loss-compensation.

Loss-compensation arrangements are also very important in the case of mergers and acquisitions. It is evident that while carrying out cross-border business restructuring operations, companies try to preserve accumulated pre-conversion losses and ensure their tax-effectiveness, i.e. current or future absorption. Generally, in some countries, costs arising from the acquisition and holding of foreign equity are not or only to a limited extent deductible which can lead to significant taxation of assumed gains. Under specific circumstances, however, the post-merger situation can offer more possibilities of loss-compensation than the pre-merger situation (e.g. when two associated companies can reflect losses only via provisions before the merger whereas, at least in many Member States, the permanent establishment created through the merger offers full cross-border offsetting of losses).

Box 36:

Survey of losses on cross-border activities within the EU by the Federation of Swedish Industries 57

Set up of the survey

In order to investigate the magnitude of this obstacle the Federation of Swedish Industries has made a survey among its member companies. The Federation has approximately 6 000 member companies from the manufacturing industry, transportation, telecommunication and information technology. The member companies count for approximately 90 % of the industrial export from Sweden. The survey was carried out by sending out a questionnaire concerning the frequency of losses on cross-border activities and inquiring to what extent it had been possible to set off the losses suffered against profits in other Member States. It was also asked to what extent these difficulties had influenced their activities and/or the organisation of their businesses. The questionnaire was distributed to all member companies (or groups of affiliated companies) having more than 25 employees, in all 1086 companies (only one questionnaire per group of companies). Out of these 1086 companies 706, or 65 %, have answered the questionnaire. A limited random telephone survey among the non-answering companies indicates that the main reason for not answering was the fact that they had no affiliate companies in any other Member State than Sweden. Therefore they had considered the survey not applicable to them.

• - 

  Out of the 216 companies having activities also in other Member States 172 companies, that is 81%, have suffered losses in one or more Member States. 1/3 of the companies have suffered losses several times.
• - 

  In 166 companies, or 96% of the companies having suffered cross-border losses, it had not been possible or only partly possible to set off these losses against profits of other companies in the group.

In 56 % of the cases this has resulted in permanent double taxation to at least some degree. Together with the answers to other questions, this shows that 77 % of the companies having cross-border activities within the European Union have suffered from a higher tax burden because of the difficulties of setting off losses against profits in other Member States. In more than 50 % of the cases the result has been permanent double taxation to some degree and in the remaining cases temporary liquidity and interest losses.

• - 

  Of those companies suffering losses on cross-border activities 73 companies indicated that the difficulties to set off losses against profits in other Member States had influenced the organisation or structure of their activities.
• - 

  A breakdown of the figures in categories of small companies (less than 100 employees), medium sized companies, (100-500 employees), and companies with more than 500 employees has been made. The breakdown shows that small companies have fewer cross-border activities, only 13% of the companies, than larger companies. However, the small companies have had greater difficulties setting off losses against profits in other Member States. Those difficulties have not influenced their organisation to any substantial degree (quite naturally as they have fewer options for restructuring their businesses).

Comments by participating companies The companies were invited to give comments on the causes and effects of the difficulties of cross-border loss-setoffs. All together, 114 comments were given. Here they have been classified into 10 different categories.

-

According to 29 comments the cross-border loss problems have caused the companies to choose a tax motivated organisation of their businesses (e.g. national subgroups, holding company structures), instead of a business motivated structure such as an organisation by business sectors. In some cases a branch structure has been chosen instead of having the foreign operations organised in separate subsidiaries which otherwise would have been the natural organisation.

• - 

  According to 21 comments the cross-border loss problems gave rise to a higher over all tax burden for European businesses and have also caused liquidity problems and reduced expansion possibilities. Also, the survival of the business have been threatened. In two cases the problems have led to a liquidation of the business.
• - 

  In 16 comments transfer pricing problems have been listed as one of the main reasons for the cross- border loss problems.
• - 

  In 12 comments a system for group taxation within the Union is recommended as a solution to the cross-border loss problems.

Conclusion

The current situation in the EU leads to the taxation of company profits which, where foreign investment is involved, do not reflect the overall result of the business activities.

In certain cases, this could result in discriminatory treatment. The non-availability or limitation of cross-border loss-compensation thus results in (economic) double- and over-taxation. Where limited cross-border compensation is available specific corporate structures may be required, thus influencing commercial decisions. Industry considers this one of the most important impediments to cross-border economic activities and in conflict with the very concept of the Internal Market.

Moreover, there appear to be good grounds to conclude that generally the company tax law of Member States contains a bias towards favouring domestic investment, thus indirectly hampering cross-border economic activities. This is in particular true in the larger Member States, because the domestic market of such States may be large enough to accommodate one important enterprise, while an enterprise of the same size operating from a smaller Member State is immediately confronted with the lack of cross-border loss compensation of some parts of its business operating in other Member States.

The effects of the above arrangements culminate in what many commentators consider violations of the basic right of free establishment (e.g. the economic problems resulting from the different treatment of permanent establishments and subsidiaries).

TRANSFER PRICING

The increasing importance of transfer pricing as an international company

tax problem and in the Internal Market

It practically goes without saying that transfer pricing issues need to be considered when analysing possible company tax obstacles to cross-border economic activity in the Internal Market. Market operators emphasise the ever increasing importance of the issue, citing unduly high compliance costs as well as clear instances of double taxation. Some have even raised the more fundamental question whether the allocation of tax revenues between Member States on the basis of the 'arm's length principle' and 'separate accounting', which lie at the heart of transfer pricing, is still the most appropriate way of "dividing the tax cake" within the EU. As a matter of fact, in the Internal Market, with its increasing market integration as well as the increasing importance of intangibles make it increasingly difficult in practice to divide profits on a traditional transaction basis.

The 1992 Ruding report identified transfer pricing as one of the most important areas for the future in international taxation and for the Internal Market and made some recommendations including, but not limited to, the ratification of the EU Arbitration Convention. However, the Ruding report did not analyse in any detail the problem of transfer pricing in the context of the Internal Market. As indicated in Part I of this study, the completion of the Internal Market and other international developments like the almost exponential growth of intra-group cross-border trade now call for a more thorough analysis. This section therefore takes a more comprehensive look at whether and how the tax arrangements concerning transfer pricing constitute an obstacle to the Internal Market.

Given the objective of this study, this section does not explicitly cover transfer pricing problems with third countries. However, transfer pricing is a global issue: double taxation, high compliance costs etc. are obstacles to cross border trade, both within the Internal Market and beyond the EU. It is nevertheless important to stress that Member States have a responsibility to address existing tax arrangements, including those with an impact on transfer pricing, which constitute an obstacle to the smooth functioning of the Internal Market. In tackling these obstacles care must, however, be taken that the

international consensus on transfer pricing is not put into question.

Basic concepts of transfer pricing and the Internal Market

The taxation of transfer pricing is one of the most complex issues of international taxation

• 58. 

  It may therefore be useful, as a basis for the following analysis, to explain

some basic concepts of transfer pricing and to consider its particular situation within the

Internal Market.

The technicalities of transfer pricing and the OECD Guidelines

The arm's length principle

It is generally recognised that affiliated companies conducting cross-border business for tax purposes must do this on market principles, i.e. act as if the business was being conducted between independent parties. The price charged for goods and services - the transfer price - therefore has to be in accordance with the so-called arm's length principle. The basis for the arm's length principle is the separate entity approach; i.e. each affiliated company in a group is for tax purposes treated as a separate entity and taxed individually on the basis that it conducts business with other group members at arm's length. For the Internal Market this means that a company has to provide separate accounting for the 15 Member States (or at least every Member State where it is active).

Article 959 of the OECD Model Convention60, and maintained and developed in the

1995 OECD Transfer Pricing Guidelines (Guidelines)61. The practical application of the

arm's length principle is complex and the Guidelines provide guidance for its application by tax administrations and taxpayers.

As indicated the application of the arm's length principle is generally based on a comparison of the conditions of transactions between affiliated parties (i.e. controlled transactions) with the conditions of transactions between independent parties. The latter transactions are generally referred to as comparables. Comparables can be either internal (i.e. transactions between the group company and third party) or external (i.e. transactions between two third parties). The Guidelines include extensive guidance on when a transaction between independent parties is sufficient comparable. Factors determining comparability include the characteristics of the property transferred or services provided, functions performed, risks assumed, contractual terms and economic circumstances and business strategies. The Guidelines explicitly recognise that comparability analysis is not an exact science, but requires an element of judgement.

Different methodologies can be applied to establish whether controlled transactions are in accordance with the arm's length principle. The methodologies are generally referred to as transfer pricing methods. The Guidelines explicitly mention five transfer pricing methods falling into two categories, which are the so-called traditional transaction methods and the so-called `other' or (transaction-based) profit methods.

Box 37:

Transfer pricing methods 62

The traditional transaction methods (hereinafter transaction methods) are the Comparable Uncontrolled Price (CUP), the Resale Price Method (RPM) and the Cost Plus Method (CP). CUP compares the price

for property or services transferred in a controlled transaction to the price agreed for property or services in comparable uncontrolled transactions. If applicable, CUP is likely to be the most direct and reliable method and is therefore preferred over all other methods. The RPM is based on the price at which a product purchased from an affiliated company is resold to an independent enterprise. The gross margin on a controlled transaction is compared with the gross margin of comparable uncontrolled transactions. The resale price to a third party is then reduced by the resale price margin and the remainder constitutes the arm's length price at which the product is deemed to have been purchased from the related company, i.e.

RPM compares gross margins. The CP starts with the gross costs incurred by the supplier of property or services in a controlled transaction. A cost plus mark is added to this cost in order to raise the price to the level at which the product would had been sold in an uncontrolled transaction. The mark up must be consistent with mark ups in comparable uncontrolled transactions. Consequently, the CP also compares

comparable gross margins on costs. This method is most relevant for production companies or service providers. The `other' or profit methods comprise the so-called transactional profit methods (profit methods) which examine the profits arising from controlled transactions. The Guidelines explicitly list two profit methods, the Profit Split Method (PSM) and the Transactional Net Margin Method (TNMM). ThePSM identifies the combined profit to be split between associated enterprises from a controlled transaction, and then splits this profit between the associated enterprises on an economically valid basis that approximates the division of profits that would have been anticipated and reflected in an agreement made at arm's length. It is important to note that the PSM differs from all other transfer pricing methods, including the TNMM, as according to the Guidelines it does not necessarily require the use of comparables. PSM is also especially relevant when one or several of the parties hold valuable intangibles.

The TNMM examines the net profit margin relative to an appropriate base (e.g. cost, sales, and assets) that a taxpayer realises from a controlled transaction. TNMM is therefore largely similar to the RPM andCP, the difference being that the first compares net margins, the latter gross margins. TNMM was a new development introduced in the Guidelines and has the advantage, compared to RPM and CP, that it does not require the same amount of detailed information concerning the cost base, i.e. direct costs and indirect costs. Different profit based methods, in addition to the five explicitly mentioned (CUP, RPM, CP, PSM & TNMM), although not specified in the Guidelines, may also be used provided they are consistent with the specified profit based (methods PSM and TNMM). The most important example of such methods is the so-called Comparable Profit Method (CPM) which originated in the U.S. CPM is to a large extent similar to the TNMM, the difference being that TNMM stresses profit per transaction (product line etc.) whereas CPM can be used on a more aggregated basis. It is sometimes argued that TNMM and CPM are in practice identical although, in fact, TNMM constitutes a limited part of CPM. CPM is often used by US companies

The Guidelines introduce the so-called arm's length range. This implies that often comparables will produce a range of figures, which are relatively reliable, and that tax authorities should not make adjustments provided the transfer prices are within the range. In this way it is recognised that the application of the arm's length principle is not an exact science. This arm's length `range' principle is also relevant for transaction methods but is generally more important in the application of profit methods.

According to the Guidelines "practical experience has shown that in the majority of cases, it is possible to apply traditional transaction methods". The Guidelines reject the use of global formula apportionment methods which work by allocating profits of a multinational enterprise on a consolidated basis among each group member according to a formula fixed in advance.

Documentation requirements

A key issue is the question of what kind of documentation a group company needs to prepare to demonstrate it has applied the arm's length principle. The Guidelines

maintain that "the taxpayer should not be expected to provide more documentation than the minimum necessary to permit tax administrations to audit transfer prices and to verify if the taxpayer has applied the arm's length principle." Generally the Guidelines aim at maintaining a balance between the right of tax administrations to obtain from tax payers as much information as possible to ascertain whether the price is or is not of an arm's length nature, and the compliance cost that any documentation rules imply for the taxpayer. The Guidelines recognise that the tax payer should make reasonable efforts, at the time transfer prices are set, to determine whether the arm's length principle is satisfied, and that tax authorities can expect or require tax payers to maintain documentation to support this. However, the amount and type of documentation required should be in proportion to the circumstances of each case. In this context the Guidelines introduce the important concept of the `prudent business manager'. This implies that the process of considering transfer prices should be carried out in accordance with the same prudent business management principles as would govern the process of evaluating any other business decision of similar complexity and importance.

• - 

  the risk of penalties and economic double taxation,
• - 

  the costs of temporarily having to finance the same tax burden twice and
• - 

  increased auditing by the tax authorities.

A general problem in this context for multinational enterprises is the one of uncertainty.

It is claimed by business that there is a risk of suddenly having a business structure which has perhaps been in place for a number of years undermined by the tax authorities who will - for transfer pricing reasons - no longer accept it from a tax point of view. Moreover, according to the business representatives in the panel it is not uncommon that a certain structure and/or transfer price (or the application of a specific method) might be acceptable to one Member State but not to another.

Transfer pricing thus represents an additional burden for a company in one Member State to set-up and/or conduct business with an affiliated company in another Member State, and instead favours domestic investments/transactions. Furthermore, as independent enterprises are not subject to transfer pricing regulations on cross border transactions they will be in a better position than multinational enterprises. Transfer pricing can also affect the location of cross border investments. All other things being equal, a company would be less likely to set-up a subsidiary or branch in another Member State with a stricter transfer pricing policy than in another Member State with more lenient rules.

Small and medium-sized enterprises can be particularly hit by these problems. Frequently they are not even familiar with the basic concepts of transfer pricing and do not have the appropriate resources and structures to deal with the problem when they, say, create a first subsidiary abroad.

The setting of intra-group transfer prices in accordance with the separate entity approach and the arm's length principle does not necessarily correspond to the prices set for business reasons (effectiveness, performance measurement etc.). This has always been the case. However, business representatives maintain that the very concept of the arm's length principle will in the future lose its underlying commercial rationale. This is because large companies, in view of their EU-wide corporate restructuring, adopt so- called Euro-pricing. As sketched out above in Part I, as a result of the price convergence expected under the single currency, one transfer price is used per harmonised (intermediate) product for the group in the whole of Europe, regardless of which production facility the goods are purchased from. Business argues that from a management point of view, this avoids intra-group disputes about price levels and permits an optimum efficiency in the structure.

Box 38:

Transfer pricing as management tool 64

Multinational enterprises are usually organised in relatively autonomous divisions which are responsible for the (non-strategic) commercial decision-making. This often includes responsibility of the division management for their own profits (profit centre structure). On the one hand, this implies that every division must be able to decide whether buying a certain (semi-finished) product with third parties or from another related division. On the other hand, it must be ensured that the objective of profitability at division level does not hamper the profit-maximising objectives of other divisions or the overall group. A fully developed transfer pricing system of a multinational enterprise would have to motivate divisional managers and employees, allow the headquarter to evaluate the performance of the divisional management and link its remuneration to it, be perceived as fair by both the selling and the purchasing unit and allow for an optimal allocation of resources within the group. It is often difficult to find a transfer price that meets all these objectives. For instance, if there is an unused production facility it would make sense, in an overall group perspective, to instruct the supplying unit to transfer its output to the down-stream affiliate at only marginal or variable cost. This means, however, that the supplier cannot recover its fixed cost and will incur a loss on this transaction. Thus, the objective of maximising the group profit conflicts with the objective of maximising the divisional profit. This kind of question has received extensive attention in the management literature. It shows that transfer prices that are set for commercial reasons are not necessarily identical to the arm's length price that is usually requested for taxation purposes.

Moreover, multinational enterprises will also have an incentive to apply non-arm's length prices in transactions with affiliated companies in those Member States who have taken a 'robust' line in the transfer pricing area. By providing affiliates based in these `robust' Member States with more than an arm's length remuneration, transfer pricing problems with those Member States' tax authorities can be avoided.

Profit shifting through transfer pricing

This introduces the issue of profit shifting. Transfer pricing can of course be used as a tax-planning tool not simply to shift profits into those Member States with a `robust' approach but also to shift profit from high tax to low tax jurisdictions by charging non- arm's length prices. It is, however, difficult to assess to what extent manipulation of transfer prices is systematically used as a profit-shifting instrument.

pricing tax planning and deliberate profit-shifting. Instead, when intra-group transactions do not meet the arm's length principle, this would mainly be caused by the complexity of the transfer pricing rules, the difficulties in finding comparables, or because no attempt to check whether intra-group prices are at arm's length has been made at all.

The introduction of new business structures - as part of a corporate reconstruction - where an affiliate of a multinational enterprise has its functions or its risks reduced, may, by some tax administrations, be considered to have been motivated mainly or solely for tax purposes. This might particularly be the case in reorganisations where functions and risks are transferred to one or several affiliates resident in tax havens or subject to a preferential tax scheme. Business claims such business reorganisations are done primarily for business reasons and not to save tax and that the tax authorities generally should accept them. In this context it is also important to note that the possibilities for such types of tax planning instruments will be reduced significantly by the current exercises in the EU and in the OECD aimed at eliminating harmful tax measures.

However, in contrast to the Ernst & Young survey comprehensive studies made by someUS researchers show that transfer pricing in a US context frequently is used to shift profit from high to low tax jurisdictions. The tightening up of the US transfer pricing rules in the early 1990s was based on studies showing that transfer pricing was used to a large extent to avoid taxation in the USA. There is also some evidence on profit shifting within the EU

• 65. 

  However, overall the available studies do not show uniform results on

this question. In short, one can therefore conclude that, while there is undoubtedly evidence for aggressive transfer pricing by companies, there are equally genuine concerns for companies which are making a bona fide attempt to comply with the complex and sometimes conflicting transfer pricing rules of different countries.

The role of tax administrations

Although it is not directly an obstacle to the Internal Market, the effect of transfer pricing on tax authorities should also be mentioned. Tax administrations also suffer from the high compliance costs of transfer pricing. Transfer pricing auditing differs from normal auditing by being complex and expensive, and by requiring highly skilled tax auditors. It takes up a lot of resources, which in many cases can only be taken away from other audit areas. Of course a Member State can reduce these "difficulties" - and some have done so - by introducing robust transfer pricing rules, notably in the form of tough (up-front) documentation rules backed up by penalties for non-compliance, which transfers the burden to business and to other tax administrations.

introduce documentation requirements etc., the alternative being that tax would instead be paid in those countries with stricter documentation requirements, stricter penalties etc. There is a risk that this will lead to a kind of race to the top between Member States, where business, the Internal Market and international trade in general can only be the loser.

The following box illustrates the increased interest that governments have taken in the issue in the last 5 years.

Box 39:

The development of transfer pricing documentation rules in national legislation and practice

19952000

USAUSAUK

AustraliaAustraliaFrance

MexicoDenmark

BrazilBelgium

CanadaPoland

S.-Korea

Another problem is the different application of transfer pricing rules in Member States. The Guidelines therefore do not eliminate all the differences between transfer pricing rules in the Member States.

Based on this analysis there is little doubt that transfer pricing currently constitutes an obstacle to exploiting the full benefits of the Internal Market as it represents a significant compliance burden which does not arise in the domestic context. The following section analyses the various technical features of this obstacle in more detail.

that the scope for finding transactions between independent parties is not large and is diminishing.

Another aspect is that multinational groups can conduct business in ways which independent companies cannot and this can give rise to problems in finding comparables. A standard example is the use of certain types of intangibles, e.g. know how, where companies are normally not willing to allow independent parties access to these intangibles which constitute business secrets. Other standard examples are headquarter services (e.g. administrative and technical services), which by definition are not relevant for independent companies, and cost sharing. Numerous other examples exist, for instance the use of contract manufacturers and companies that perform contract R&D who carry few risks and therefore can be remunerated on a cost plus basis. These are much more commonly used within multinational enterprises than outside by independents. When the whole business structure of multinational enterprises differs so fundamentally from that of independent parties it can reasonably be assumed that it is difficult to find comparables. In sum, the fact that intangibles are becoming increasingly important makes it more difficult to find comparables. Increased reliance on intangibles (both production and marketing intangibles) is caused by the new business structures, new technologies (Intranet etc.) and because products are becoming more complex.

All these structural constraints affect the search for comparables and the quality of (any) comparables which are identified. As mentioned above transaction methods are generally preferred to profit methods and they rely to a great extent on multinational enterprises having internal comparables. It is often the case that a multinational enterprise does not have transactions with independent parties, other than when the product etc. is sold to the ultimate end user, or that if it does, these transactions might not actually be comparable, for instance because the sales to independent parties only take place on marginal markets, because of low volume etc.

Pragmatic solutions for the lack of "comparables"

As a result multinational enterprises in practice often have to search for so-called external comparables, i.e. comparables between two independents parties available from commercial databases

the use of commercial databases is the costs of obtaining access to the databases and the costs of performing the searches.

In short, it is difficult for business to find comparables and that the comparables available generally are not the internal comparables at the transactional level which tax authorities generally prefer to see. Often only external profit comparables from commercial databases are available which then often give rise to disputes with and between tax authorities.

"Comparables" and transfer pricing methods

It is often maintained that TNMM is much more widely used than generally expected, for instance by multinational enterprises. In this context it is noteworthy that some tax administrations which tend to be Anglo-Saxon OECD members who are quite experienced in the transfer pricing area - hold that business, via commercial databases, has easy access to some of the information necessary to apply the TNMM, but that they often do not apply the method in accordance with the Guidelines opting for a more simplistic approach. Compared to transaction methods, TNMM therefore can be an easier (cheaper) solution. The method is sometimes chosen even when transactional methods (subject to necessary adjustments) actually could be used. Another reason mentioned for using TNMM is its similarity with CPM, which is often used in the USA.

There are, however, good reasons to believe that the particular situation in the EU Member States is still different from the USA in this respect and that profit methods are not used. The Ernst & Young survey indicates that in practice transaction methods - especially CUP and CP - are in fact the most commonly used methods for all transaction types in clear preference to profit-based methods. It therefore seems that business in general has, up until now, managed to overcome the theoretical problems of finding and applying internal (or external) transactional comparables for setting transfer prices. It should nevertheless be noted that the survey does not reveal the quality of these transactional comparables. For instance it cannot be ruled out that in applying the CP, business often apply a "standard mark-up" of for instance 5-15% without any support from independent transactions or companies. More importantly, the tax problems relating to transfer pricing can be expected to grow in the future. The debate on the applicable method and the conditions for its use will become increasingly important in the EU.

guidelines or statements of practice stating their particular position in areas where the Guidelines are not clear. One example of different treatment is intra-group services and the question of whether the service provider shall earn a profit on the services or just have his expenses covered. In practice a cost plus based approach is often used for intra- group services and especially for management services by the headquarter. The question therefore often arises whether the service provider (often the headquarters) must charge a mark-up on services provided to foreign affiliates (often subsidiaries). The OECD Guidelines deal with services in Chapter VII.

In paragraph 7.2. of the Guidelines it is mentioned that most multinational groups of enterprises arrange for a wide scope of services to be available for its members, in particular administrative, technical and financial services, and that such services also may include management, co-ordination and control functions for the whole group. The Guidelines address the issue of profit versus cost in paragraphs 7.33, 7.34 and 7.37. As a general rule the service provider must make a profit, as an independent party will not provide services at cost (7.33). However, there are exceptions (7.33 and 7.34), e.g. where the value of the service provider does not exceed the costs. For instance for headquarter services provided to the whole group including the headquarters it can often be asked whether the value of the services exceeds the cost of providing the service, or whether the "service" is in fact simply a sharing of costs.

Some Member States take the view that the service provider must make a profit, as an independent party will not provide services at cost. This is also the case, where the service provider itself also uses the services (which will often be the case for headquarters). Other Member States take the position that if the service provider itself use the services function, then the arrangement must be considered a cost sharing one to which no mark-up can be added. The argument is that all parties including the service provider will benefit from the services by reduced costs. The first approach does not seem any more correct than the second, and vice versa. Sometimes there are ways of getting around this problem, for instance by setting up a special entity, but these will lead to additional costs and unnecessarily complicated business structures.

There are some other important issues where the Guidelines are not clear and/or where they can be difficult to apply in practice. These concern, for instance intra-group services in general, cost sharing and cost contribution agreements, losses, the detailed application of profit methods and market penetration issues. In the context of this study, the question arises whether EU Member States should develop a common stance on these issues.

As illustrated above, documentation requirements overall have increased within the EU in the sense that some Member States either by legislation or by circular letters have introduced whole new documentation rules or tightened existing requirements. This trend will probably continue. The majority of the Member States have not (yet) taken such initiatives. However, this does not mean that documentation requirements have not increased in these Member States as such an increase also can take place via the audit

process.

Concrete estimates of the resulting compliance costs

According to information given by business representatives the tax compliance costs of transfer pricing are quite material. This cost results from the obligation for enterprises to determine what prices could be regarded as arm's length including finding comparables, assembling the related documentation and defending these prices in audits, etc. According to some estimates, medium sized multinational enterprises spend approximately 1 to 2 million each year on complying with transfer pricing rules. Large multinational enterprises incur compliance costs related to transfer pricing of approximately 4 up to 5.5 million a year. These figures do not include the costs and risks of double taxation due to transfer pricing disputes. Other estimates by representatives of the manufacturing industries indicate even higher figures.

Quantitative information on transfer pricing

The views of the business community and of tax administrations on the crucial tax problems on transfer pricing appear to differ significantly. While business representatives maintain that double taxation cases and subsequent disputes are almost daily practice, tax administrations claim that real double taxation cases are relatively rare and always solved. According to tax practitioners this different perception can partly be explained by the fact that in view of the insufficient redress possibilities businesses often "give up" and accept double taxation which would be too costly and burdensome to remove, if there is prospect of achieving this at all. Moreover, the difference in perception may in part also stem from the fact that transfer pricing rules must be applied on a self-assessment basis. Even if only a small proportion of companies are audited for transfer pricing, every company must apply transfer pricing rules in computing profits and maintain the necessary documentation in case of audit. Thus the compliance burdens do not depend upon either an audit or (even if there is an audit) a dispute over transfer pricing policy.

The Ernst & Young transfer pricing survey

Ernst & Young published their latest survey on transfer pricing at the end of 199967. The

1999 survey includes 19 countries including 9 Member States: Argentina, Australia,

Brazil, Canada, Denmark, Finland, France, Germany, Italy , Japan, Korea, Mexico,

Netherlands , Norway, Spain, Sweden , Switzerland, the UK and the United States. The survey was conducted by Consensus Research International, a London-based Research Agency, and included interviews with the person responsible for international tax matters in leading multinational organisations. 582 parent company interviews were held, and telephone interviews were conducted with 124 subsidiaries of foreign-owned parents. This survey, which was commissioned by a private tax consultancy firm, represents the only available large-scale work in this area and its findings give valuable insights to business perception of the subject.

The survey reveals that business considers transfer pricing to be the most important international tax issue for the future. One of the reasons is that it sees a clear connection between transfer pricing and double taxation. Business reports that in 42% of cases of adjustment this resulted in double taxation . This is primarily because firms do not generally refer cases to mutual agreement procedure, as they consider the procedures take too long and take up too many resources in relation to the exposure. Advance Price Agreements are increasingly considered as a possible solution to avoid double taxation.

Business, as in the 1997 survey, continues to identify maximisation of operating performance, not optimising tax arrangements (i.e. tax planning), as the most important factor in shaping transfer pricing policies; consequently the majority of companies do not consider transfer pricing to be a tax planning instrument, but instead largely a compliance exercise.

The survey also revealed that the number of transfer pricing audits has increased overall. Nearly two-thirds of the multinational enterprises headquartered in the 12 countries included in the 1997 survey, report in the 1999 survey that their transfer prices have been subject to tax authority examination (40% of the multinational enterprises in the 7 new countries), and 75% of multinational enterprises expect to face a transfer pricing audit somewhere in their organisation within the next two years.

still rely on historical practice or cost-methods (i.e. without profit element) as the price

setting mechanism.

Commission Services questionnaire on dispute settlement mechanisms in the area

of transfer pricing

In June 2000 the Commission Services circulated a transfer pricing questionnaire to the tax administrations of the Member States

• 68. 

  14 out of 15 Member States responded to

the questionnaire. In part I of the questionnaire , Member States were asked for information on their experience with the mutual agreement procedure (within the context of double tax agreements) from 1995-1999. The questions included: the number of the mutual agreement procedures their tax administration had been involved in, the number of these relating to other Member States, how many requests from tax payers they had accepted, whether the mutual agreement procedures were initiated by the Member State itself or by the treaty partner, to what extent it was possible to reach an agreement, the length of the procedures (average, longest and shortest period), why the mutual agreement procedures were not successful and in which areas tax administrations tend to have the greatest difficulties in reaching agreement. Member States were asked to complete a table.

Similar questions were asked in part II with respect to the EU Arbitration Convention plus questions referring to the second phase (panel) in the Convention. Part III

concerned administrative issues about co-operation with other tax administrations, the use of penalties and the possibilities of suspension of tax collection in cases of income adjustment.

The answers to part I and II are summarised in 3 tables at the end of this subsection. The tables reveal that 127 intra-EU transfer pricing adjustments cases have been referred to the mutual agreement procedures or to the EU Arbitration Convention from 1995-1999 (total number of cases, i.e. including third countries, is 413 cases). This gives an average of approx. 25 adjustments per year within the EU (83 including third countries), or approximately two adjustments per Member State. This does not seem to be an alarming number of adjustments. Furthermore, although the number of adjustments from 1995- 1996 to 1997-1999 have gone up by more that 100%, this development seems to have stopped in 1998.

respect to the length of the procedures, the calculated average duration of the mutual agreement procedure intra-EU is 20 months.

In the context of the EU Arbitration Convention hardly any cases are being rejected, and no Member States reports having used the so-called penalty-clause in Article 8. The questionnaire does not reveal the average duration within the EU Arbitration Convention, as it was expected that the vast majority of cases would be solved within 2- 3 years. However this is not the case. Of the 1995-cases only 67% have been solved (1996; 48%, 1997; 48%), which must be considered to be quite disappointing. Furthermore, no case has so far despite the fact that a number of cases are well over 2 years old - reached the second phase in the sense that no advisory commission has been set up to date (one Member State reported 3 cases where the second phase has been initiated). The material does not reveal to what extent the lengthy procedures are because businesses have appealed cases to the national courts etc. Despite that, it seems fair to conclude that the objective of maximum 3 years of duration has not been achieved, and that for some reason Member States do not initiate and progress the second phase after the end of the 2 year negotiation period.

Table 47

Mutual agreement procedures (MAP) 69

PeriodTotal numberProcedure initiatedSuccessful completionClosed cases

Of new casesYesNoYesNo

199522 (74)22 (73)0 (1)17 (60)5 (13)18 (63)

199615 (75)15 (75)0 (0)12 (60)3 (15)12 (67)

199726 (76)25 (75)1 (1)11 (40)14 (35)13 (43)

199841 (111)40 (110)1 (1)12 (40)28 (70)12 (40)

199931 (85)30 (84)1 (1)10 (38)20 (45)14 (42)

TOTAL I135 (421)132 (417)3 (4)62 (238)70(178)69 (255)

TOTAL II 6770 (35471)64 (350)3 (2)31 (119)35 (89)35 (228)

PeriodSuccess-rate in %Duration (in month)

Of new casesOf initiated casesOf closed casesAverageShortestLongest

199577 (82)%77 (82) %94 (95) %

199680 (80) %80 (80) %100 (90) %

199742 (53) %44 (53) %85 (93) %

199829 (36) %30 (36) %100 (100)%

199932 (49) %33 (45) %71 (90)%

Total I/II46%47% (57%)90%18 (21)1 (1)60 (72)

Table 48

The EU Arbitration Convention

Procedure initiatedNumber of casesEnd of the first phase

PeriodTotal numberStill in the first phaseSolvedFailed

YesNo

of new cases

72

199518 (19)18 612

199625 (26)2231012

199740 (44)39118192

199836 (43)36 24111

199947 (54)47 389

Total I166(186) 162496643

Total II83 (93)8144832-

Second phaseSuccess-rate in %

PeriodInitiatedIn progressSolvedOf requested casesOf initiated cases

1995 67%67%

1996 48%55%

19972 48%48%

19981 31%31%

1999 19%19%

Total I373 30%40%

Table 49

Total Mutual Agreement Procedures (including EU Arbitration Convention) of

Member States

PeriodTotal number of new cases

199530 (82)

199631 (91)

199752 (104)

199871 (141)

199968 (122)

Total I25475 (54076)

Total II12777 (41378)

Part I and II of the questionnaire also gave some indications about where Member States consider the application of transfer pricing rules to cause problems. These include: disagreement of use of comparables, transactions involving intangibles, the fact that the mutual agreements may concern periods in the past, for which enterprises didn't have as detailed documentation as today, use of profit methods (especially in the case where the whole group is loss making) and lack of adequate information. One Member State requested a co-ordinated approach with respect to conducting functional analysis.

sense in transfer pricing cases. Most other Member States reserve the use of these types of penalties for cases which involve an element of tax evasion or gross negligence, and report that penalties are in practice not levied in transfer pricing cases.

With respect to the possibility of suspension of tax liabilities, it can with care - be concluded from the answers that some Member States in principle provide for suspending of enforcement if the adjustment is appealed; however the practical application of these rules are uncertain. In cases of requests for mutual agreement suspension facilities are only rarely available.

Double taxation and dispute settlement mechanisms

The need to avoid or at least swiftly remove double taxation in transfer pricing

Double taxation is a serious obstacle to the Internal Market. Double taxation created by transfer pricing rules (even if legal) should be avoided in principle. If this is not possible in practice, then it is imperative to have appropriate dispute settlement mechanisms that relieve double taxation as quickly and efficiently and in as many cases as possible, and with the lowest possible costs for business and tax administrations.

When a tax administration makes an (upward) income adjustment (primary adjustment) the multinational enterprise is immediately subject to double taxation. This double taxation can be relieved if either the tax authorities of the other state accept a (downward) income adjustment (corresponding adjustment), or if the tax authorities making the primary adjustment subsequently reverse the adjustments.

There is the further problem that some tax legislation in order to make the actual allocation of profits consistent with the primary adjustment might assert a constructive transaction (as dividends, loans or equity contributions). The secondary transaction might lead to source taxation or influence the availability of loss relief claims. This secondary adjustment might not be accepted by the other tax jurisdiction involved.

It is fundamentally important that the double taxation is relieved, but other important issues in this context include:

• 2. 

  Mutual Agreement Procedures (MAP) which represent a special dispute settlement mechanism vested in the bilateral double tax treaties.
• 3. 

  The EU Arbitration Convention (Convention)79 which is a special EU dispute

settlement mechanism. The convention only covers transfer pricing, i.e. whether or not intra-group trade is at arm's length.

• 4. 

  Advance Pricing Agreements (APA). These agreements form a means for the taxpayer to request a 'binding transfer pricing ruling' from the tax administration(s)

on the treatment of a future transaction involving the setting of transfer prices.

The suitability of the various approaches within the EU is very different.

Transfer pricing cases are not suitable for litigation in national courts. This is because transfer pricing is not a juridical discipline as it mainly focuses on economics ("fractional analysis") and because cases tend to be very fact specific with substantial amounts of background material. Litigation of transfer pricing cases in national courts therefore tends to be very lengthy. Furthermore, litigation is a one-sided approach that only deals with the problem from the perspective of one of the jurisdictions. Hence, litigation does not guarantee elimination of double taxation. On the contrary, if double taxation is not (fully) relieved in national courts the possibility of avoiding double taxation in a subsequent mutual agreement procedure will be reduced. Nevertheless, there are significant number of transfer pricing cases pending at national courts.

Moreover, litigation of transfer pricing cases is often very expensive for both business and tax administrations. Some indication of the costs of litigating transfer pricing cases is given in the context of OECD meetings. In any event, all available figures suggest that the costs for business (or for EU tax administrations) of preparing and litigating transfer pricing cases is substantial. The general dispute settlement mechanism is therefore unlikely to be litigation.

Advance Price Agreements differ from the other dispute settlement mechanisms as they aim to avoid any dispute arising at all. APAs are generally not yet very developed in theEU. They are considered in more detail below as a possible way forward.

the case80. Timely decisions and guarantee for relief of double taxation are the two most

important objectives, and the EU Arbitration Convention should therefore be the prevailing dispute settlement mechanism within the EU.

Business generally recognises the implementation of the EU Arbitration Convention as a major improvement compared to the MAP; mainly because of the arbitration phase for cases where the competent authorities cannot agree upon a solution. However, the EU Arbitration Convention suffers from shortcomings as discussed below

81.

Box 40:

The Mutual Agreement Procedure (MAP) According to Article 25 of the OECD Model Tax Convention

Introduction Legal basisMost double tax agreements include a provision equal or similar to Article 25 of theOECD Model Tax Convention providing for a mutual agreement procedure. It should be noted that the double tax treaties between Member States can deviate from Article 25 of the OECD model tax convention.

Coverage of transfer pricingMember States generally recognise that the MAP covers transfer pricing cases.

Right for taxpayer to request for MAPThe taxpayer generally has the right to initiate the procedure.

When can the taxpayer initiate the MAP?Taxpayer can initiate before the primary adjustment is made; it is sufficient that an adjustment is likely to take place. The taxpayer must request the procedure not later than 3 years after this time.

Who does the taxpayer contact?The relevant authority is the so-called "competent authority", which is appointed in the double tax treaty. The competent authority is typically a section of the central tax administration. The competent authority, does not normally, if ever, perform auditing activity.

Description of the MAP procedureThe MAP in principle comprises two phases/stages. In the initial phase the competent authority contacted by the taxpayer is required to reconsider the case. The second stage consists of the negotiation phase between the two competent authorities.

Only obligation to negotiate not to reach agreement!The MAP only obliges competent authorities to negotiate, not to reach a solution. Also, the MAP does not include any time limits within which agreement should be reached.

Does internal time limit rules exclude implementation of an agreement?No, the MAP provides for any agreement under MAP to be implemented irrespective of national time limit rules. It should be noted however, that five Member States had made reservations to this rule.

Box 41:

The EU Arbitration Convention comparison with the Mutual Agreement Provisions

in Double Tax Treaties

Introduction Legal basisThe EU Arbitration Convention is a convention between Member States. The legal basis of the EU Arbitration Conventions is Article 293 (ex 220) in the EC Treaty. The Convention is not yet applicable in Austria.

Generally the same instruments as in MAP but an arbitration procedure on topThe EU Arbitration Convention to a very large extent includes the same instruments as the mutual agreement provisions in the double tax agreements. However most importantly if competent authorities cannot agree to solve double taxation then the case is referred to a so-called advisory panel (arbitration panel). Therefore, relief of double taxation is in general guaranteed.

The remaining sections list the special features of the EU Arbitration Convention (compared the normal mutual agreement procedure).

Four phases/stagesInitial phase - The competent authority contacted reconsiders the case 1

st phase - Negotiation between the competent authorities competent authorities

have 2 years to reach agreement. If they do not succeed they must set up an advisory panel (arbitration panel) 2

nd phase - Advisory panel considers the case the panel must make a decision

within 6 months 3

rd phase - When the advisory panel has made its decision competent authorities

have another 6 months to reach an agreement. If they fail to do so the decision from the advisory panel becomes final and competent authorities must comply.

When does the 2-year period of the 1Member States hold different opinions on the precise starting point.

st phase

start?If an adjustment is appealed to national courts/tribunals, the two-year period of the 1

st phase does not start running until the date on which the judgement of the final

court of appeal was given.

Penalty clauseMember States are not obliged to initiate the mutual agreement procedure or to set up the advisory commission if one of the enterprises concerned is liable to a serious penalty. If proceedings on the penalty issue are pending, the competent authorities may stay the proceedings until the penalty issue has been concluded. Member States have, in an annex to the Convention, individually defined what they consider to be liable to a serious penalty.

Shortcomings of the arbitration convention

Suspension of tax deficiencies

Member States normally immediately enforce an income adjustment, i.e. collect any tax underpayment. The EU Arbitration Convention does not include rules on suspension of the collection of tax and neither does the MAP. Sometimes multinational enterprises can avoid having temporarily to finance the "same" tax burden twice by appealing to domestic courts or tribunals. However, this generates other problems. According to Article 7 (2), second sentence, when a case is referred to national courts/tribunals, the two year-period of the first phase does not start running until the date on which the judgement of the final court of appeal was given. Furthermore, as described below, in some cases the second phase of the EU Arbitration Convention cannot be initiated unless the enterprise gives up its possibility of a court appeal. In short, multinational enterprises are trapped in a dilemma of either having to give up their right to have the tax collection suspended or diminishing the possibility of having double taxation abolished.

Transfer pricing cases differ from other tax disputes in the sense that the question is not whether the tax payer (i.e. the multinational enterprises) should pay tax or not, but rather whether the tax should be paid in state A or state B. The OECD transfer pricing guidelines recognises the problem for business of having to (temporarily) pay the same tax twice. The Guidelines recommend that countries adopt rules allowing for the suspension of tax liabilities or underpayments:

"A first problem is that the assessed deficiency may be collected before a corresponding adjustment proceeding is completed, because of a lack of domestic procedures allowing the collection to be suspended. This may cause the multinational enterprises group to pay the same tax twice until the issues can be resolved. Countries that do not have procedures to suspend collection during a mutual agreement procedure are encouraged to adopt them where permitted by domestic law, although subject to the right to seek security as protection against possible default by the tax payer"

82

The penalty clause

According to the so-called penalty clause in Article 8 of the Convention, Member States are not obliged to initiate the mutual agreement procedure or to set up the advisory commission where legal or administrative proceedings have resulted in a final ruling that the actions giving rise to the adjustment of profits renders one of the enterprises concerned liable to a serious penalty. If judicial or administrative proceedings are initiated with this in view (serious penalty) and are being conducted simultaneously with any of the proceedings of the Convention, the competent authorities may stop the latter proceedings until the judicial or administrative proceedings have been concluded.

According to the above-mentioned Commission Services transfer pricing questionnaire, Member States have not (yet) used this `serious penalty' clause to refuse the use of theEU Arbitration Convention. This, however, does not imply that one can automatically conclude that the penalty clause does not restrict the use of the Convention. The question of whether an enterprise is subject to penalty or not will have been decided upon in parallel to the question of the income adjustment. Enterprises having their taxable income adjusted and being subject to serious penalties cannot request application of the Convention, and the competent authority therefore does not need to specifically deny the use thereof

84.

Member States have in individual declarations annexed to the Convention listed what constitutes a serious penalty. The definitions of serious penalties differ among Member States. For instance some Member States only include cases of intent whereas others also include negligence. Another approach is to include cases where the enterprise has been subject to a fine exceeding a certain threshold.

The penalty clause is problematic because the issue of penalties should not be linked to the issue of the possibility of having double taxation abolished, as this has the effect that double taxation becomes a penalty. Member States are, subject to restrictions of the EC Treaty, free to penalise non-compliance with transfer pricing regulations. However, this should take place openly and transparently and not be hidden or disguised as double taxation. Furthermore, it is problematic that the definitions and practices of the penalty clause are inconsistent among the Member States, as this leads to unequal treatment among EU enterprises.

starts when the tax authorities receive a request from the taxpayer. This is also the position of two other Member States which, however, express the view that a request cannot be made until the tax authorities have actually made the adjustment, as no double taxation will occur until this point. One Member State takes the position that the two- year period does not start until all necessary information has been provided to the tax authorities. The answers to the questionnaire thus confirm the differing views, and a substantial number of Member States further respond that they would like to have clarification on this point.

Interpretation issues

Apart from the question of when the two-year period of the first phase starts, there are other examples where the EU Arbitration Convention is not clear. For instance one Member State would like a clarification of whether thin capitalisation rules are covered, and another Member State would like to have a set of guidelines on the application of the panel phase, including the establishment of the advisory commission.

Furthermore, according to Article 3 (2) of the Convention any term not defined in the Convention shall, unless the context requires otherwise, have the meaning, which it has under the double taxation convention between the Member States concerned. Examples of terms not defined include "enterprise", "permanent establishment" and when companies are "associated". The Convention as it stands does not therefore guarantee relief of double taxation if Member States apply a different interpretation of these definitions. The following describes the concrete problems that can arise.

The treaty network between Member States is not complete. In a potential case where there is no double tax treaty between the Member States, it is therefore uncertain whether these terms will be interpreted in line with the OECD model tax convention or according to domestic legislation. Furthermore, key definitions in the double tax treaties are not always defined in the treaty itself, but refer back to the domestic legislation of each Member State

• 85. 

  The term "enterprise" and the question of when companies are

"associated" might therefore be defined according to each Member States internal legislation. This lack of definition of "associated" might be problematic as Member States apply different definitions in their domestic legislation. Some Member States require a fixed threshold of the direct and indirect holding of share capital and/or voting rights; the "normal" threshold is 25%, but in some Member States it is higher (e.g. 51%). Other Member States take into account the facts and circumstances of each case, and apply a kind of de facto control. If for instance a Member State (applying a threshold of 25% according to domestic rules) makes an adjustment in a case where a parent company holds 45% of the shares in the subsidiary, whereas the other Member State in its domestic rules applies a threshold of 51%, then there would be a risk that this second State would not consider the companies to be associated and would thus consider the Convention to be inapplicable.

Lack of guidance on the panel phase

Only one Member State86 has (three) cases that have proceeded to the second Panel

phase. In its answer to the Commission Services transfer pricing questionnaire this Member State suggested that the arbitration phase should be explained in more detail. Procedures for setting up the advisory commission, in particular the appointing of the chair

87 could be included in a code of best practice. The EU Arbitration Convention

includes some procedure rules, e.g. information, business rights to appear or be represented before the advisory commission, costs etc. It is also stated that the advisory committee must deliver their opinion "within 6 months from the date on which the matter was referred to it". However, there are numerous other unresolved issues, some of which are outlined below.

One is the important question of when precisely the 6-month period starts running. The most obvious starting point would be the cut-off date of the two-year period of the first phase, leaving it up to the involved Member States to get the second phase process started quickly. However, it could also be argued that a case cannot be referred to an advisory commission until this has been (finally) established.

In that context it should also be noted that the Convention does not include rules on how the advisory commission organises its work. For instance who should call for meetings, what notice periods are required etc. The deadline of 6 months is very tight (but there are no consequences linked to non-compliance), and it therefore seems to be important to establish rules which would improve the likelihood of meeting the deadline.

The advisory committee is not a fixed one; a new committee is set up for each case. As different committees are not required to take decisions taken by other committees into account, and as publication of decisions by the advisory committees are not mandatory, there is a risk of different treatments. This could also mean that the possibility of establishing a common `jurisprudence' and series of precedents in the transfer pricing

area is missed.

Redress for the taxpayers

As mentioned above, according to Article 7(3) of the Convention, Member States whose internal laws do not permit the competent authority to derogate from decisions from their judicial bodies are not obliged to set up a panel, unless the enterprise of that state gives up its possibility to court appeal. The UK and France in Declarations on Article 7(3) positively declared that they will apply this provision. Denmark and apparently Belgium also use the provision

88.

Ratification problems

Unlike an instrument of Community law, the Arbitration Convention needs to be ratified by the Parliaments of all Member States. As mentioned above, the Ruding report already referred to this problem and urged Member States to accelerate the sometimes lengthy ratification procedure. Interestingly enough, currently both the extension of the Convention to Austria, Finland and Sweden

89 and the prolongation of the convention

beyond its original expiry date 200090 are again still not ratified in all Member States.

Therefore, transfer pricing dispute cases that have arisen since 2000 can currently not be dealt with under the Convention. This will only be possible retroactively when the prolongation has been ratified in all Member States. Evidently, this situation tends to increase the compliance costs and general problems of uncertainty by business operators. It should be noted, however, that the prolongation protocol contains a clause which in future provides for the automatic prolongation of the Convention when there is no timely objection raised by a Member State.

Conclusion

Transfer pricing is not only a major issue in the international tax arena but also a specific and important taxation problem within the Internal Market. The use of the arm's length principle is becoming increasingly difficult to apply, and transfer pricing rules and practices among Member States differ significantly. One common feature among tax administrations, however, is the increased focus on the issue, notably through increased documentation requirements and audit efforts.

The effects for business are double taxation and, most importantly, high compliance costs combined with penalty rules. Although, the EU Arbitration Convention constitutes a major accomplishment compared to the traditional mutual agreement procedures within the bilateral double tax treaties, it contains numerous technical difficulties and features certain provisions that are dissuasive to companies. In sum, many aspects of the tax treatment of transfer pricing constitute a complex obstacle for the Internal Market which hampers efficiency, effectiveness, transparency and simplicity.

DOUBLE TAXATION CONVENTIONS

The requirement to avoid double taxation in the Internal Market

Article 293 of the Treaty requires Member States ", so far as necessary, to enter into negotiations with each other with a view to securing for the benefit of their nationals [...] the abolition of double taxation within the Community". The purpose of this provision is to ensure that cross-frontier activities are not at a disadvantage compared with national activities. Generally, neither discrimination nor double-taxation resulting from the transnational character of an operation can be tolerated in the Internal Market. This marks an important difference in comparison to parties that are not linked in an integrated market and conclude a double taxation treaty. Double taxation treaties are designed to address double-taxation problems but in the case of divergent interpretations and similar problems no "higher" treaty forces them to find a solution. EU Member States must also however have regard to the Internal Market requirements concerning non-discrimination and the four fundamental freedoms, enshrined in the EC Treaty. The following box gives an overview about the relevant Treaty articles in this respect.

The existing network of bilateral tax treaties between Member States goes some way towards meeting these objectives. However, these existing tax treaties are far from sufficient to meet the requirements of the Internal Market. The following section analyses the tax obstacles and double taxation cases which remain intrinsically unsolved. Given the complexity and variety of the issues involved, it is not possible to present a thorough analysis of the technical details. However, the essential nature of the different obstacles to cross-border economic activity in the Internal Market that shortcomings in double taxation treaties within the EU create or fail to remove are identified.

Box 42:

Articles of the EC Treaty which impose obligations concerning non-discrimination

and the fundamental freedoms of the Internal Market

• - 

  Article 10 requires Member States to take all appropriate measures, whether general or particular, to ensure fulfilment of the obligations arising out of the Treaty or resulting from action taken by the institutions of the Community
• - 

  Article 12 of the Treaty prohibits any discrimination on grounds of nationality.
• - 

  Article 39 guarantees freedom of movement for workers within the Community, including the abolition of any discrimination based on nationality.
• - 

  Article 43 prohibits restrictions on the freedom of establishment of nationals of a Member State in the territory of another Member State.
• - 

  Article 48 requires that companies formed in accordance with the law of a Member State and having their registered office, central administration or principal place of business within the Community are to be treated in the same way as natural persons who are nationals of Member States.
• - 

  Article 49 prohibits restrictions on freedom to provide services within the Community.
• - 

  Article 56 prohibits restrictions on the movement of capital between Member States and between Member States and third countries, subject to certain caveats contained in Article 58.
• - 

  Article 94 requires the Council to issue directives for the approximation of such laws, regulations or administrative provisions as directly affect the establishment or functioning of the Internal Market
• - 

  Article 211 requires the Commission to take steps, including formulating regulations and delivering opinions, in order to ensure the proper functioning and development of the common market.
• - 

  Article 294 requires Member States to accord to nationals of other Member States the same treatment as they accord to their own nationals as regards participation in the capital of companies or firms.

The incomplete treaty network within the EU and its insufficient scope

The network of bilateral tax treaties on income and capital between Member States of the European Union is still not complete. There are at present about 97 bilateral tax treaties on income and capital in force between Member States one of which is a multilateral convention involving three Member States. The total possible would be 105.

agreements on inheritance and gift taxes is an impediment to the free movement of persons within the Community, especially owners of enterprises and companies, particularly of small and medium-sized enterprises. The members of the Panel that assisted the Commission services with this part of the study also highlighted this issue, pointing out that cross-border inheritance tax problems are increasing with the growing numbers of company transfers. As explained in more detail below, this problem primarily concerns small and medium-sized enterprises.

Generally, the absence of certain tax treaties can mean that substantial differences between the taxation in an EU Member State of taxpayers resident in other EU Member States with whom a tax treaty has been concluded compared with taxpayers of a thirdEU Member State with whom no treaty has been concluded yet. This can lead to instances of discrimination.

Double taxation cases unresolved by double taxation treaties

Double-taxation treaties and Internal Market requirements

Even where tax treaties are in place, there are a number of areas where they are inadequate to meet the requirements of the EC Treaty, in particular concerning respect for the four freedoms and the elimination of double taxation. It is self-evident that the fact that some of the treaties are quite old tends to exacerbate the problem. Of the 97 tax treaties on income and capital in force between Member States 36 are more than twenty years old. At least some of the old treaties do not reflect the current tax law of the two countries.

As already mentioned above as regards transfer pricing disputes, the Mutual Agreement Procedure in tax treaties does not oblige the two Contracting Member States to eliminate double taxation. While this procedure, which is contained in Article 25 of theOECD Model Convention and is included in most tax treaties, must be initiated in all cases of double taxation, it does not require the administrations concerned to reach an agreement. In practice, therefore, this instrument is incapable of resolving all cases of double taxation. This is incompatible with the Internal Market and effectively creates an obstacle to cross-border economic activity therein.

of the conventions and the shaping of anti-abuse provisions such as limitation of benefits clauses and the provisions safeguarding the application of thin capitalisation and controlled foreign company rules.

91

Hence, it appears fair to say that the equal treatment principle enshrined in the EC Treaty will make it increasingly difficult to justify inequalities of treatment between resident and non-resident individuals or companies in a similar situation under one identical tax treaty. This should not be misread as meaning that any difference in treatment of EU residents under double taxation treaties is automatically violating basic Treaty rights. In many situations, however, this question may be legitimally raised.

Triangular situations and third countries

More specifically, the equal treatment principle can influence the extent to which Member States' tax treaties can differ not only from each other but also and notably the permissible extent of differences between treaties of Member States with third countries. In this context it is noteworthy that already the Ruding Report suggested, in an annex, that if capital-exporting Member States grant fictitious tax credits (tax sparing) on income subjected to a special tax incentive in one other Member State they should grant it to all Member States in approximately the same economic situation.

Moreover, bilateral tax treaties are not normally capable of addressing triangular situations which can cause double taxation to go unrelieved. An example would be where a corporation is incorporated in one Member State, managed and controlled in a second and derives income from a third. The difficulty arises because two tax treaties are involved - that between the first State and the third State and that between the second State and the third State. The third State may withhold tax at source on the basis of one of those two tax treaties which will not be credited by one of the other States because it applies the other tax treaty which does not provide for a withholding tax at source.

According to businesses representatives many of these complex (legal or practical) problems ultimately boil down to a very fundamental and simple question. Say, Member State A has, due to a strong negotiating position, reached an advantageous clause in its bilateral tax treaty with a given third state. Member State B, on the other hand, had to accept a less advantageous solution on this matter in its bilateral treaty with that third state. Is it really in line with Internal Market requirements that companies resident in A can benefit from this advantage whereas companies that are resident in B cannot? What is the situation of permanent establishments that B-companies might have in A? Despite their underlying legal complexity the basic relevance of these fundamental questions can hardly be denied as, after all, EU Member States adopt a single common trade policy on goods (and, according to the Nice-Treaty, also some services) towards the rest of the world

impacts on business behaviour and investment decisions, thus provoking economically sub-optimal and welfare-reducing decisions. The existance of typical "treaty shopping routes" for EU companies in designing their business relations with the USA is revealing in this respect.

Box 43:

The "Saint-Gobain" triangular case - far-reaching impact on tax treaties

The "Saint-Gobain case" (case C-307/97) which concerned the justification for distinguishing for tax

purposes between permanent establishments and subsidiaries is an example of an issue which was resolved by the European Court of Justice but where treaty practice still seems to be inconsistent.

Compagnie Saint-Gobain is the German branch of Compagnie de Saint-Gobain SA which is established, managed and controlled in France. German domestic law, and German tax treaties with the United States and Switzerland, allowed certain tax advantages to German resident companies receiving dividends from abroad which were not available to a permanent establishment in Germany of a corporate enterprise registered in another Member State. The Court noted that the position of companies not resident in Germany but with permanent establishments there, and companies resident in Germany was objectively comparable. Consequently, refusal to grant the tax advantages in question to the permanent establishments meant that the latter were being treated differently from resident companies and amounted to restricting the freedom of companies to choose the form of their secondary establishments.

Most commentators and scholars believe that this judgement will have a far-reaching impact on the tax treaties of Member States. The ruling implies that the freedom of establishment laid down in Article 43 of the Treaty must be interpreted as meaning that a distinction of any kind for tax purposes between subsidiaries and permanent establishments will no longer be acceptable. It has already been clear, since the "avoir fiscal case" (case 270/83), that Member States must grant permanent establishments situated in their territory the same tax benefits as those applicable to resident companies. But the Saint Gobain case goes a step further. It implies that a permanent establishment must be granted treaty benefits under tax treaties, including those with third countries, concluded by the State where the permanent establishment is

located.

The overriding effect of EU legislation

The fact that EC Directives and the Arbitration Convention take precedence over the bilateral tax treaties but are not reflected in the provisions of the treaties also creates complications and difficulties of interpretation for taxpayers. As regards new EU proposals with an impact on company taxation, the proposed Interest and Royalties Directive and the proposed Directive on Mutual Assistance in recovery, once they take effect, should, for example, be reflected in bilateral tax treaties, as should the provisions of the Code of Conduct for business taxation.

Finally, bilateral tax treaties based on the OECD Model Double Taxation Convention often do not resolve many of the instances of double taxation which have been described in other sections of this Part of the study. They do not normally provide a solution to the problem of cross-border loss compensation, or a definitive solution to the costs and risks of double taxation due to transfer pricing disputes. Some but not many tax treaties include a clause to provide tax relief for cross-border pension contributions paid by posted workers (see section 7.2 below), and the treaties do not normally include a provision to deal with the tax treatment of stock options (see section 7.3 below). TheOECD Model and most double taxation treaties do not include procedures to ensure that the more favourable treatment applicable under bilateral agreements as compared to domestic rules is applied speedily and with the minimum of administrative complexity, so as to facilitate cross-border investment.

Conclusion

The analysis has shown that there are a significant number of issues of double taxation which are not currently being properly addressed by the bilateral tax treaties in place between Member States or by domestic tax provisions. This is because they do not cover all bilateral relations between Member States, they do not achieve complete abolition of either discrimination or double taxation and, in particular, they never provide any uniform solution for triangular and multilateral relations between Member States. The number and extent of the complexities and difficulties in this area will increase when the European Union expands. Furthermore, where tax treaties exist, there are many conflicts and inconsistencies between the provisions of these tax treaties and the provisions contained in the Treaty and in EC Directives in the tax field.

Business organisations have on many occasions urged a clarification of the conflicting rules in this area as well as the elimination of double taxation problems not addressed in bilateral tax treaties. Many commentators, including some Members of the Panel assisting the Commission services with the study, have gone so far as to suggest that differences between Member States' tax treaties with each other and between Member States' tax treaties with third countries create distortions and serious problems of treaty shopping. As a result of these problems, the economic behaviour of EU companies is unduly distorted, and overall, EU welfare is reduced.

costs, including increased taxes, that the employee may incur following such a removal.

In addition, it is worth noting that any tax treaty issue facing the employee will normally result in compliance problems for the employer. Among the various tax-related labour costs connected with cross-border activities, the following two categories are worth singling out:

• - 

  costs for occupational pension arrangements , since they probably represent the

single most important cost; and

• - 

  costs for employee stock option plans , since the use of such plans has become more

and more common over the last few years.

Furthermore, both in the case of pensions and stock options, there are considerable risks that double taxation may occur.

There is only little empirical evidence on the relative importance of these problems but both the legal analysis and the indications of the representatives of the various parties concerned, i.e. employers, employees and tax administrations, support the view that the cost related to these problems is substantial. Given the relatively recent character of the phenomena underlying these problems, it is, moreover, only logical that the empirical economic data is currently somewhat limited.

Costs for occupational pension arrangements in cross-border situations

As more and more companies and employees strive to take full advantage of the Internal Market, resulting in an increased mobility of the work-force

93, the problems relating to

occupational pension arrangements are growing in importance. Companies often complain that the diversity, complexity and specificity of the provisions developed over the years at national level are so great that they constitute a major obstacle to cross- border activity.

In many cases Member States' tax laws de facto prohibit cross-border membership of occupational pension schemes. The result is that a company, or group of companies, operating in several Member States have to establish separate pension arrangements or pension schemes in each State where it has activities. The costs incurred may be substantial. Calculations presented by industry indicate that the cost of setting up and managing separate funds could amount to around 40 million per annum for a large pan-European business

stem from the difficulty to obtain tax relief for employer and/or employee cross-border contributions to supplementary occupational pension schemes.

In certain Member States there exist pension schemes that only accept members who are resident in the State where the scheme is established. Even disregarding the tax aspects, employees may therefore in practice not be able to remain affiliated to their old scheme when moving to another Member State on a more or less permanent basis.

In this context it is worth recalling that through Council Directive 98/49/EC95 , which is

to be fully implemented by 25.7.2001, posted workers (as defined in Regulation 1408/71) will have the legal right to remain within their old scheme in the home State. Still, a vast majority of Member States do not grant cross-border contributions paid by or on behalf of posted workers the same tax privileges as for domestic contributions.

The inconsistent interplay of national tax arrangements will often lead to double- taxation, since the employees may end up being taxed for pension contributions made by them and/or by the employer on their behalf, while the pension benefits are also taxed, although at a later stage. Companies frequently have to compensate their expatriates for the extra taxes incurred (even if they are levied at different points in time). They may also have to pay compensation where pension rights are lost due to long vesting periods or because the employees cannot stay within their old scheme.

Box 44:

Tax-related labour compliance costs

Employers involved in cross-border activities are faced with the problem of keeping up to date and complying with a variety of obligations imposed on them in that capacity by Member States. These typically include the liability to pay social security contributions and unearmarked payroll taxes as well as the liability to withhold and pay wage taxes. To determine the liability in respect of wage taxes the employer must in fact also have a clear picture of where the employees are liable to pay income tax on their salaries. To get a better understanding of the difficulties facing an employer in the above respects, short summaries of the various rules applicable are presented below.

Social security contributions

Council Regulation EEC/1408/71 provides rules at Community level on the application of social security schemes to workers moving between Member States. The main rule is that the place of employment (activity) governs which scheme is applicable. In this way it makes no difference for the employer if he hires people of one nationality or the other. Normally, the same rules will thus apply to all workers employed at a certain work-place. Under Regulation 1408/71 frontier worker means any employed person (or self-employed person) who pursues his occupation in the territory of a Member State and resides in the territory of another Member State to which he returns as a rule daily or at least once a week. Such a frontier worker will normally follow the main rule but has some additional rights in the Member State where he/she is resident.

However, in the case of workers who are posted to another Member States under the conditions laid down in Article 14(1) of the Regulation special rules apply: Posted workers will remain affiliated to the social security scheme of the Member State from which they are posted, i.e. the Member State where they are normally employed. The time limit laid down in Article 14(1) may, pursuant to Article 17, be extended through an agreement between the Member States involved. In practice, these so called Article 17 agreements are often concluded for up to 60 months. The liability to pay social security contributions follows indirectly from these Community rules; the contributions are thus payable to the Member State whose scheme is applicable to the worker in question. Some Member States have very strict requirements for the payment of these contributions. Although these are the same for domestic and foreign businesses, the latter are faced with the additional problem of handling two or more systems. They may thus be forced to bring in specialist advisers at an additional cost.

Unearmarked payroll taxes

"Social security taxes" are levied in a number of countries. They constitute a non-negligible part of non- wage labour costs and could thus affect one-off localisation decisions.

Determining the Member State in which the employee is liable to pay tax on his/her salary and consequently the employer's liability to withhold and pay wage taxes

Normally the division of taxing rights between Member States will be governed by double taxation

agreements: Out of 105 possible bilateral relations, 98 are presently covered by such agreements. Generally, the division of taxing rights in respect of dependent activities (i.e. salaries and other remuneration stemming from employment) follows the OECD model tax convention. Where the employee is attached to a permanent establishment in the other Member State, as will normally be the case, his/her remuneration will be taxable in that Member State (to the extent that it derives from activities exercised/performed there). In the exceptional case where the worker is not attached to a permanent establishment in the other Member State, the 183 days rule found in the OECD model tax convention could come into play. Where the work lasts no more than 183 days, the worker (who would be regarded as posted under the social security legislation) would remain liable to tax in his/her State of residence (origin). In new tax treaties the counting of days is normally done on any twelve months basis, but there still exist a number of treaties where the counting is per calendar year. Where the posting, due to unforeseeable events, is extended beyond 183 days, the taxing right would normally, with retroactive effect, pass to the Member State where the activity is exercised (the Member State of posting). Problems will then often arise for the employer with respect to wage taxes due. All Member States except one (France) oblige employers to withhold and pay wage taxes. The amounts will vary with the tax burden on labour.

If the taxing rights move from France to another Member State, following such a prolongation of the stay, a situation might arise where the employee will be able to claim reimbursement of taxes paid by him/her in France, while the employer will be faced with an obligation to pay wage taxes in the other Member State.

Whereas the division of taxing rights in respect of seconded workers is relatively similar under the existing tax treaties, the taxation of workers crossing the frontier on a regular basis vary considerably. Neighbouring countries, when concluding a double taxation agreement, often provide for special provisions for such workers. Generally these provisions are linked to geographical notions such as where the worker is living and where he/she is working; there are conditions relating to how often (daily or, sometimes, weekly) the worker must return to his/her State of residence for the special rules to be applicable etc. These workers are often referred to as frontier workers but, when referring to double taxation agreements, the concept of frontier worker will more often than not be different from the one used in Community acquis on social security (see above).

In the future, completely new questions may arise in the light of developments in communications technologies (internet, e-mails, teleworking, video conferences, etc): How will the traditional cross-border movements evolve and what criteria will be used to categorise "virtual" frontier work?

Companies making use, or wanting to make use, of employee stock option plans are in practice faced with tax problems at two levels:

• - 

  first, there is the tax treatment of costs of the company for the employee stock option

plan itself

• - 

  second, there is the tax treatment of the stock options in the hands of the employees,

which, indirectly, could have a significant bearing on the costs incurred by the employer company.

Different rules in Member States

A majority of the fifteen Member States have special rules in place (January 2001) as regards the taxation of stock options. These may relate to the tax treatment of the costs for setting up and running an employee stock option plan and/or the tax treatment of the stock options in the hands of the employee. Some Member States make a distinction for taxation purposes between approved and non-approved stock option schemes. In most cases, these rules have been introduced or changed during the last three years. Those Member States without special legislation tend to rely on their general tax rules relating to earned income as regards the taxation of the stock options in the hands of the employee.

The accounting costs, financial and legal costs etc. for setting up and running the employee stock option plan do not normally constitute a problem with regard to the right of deduction, at least not where the company incurring the costs is identical to the employer.

However, practices seem to vary more as regards whether the employer is allowed to deduct an amount corresponding to the benefit taxed in the hands of the employee. It may, for instance, be decisive whether the shares granted under the plan are being issued specifically for that purpose or whether they have been previously acquired, but, whichever is the case, the picture seems to be far from clear-cut. In addition to the compliance problems that could result from this, the deductibility or non-deductibility of the benefits could have a bearing on the application of double taxation conventions; this is especially the case where the stock options are issued to employees of a permanent establishment.

variation in tax treatment is largely due to the fact that Member States identify (i) different taxable events and (ii) different measures of income, although other factors

also play a role.

Box 45:

Example on stock options

X and Y are granted stock options by their employer, company Z, which has plants in Belgium, Finland and Sweden. There are certain conditions linked to the stock options: the vesting period is two years, meaning that the options cannot be exercised until two years after they are granted. Furthermore, X and Y have still to be employed by Z when actually exercising the options, i.e. when acquiring the shares.

At the moment the options are granted, X is working and living in Sweden, and Y in Belgium. Towards the end of the vesting period X moves to Finland, while Y goes to Sweden to replace X. After three years both X and Y decide to exercise their options, at which point in time X is working and living in Belgium, and Y in Finland. After yet another year (without change of residence) both X and Y sell their shares, which have increased considerably in value.

X could escape taxation completely: When being granted the options X is living in Sweden, which normally taxes when the options are exercised, but in cases of inbound or outbound transfers of residence, taxes if the person was resident there when the options could first have been exercised, i.e. at the vesting.

At the vesting, X is however living in Finland, which taxes when the options are actually exercised; when exercising the options, X is living in Belgium which taxes at the time the options are granted. Finally, the sale of the shares may not trigger any tax since Belgium does not tax capital gains.

Y on the other hand risks being taxed three times: in Belgium since it taxes when the options are granted;

in Sweden since the options could have been exercised while Y was residing there; finally, Y may be taxed in Finland, since it taxes when the options are actually exercised. However, Y should get at least some relief since the gain will be time-apportioned, i.e. Y should only be taxed on the value accrued during Y's stay in Finland. Finally, Y will be subject to tax on capital gains when selling the shares.

Y's situation of course raises the question whether no alleviation could be achieved under the double taxation conventions between the Member States concerned. The problem then arises that there are no specific provisions on the taxation of stock options in the conventions. In addition, there is the timing problem.

Since there are conditions of continued employment linked to the stock options, it could be argued that the benefit of the stock options refers not only to current or past employment but also to the vesting period of two years or possibly to the whole period until the options are exercised. Furthermore, under Article 15, Dependant personal services, of the OECD Model Convention, the taxation rights of the source country are not formally restricted to income that arises during the time the employee is present in the source State but refer to "remuneration derived from employment exercised in a State".

shares acquired under the option. Finally, countries may consider the leaving of the country by an employee holding stock options as a taxable event.

In practice, it seems that none of the EU Member States (at least none of those having special provisions in place) treats the sale of the shares acquired under the option as the one and exclusive taxable event. In other words, where the option as such is deemed to have a value, this will be taxed at an earlier stage. In this context it is worth noting that the UK tax rules are designed in such a way that, where the conditions for favourable tax treatment of the employee stock options are fulfilled, no benefit is deemed to have arisen, and that for that reason taxation will only take place when the shares are sold.

In most Member States it is thus the exercise of the option that triggers taxation. Where, however, the option is tradable or unconditional, or where the vesting period is very long, taxation would in some of these States be brought forward to the granting. In one Member State (Belgium) the granting is the main taxable event. Furthermore, a few Member States, e.g. Denmark, the Netherlands and Sweden have special rules relating to transfers of tax residence, inbound as well as outbound (January 2001). On this basis, the risk of non-taxation currently seems to be somewhat more frequent than that of

double-taxation.

The measure of income

The measure of income used is necessarily linked to the definition of the taxable event and will thus also vary. If the options are taxed at the moment they are granted , the value would normally be estimated on basis of the value at that time of the underlying shares (minus any amount paid for the options). For options not quoted on the stock exchange the evaluation could pose considerable difficulties. Taxation at this point in time may be advantageous where the value of the shares increases, but it also involves a risk from the tax payer's point of view, as he/she cannot be certain that it will ever be favourable to actually exercise the option.

If the options are taxed at the moment of exercise , the taxable benefit may be determined as the value of the shares acquired (minus any amount paid for the options, i.e. the strike price).

Double taxation conventions

The division of taxing rights in respect of employee stock options has so far never been expressly regulated under any double taxation convention concluded. Therefore, current double taxation agreements are no guarantee against double taxation or non-taxation of share options received by employees in one country and exercised in another.

First, there are the problems linked to the timing of the taxation under the various systems applied by Member States as described above. Second, there are problems linked to the categorisation of the income under double taxation conventions:

Generally, it would be looked upon as remuneration for services rendered. As such it would normally fall under Article 15, Dependent personal services, of the OECD model convention (or, in the case of board members, under Article 16, Directors' fees). Another possibility is that the advantage of (gains from) the stock option could be viewed, wholly or partly, as capital gains falling under Article 13; the argument would be that the holding of the option and the exercise constitutes an investment.

If categorised as remuneration for dependent personal services, the question could arise, for someone moving (or having moved) from one country to another, whether the advantage refers to past services, to future services (to be carried out by the employee during the lifetime of the option), or, possibly, to both. This could be decisive for the division of taxing rights between the contracting states.

Given the various factors involved, it is conceivable that one employee moving between Member States may completely escape taxation while another could, in extremis, be taxed three times (or, if taking into account capital gains taxation, even four times) for the same stock options. Furthermore, there is no guarantee whatsoever that a Member State taxing the stock options at some stage, would take account of taxes paid earlier or later, in other Member States, in respect of the same stock options.

Conclusion

enterprises can only be properly assessed by examining the cumulative effects resulting from the current structure of value added tax (VAT).

The particular situation for small and medium-sized enterprises

There is no common Community tax definition of small and medium-sized enterprises.

At EU level, small and medium-sized enterprises are generally defined as an enterprise which has fewer than 250 employees and an annual turnover not exceeding 40 m. or an annual balance sheet total not exceeding 27 m. , and in which no enterprise or enterprises which themselves are not small and medium-sized enterprises own 25% or more of the capital or of the voting rights

• 96. 

  In 1998, more than 99.8% of the 17.9

million enterprises in the EU were small and medium-sized enterprises under this definition, employing 66% of the private-sector workforce and generating 56.2% of total turnover

97.

"Small and medium-sized enterprises" include a broad variety of businesses, ranging from traditional craftsmen doing occasional business abroad to international start-up companies created by big multinationals (although the latter are not really covered by the above EU definition in the strictest sense). Small and medium-sized enterprises are often run as non-incorporated companies that are not subject to corporation tax (this aspect is considered separately below). As alluded to in Part I of the study, in comparison to the situation at the beginning of the 90s, small and medium-sized enterprises today need to "go international" much earlier, thus accentuating their possible specific cross-border tax problems.

All the various tax obstacles considered above also concern small and medium-sized enterprises with cross-border activities. Their relative importance and assessment may, however, differ. Due to the smaller business size, some tax obstacles may be more relevant than others. The often specific nature of their business activities means it that some of the cross-border tax obstacles are particularly felt by small and medium-sized enterprises (and in particular technology-driven start-up companies). Moreover, in the domestic context most Member States apply special tax arrangements for small and medium-sized enterprises that need to be taken into account. These arrangements essentially concern the determination of the tax base, flat-rate arrangements and other simplified methods of profit determination. Some Member States also grant specific lower rates. The combination of both effects - the particular importance of cross-border obstacles for small and medium-sized enterprises and relief for domestic tax problems - may even increase the hurdle for starting cross-border business for small and medium- sized enterprises.

Generally, small and medium-sized enterprises have particular difficulties in meeting the compliance costs resulting from the need to deal with up to 15 different taxation systems. This will sound obvious to most tax practitioners. However, this is also supported by scientific and quantitative evidence because, as indicated above, all available studies suggest that compliance cost are regressive to size and put a disproportionately higher or even prohibitively high burden on small and medium-sized enterprises compared to bigger companies

98.

More specifically, particularly burdensome compliance costs relate to the lack of transparency and the variety of domestic tax laws (including on other taxes than company taxation), administrative structures and administrative requirements, including differences in accounting and bookkeeping rules, as well as frequent changes to those same laws, etc. Moreover, many administrative aspects of company tax law directly impact on the cash-flow of the enterprises. The costs of timing differences between the receipt of income and the payment of the taxes levied thereon as well as time lags in the reimbursement of tax paid are important examples. Similar refund problems arise with the wrong levy of taxes which are not due or the amount of which has been assessed incorrectly by the tax authorities. The same holds for taxes which have to be refunded because of jurisprudence. Given their limited reserves, small and medium-sized enterprises are particularly hit by these delays in general tax reimbursements and the recovery of wrongly levied taxes (including VAT). The Commission services are aware of concrete cases in which national tax administrations refuse to accept the deductibility of business expenses unless the complete bookkeeping of the foreign enterprise and all relevant documents are presented in a certified translation into the language of the Member State in question. For small and medium-sized enterprises such costs can be extremely difficult to bear and put into question the commercial rationale of doing business in that state.

The effects of tax competition, as outlined above, appear in most cases to put small and medium-sized enterprises at a relative disadvantage as they are often not in a position to exploit the opportunities opened by appropriately designed tax schemes. Generally, small and medium-sized enterprises do not have tax planning possibilities which match those of big companies

. This can be illustrated with regard to dividend taxation:

offset against profits. This is far from being true for small and medium-sized enterprises. In this context, it is important to remember that business start-ups are almost by definition small- and medium-sized enterprises.

Last but not least, small and medium-sized enterprises face specific cross-border problems relating to other taxes than company taxation. For instance, VAT is a problem of particular relevance to small and medium-sized enterprises. It should also be noted that the transfer of small and medium-sized enterprises, often family businesses, entails a number of tax problems that are often more difficult and onerous than for big publicly quoted companies. Cash-effective gift and inheritance taxes are a prominent example. There is anecdotal evidence of small and medium-sized enterprises relocating to other Member States mainly to ensure that the later succession of the business to the family heirs takes place without a tax burden that is perceived as being disproportionate if not confiscatory. These problems fall outside the scope of the mandate for this study and are therefore not considered in detail. They should, however, not be overlooked

100.

In short, the assessment of the tax obstacles is to some extent different for small and medium-sized enterprises. It may be recalled, however, that the Ruding report found thatEU corporate tax systems are mostly neutral with respect to the size of the business. Specific tax incentives for small and medium-sized enterprises do not alter these findings. The above reflections do not provide any reason to question this analysis.

Company tax obstacles and partnerships

The importance of partnerships varies considerably between Member States. For instance, in Germany, +/- 85% of all businesses, some of which are large multinational enterprises, are run under one of the various legal forms of 'partnership' available under German company law whereas in other Member States only small groups of individual entrepreneurs or very small businesses choose this legal form. Recently, partnerships are becoming increasingly important as legal form for business start-ups in the "new economy" and/or for joint ventures of large multinational enterprises involved in R&D, telecommunications, e-business and similar commercial fields

• 101. 

  There the intention is,

among other things, to spread the burden of initial losses (between companies in different jurisdictions) as such 'joint branches' allow losses to be shared with another partner and for cross-border loss compensation

being homogenous. Frequently, what is considered a 'partnership' in one Member State would be considered a 'company' by another. For instance, it is unimaginable under, say, Italian law that a hybrid legal form like the 'GmbH & Co KG ' in Germany is a partnership (which is its legal status under German company law). Agreement on a precise definition seems almost impossible as these terms are embedded in deeply rooted cultural, company law and civil law traditions. Whereas stock corporations and other forms of incorporated limited liability companies can relatively easily be defined, it is generally possible to distinguish on an EU-wide basis:

• - 

  CIVIL LAW PARTNERSHIPS - in some Member States with a civil law tradition, any

-

gathering of a group of people for a common purpose which is not organised in a specific legal form constitutes such a partnership which may, dependent on the nature of its activities, generate taxable business income;

• - 

  GENERAL OR ORDINARY PARTNERSHIPS - like an individual entrepreneur, all partners

-

have unlimited liability;

• - 

  LIMITED PARTNERSHIPS - some partners have unlimited liability, some have limited

-

liability only; in some countries it is not uncommon that the partner with limited liability is a limited liability company;

• - 

  SILENT PARTNERSHIPS - the enterprise does not act as such towards third parties; at

-

least one partner has unlimited liability; the other partners have either unlimited or limited liability;

• - 

  SLEEPING PARTNERSHIPS - the 'sleeping' partner is not made known to third parties;
• - 

  EUROPEAN ECONOMIC INTEREST GROUPING (EEIG) - a Community law103 instrument

for facilitating co-operation between firms established in a number of Member States.

In most countries, 'partnerships' are treated as 'transparent' entities, i.e. they are not taxable entities but used for the purpose of computing taxable income to be attributed to the partners. The profit share attributed to the individual partners is then subject to personal (in most cases progressive) income tax, or corporate income tax when the partner is an incorporated company

another form of profit-distribution). Those are only some criteria for categorising the tax treatment of partnerships in Member States.

In cross-border situations, this gives rise to a significant risk of double taxation.105

Member States may or may not consider a 'partnership' to be a company for the purposes of their tax treaties. And even when the classification of a 'partnership' is the same in the residence country and in the source country, as shown above, the tax treatment can be different. Member States' qualification of partnerships as resident for treaty purposes may also be different as either the residency of the entity or that of the partners can be relevant. As regards the taxation of non-resident holders of partnership interests, again whether the partnerships is 'transparent' or a taxable entity is decisive and can lead to double-taxation in a cross-border situation (taxation of the income both at the level of the partnership and the partner). Similarly, double taxation can arise in transactions between the partnership and its partners. A good example of this is the treatment of cross-border interest payments on a loan from a partner to the partnership: in principle the interest is taxable in the hands of the partner and should be deductible while computing the income of the partnership; without deductibility, double taxation occurs. Other problems relate to the granting of tax credits.

It clearly follows from the foregoing that partnerships need to be taken into account when considering possible targeted or comprehensive remedial measures to the various tax obstacles. It is noteworthy that the OECD is considering these questions with a view to possible changes of the OECD model tax convention. In this context, it has published a comprehensive report on the subject

106.

The cumulative effects of VAT difficulties for small and medium sized

enterprises

Strictly speaking, value added tax (VAT) falls outside the scope of an analysis of company tax obstacles in the Internal Market. Some of the basic problems encountered by enterprises are however similar and/or linked to company tax issues. Moreover, it would look somewhat odd to neglect VAT, the application of which is so closely linked to the technical functioning of the Internal Market. In particular, the impact of difficulties encountered especially by small and medium-sized enterprises in the field of direct taxation can only be evaluated properly if seen in the context of the cumulative effects resulting from the current structure of VAT. Generally, indirect taxes are relatively more important for the service industry and for retail trade. But all companies are subject to VAT and thus confronted with any problem that the application of the tax might give rise to resulting from cross-border economic activity within the EU. However, small and medium-sized enterprises are particularly hit by such problems.

form of sub-contracts requiring the application of specific (complex) VAT rules. Generally, the fixed cost caused by VAT obligations constitutes a larger share in the relatively small turnover of small and medium-sized enterprises as opposed to larger companies. Therefore, this section focuses on the specific VAT problems of small and medium-sized enterprises. This does not mean that large companies do not face the same or similar problems, they can build on a bigger infrastructure to deal with them in practice.

As already suggested in the case of direct taxation, for VAT purposes, small and medium-sized enterprises are not confronted with a single market but - despite the degree of harmonisation achieved in the VAT field - 15 different jurisdictions and 15 different sets of rules. However, in respect of VAT, the difficulties for small and medium-sized enterprises begin even earlier, since almost all cross-border supplies of services and goods are governed by VAT rules. The complexity of the VAT rules, generating high expenditure on specialist tax advice and high compliance costs, create access barriers for small and medium-sized enterprises to setting up business in or with another Member State.

107

Small and medium-sized enterprises when considering supplying services or goods into another Member State, must first identify the correct jurisdiction in order to determine the applicable VAT rules. The jurisdiction is determined by the place of taxable transactions

• 108. 

  Under the current VAT system, there are some 25 rules applicable to

determine the place of supply of taxable transactions. The various thresholds for the application of specific rules that are provided in the 6

th VAT Directive are also

implemented differently in various Member States.

Thus, this is the moment when small and medium-sized enterprises realise that they are about to enter a tax jungle of complex and complicated rules in which they need professional tax advice. At this early stage and long before any profit is realised, small and medium-sized enterprises already have to invest considerable financial resources for the acquisition of special tax know-how, creating an access barrier to doing business in or with another Member State. This barrier is substantial, because the expense at acquiring this specialised tax knowledge is also considerable.

obligations under the rules of the jurisdiction of establishment have to be understood (tax representation, invoicing, declarations, payments, deduction or reimbursement procedures, etc.). And even then, in practice many technical questions will be handled differently by the tax administrations of the different Member States concerned, leaving the small or medium-sized enterprise with a substantial risk of non-compliance.

Box 46:

Difficulties for small and medium-sized enterprises to live up to the requirements

of the present EU system of VAT

A small but expanding Dutch business (A) decides to explore other Member States for supplying remote controlled alarm and surveillance systems, their maintenance, and also for supplying technical and legal advice for making best use of these systems (negotiation with insurance companies, trade unions, etc).

If A presents his goods and services to a potential client in another Member State, he may incur business expenditure (fuel, food, hotel accommodation) there. The VAT on this expenditure can only be recovered by means of a lengthy and complicated procedure.

If A succeeds in finding a client in another Member State, he will have to find out where the place of VAT taxation of his own supplies is. He may have to register in one or even more Member States and/ or pay a fiscal advisor to obtain a clear picture on the rules to be applied.

Where his supplies turn out to be taxable outside the Netherlands, he may - at least currently - have to nominate a fiscal representative in the country/ countries of supply. Even if A can escape that obligation, the complexity of the rules will force him to pay a fiscal advisor to comply with his fiscal obligations in the country of supply (invoicing, tax declaration, tax payments etc.).

It is here that the access barrier to cross-border trade becomes fully visible to small and medium-sized enterprises. Where fiscal representation is obligatory (e.g. building services or building related services, including the setting up of stands for exhibitions and fairs, an area in which many small and medium-sized enterprises are doing business), the costs of that tax representation usually cover or even exceed the profit which could be gained from doing business in another Member State. This obligation, still applicable in some Member States, will however disappear shortly

Similarly, the costs of documentary proof for intra-Community acquisitions of goods are huge (returns, id.-no, supplier-id., Intrastat, etc.) and often equal or exceed the benefits of lower prices for imports from other Member States.

The conditions and requirements of proof demanded for reclaiming input taxes under the 8

th VAT Directive create another costly barrier for small and medium-sized

enterprises. The procedural requirements differ from Member State to Member State, there are long processing times and considerable language barriers for claiming back input taxes. This often leads to additional need for tax advice and increases costs as well as cash flow problems. In particular, small and medium-sized enterprises have considerable cash flow problems when VAT is charged on the basis of estimates and payment is due prior to receipt of payment from the customer.

Conclusion

Due to their size and limited resources, small and medium-sized enterprises are particularly hit by the various company tax obstacles that have been identified so far and suffer from disproportionately high compliance costs for cross-border economic activities. The overall welfare losses implied by the need to deal with a multiplicity of tax jurisdictions are strongly felt by small and medium-sized enterprises which are indeed often deterred from entering the Internal Market, although the economic pressure on them for doing precisely this is constantly growing. Administrative burdens and problems in relation with cross-border loss-offset are the most important obstacles in this context.

VAT problems are a particular concern for small and medium-sized enterprises and these significantly exacerbate the overall impact of cross-border company tax obstacles.

In an international context, the use of partnerships can cause numerous complex tax problems. This is particularly true with regard to the application of double-taxation treaties. Moreover, the question arises as to whether specific EU arrangements for corporations should also be accessible for partnerships and if so, under which conditions.

PART IV:

REMEDIES TO THE COMPANY TAX OBSTACLES IN

THE INTERNAL MARKET

IV.A

The framework for possible remedies

3. THE CASE FOR STUDYING BOTH TARGETED AND COMPREHENSIVE REMEDIAL

MEASURES

There are in principle two types of measures to address the above tax obstacles. The first possibility would be, on the basis of the separate analysis of the obstacles one by one, to try separately finding a targeted solution for each specific problem. Thus, a number of specific actions for improvement can be identified to minimise individual obstacles. The second possibility would be to devise more comprehensive, all-embracing approaches which could eventually minimise, or remove altogether, the obstacles in a more unified manner. This approach would also, at least in principle, be more in line with the philosophy of a Single Market. In this perspective the underlying idea of such comprehensive approaches represents a legitimate ambition and ultimate goal for the development of company taxation in the EU. These two basic approaches are not mutually exclusive but priorities will have to be established. There are numerous concrete suggestions put forward in the literature under both approaches.

On the one hand, there are good reasons to believe that some of the targeted measures appear to be important regardless of whether or not comprehensive solutions are introduced. Some of the above obstacles are, for various reasons, of a nature that makes it almost impossible to address them by means of comprehensive schemes (e.g. those relating to stock options, partnerships). Therefore, some targeted measures appear to be necessary in any event. Moreover, as will be explained, most of the comprehensive approaches may be conceptualised as optional schemes and would thus be operated as a parallel system on top of the traditional domestic company tax system. Thus, the non- participating companies would still be faced with at least some of the obstacles. It also seems that certain types of companies, for instance in specific industries, would by definition be excluded from comprehensive schemes. Finally, it is likely that any comprehensive scheme would, at least to a certain extent and in a transitional phase, create problems of its own and, possibly, fail to address the interplay of companies operating under comprehensive but different schemes (e.g. mergers). These considerations by no means put into question the value and potential benefits of comprehensive approaches. It only underlines that accompanying targeted measures will still be necessary even if comprehensive schemes are realised.

It follows from the foregoing that it is necessary to analyse both basic approaches and the various detailed technical ideas and suggestions behind them in this study. The final assessment of which approach is better or whether a combination of both is appropriate can, of course, only be made after having carried out this analysis. However, it is also clear that only a truly comprehensive approach can eventually match the requirements of the Internal Market.

Before starting to consider the two basic approaches for devising remedial measures in detail, two important general issues are explored: the role of the European Court of Justice in and the possible impact of financial accounting rules on remedying tax obstacles in the Internal Market.

THE ROLE OF THE EUROPEAN COURT OF JUSTICE IN REMEDYING TAX

OBSTACLES IN THE INTERNAL MARKET

Regardless of the basic approach to remedies, it is important for a proper assessment of possible solutions to look first into the guidance given and limitations indicated by the jurisprudence of the European Court of Justice. In the absence of political solutions taxpayers have been compelled to have recourse to the legal process to overcome discriminatory rules and other obstacles. In consequence, the European Court of Justice (ECJ) has developed a large body of case law on the compatibility of national tax rules with the Treaty. National courts are also increasingly being asked to give rulings in this area. As repeatedly referred to in the analysis so far, since the publication of the Ruding report, the law on direct tax matters in the EU has developed at an unprecedented rate, and today the jurisprudence of the ECJ strongly influences almost all aspects of company tax law. There is good reason to believe that this influence will continue to increase. While the ECJ has unquestionably made a significant contribution to the removal of tax obstacles for companies, it is unlikely that the interpretation of the Treaty is sufficient to address all tax obstacles to cross-border activity.

These recent developments have two implications for this part of this study. First, any possible solution to the existing tax obstacles must take account of the existing judgements of the ECJ. Secondly, it is necessary to analyse to what extent the European Court of Justice jurisprudence calls for further co-ordination between Member States, i.e. beyond the mere correction of a Treaty infringement.

closely linked to, both international law and the legal systems of the Member States110.

Since the creation of the European Communities this hybrid nature of Community law has required clarification by an independent and non-political institution. In the Treaty this role is assigned to the Court, which under Article 220 has the primary tasks of interpreting Community law and ensuring its uniform application throughout the Community.

By carrying out these tasks the Court has actively contributed to the development of Community law. In addition to the questions of law on which it has ruled, the Court has acknowledged and confirmed that a number of general principles of law are inherent in the Community legal order. By far the most fundamental of these general principles are the supremacy and the direct effect of Community law. The principle of supremacy ensures that Community law has primacy over conflicting national law, while the principle of direct effect means that individuals can invoke their Community rights directly before national courts

111.

These two principles, derived by the Court from the Treaty, are essential elements of the

acquis communautaire (the state of Community law to date) and the reason why the role of the Court in the Community legal order has been and continues to be so important. They provide for full and effective protection of Community rights and have established a framework for review by national courts of all national measures that fall within the purview of Community law. Without them, Community law would be of greatly less significance.

The role of the Court can be observed from many different angles, but this section focuses on its case law, particularly that which has tested the compatibility of national tax rules with Community law. As the Court actively contributes to the development of Community law and the interpretation of Community law falls, in the last resort, within the exclusive competence of the Court, its role has often been assimilated with that of a supreme or constitutional court within national legal systems. Its jurisprudence has established a number of precedents which are not only referred to and followed by the Court itself, but also provide guidance to national courts when they, in accordance with the principle of supremacy, apply Community law and set aside conflicting national rules.

The tangent at which Community law and national laws on direct taxation meet is a result of the combined application of the four freedoms and the principle of equal treatment. In 1986, the Court extended its case law on the four freedoms to the sphere of direct taxation when it gave judgement on Case C-270/83 Commission v France , commonly known as "avoir fiscal" . The Court held that a national tax law which refused a dividend imputation tax credit to permanent establishments of foreign (non- resident) companies, whilst granting it to resident companies, was contrary to Community law. Unsurprisingly, this decision caused a great deal of confusion among practitioners of international tax law at the time as for them it was practically unheard of that non-residents and residents could not be subjected to different treatment.

Since the decision in avoir fiscal , the jurisprudence in this area has developed rapidly and it is perhaps fair to say that of all the Community institutions, the Court has so far proved to be the most efficient at removing tax obstacles to cross-border economic activities within the Community. The reminder of this section thus extracts recurrent themes from the Court's case law in order to establish what progress it has made, and to consider how this progress could be furthered and reinforced.

The principle of equal treatment and the four freedoms

Non-discrimination

The principle of equal treatment, which the Court has derived in part from the Treaty112

but also from the national laws of Member States, has had a decisive influence on the interpretation of the Treaty itself. It is of particular importance to, and forms a fundamental element of, the provisions of the Treaty which establish the Internal Market. As pointed out repeatedly above, violations of the equal treatment principles generate tax obstacles to cross-border economic activity in the Internal Market.

In general, discrimination can be defined as treating similar situations differently, or different situations alike. This definition is used in both Community law

113 and

international tax law but the notion of discrimination in Community law differs from that of international tax law. Under Community law it is sufficient that situations are materially similar, whereas international tax law rules are based on the assumption that residents and non-residents are in a different situation and can therefore legitimately be subject to different treatment

In Community law prohibition of discrimination is a common thread for the four freedoms provisions (the free movement of goods, persons, services and capital, and the right of establishment). They give specific expression to the prohibition of discrimination on grounds of nationality in Article 12 of the Treaty, itself a manifestation of the general principle of equal treatment. Nationality is a concept commonly attributed to natural persons but Article 48 of the Treaty includes as beneficiaries of these provisions also companies constituted in accordance with the law of a Member State and having their registered office, central administration or principal place of business within the Community. Accordingly, a company is considered a national of the Member State in which it has its seat.

Thus Article 43 of the Treaty prohibits a Member State from imposing discriminatory restrictions on the freedom of companies of other Member States to establish themselves within the territory of the first Member State through a branch or subsidiary. Restrictions on the freedom to provide services are prohibited under Article 49. Article 56, in conjunction with Article 58 prohibits discriminatory restrictions on the free movement of capital and payments. So far the Court has found cases in the area of direct taxation to be within the scope of the four freedoms provisions of the Treaty and the prohibition of discrimination as reflected in them has therefore been sufficient. To date the Court has not examined Member States' tax laws directly in the light of Article 12 of the Treaty, the general prohibition of discrimination in Community law.

In a tax context, differences in treatment include increasing the tax burden (due to a higher rate or wider base) and procedural disadvantages, for example in the way the tax is assessed or collected. In its case law the Court has acknowledged that residents and non-residents may be in materially dissimilar situations because of the differences in taxing rights exercised over them. Nevertheless, it looks closely to see whether the differences are in fact material. In accordance with the objectives of the Treaty the Court has adopted a narrow interpretation of derogations from its fundamental principles of free movement and thus has taken the view that it is sufficient that the situations of the resident and non-resident individuals or undertakings are substantially the same. For example, the advantages commonly refused to permanent establishments under domestic rules and double taxation conventions have been considered discriminatory even though there are differences in the treatment of permanent establishments and subsidiaries. It is sufficient that both are subject to tax on profits

be comprised of a high proportion of foreign nationals constitutes covert discrimination as it has in fact the same effect as a measure overtly targeted at foreign nationals. In a tax context this means that, at least in principle, differences in the treatment of non- residents, being mainly foreign nationals, are likely to be considered discriminatory.

Justifications

The principle of equal treatment can only be overlooked if the discriminatory national rule can be justified on the ground of "imperative requirements of public interest" (objective factors other than nationality) and the adverse treatment is proportionate to the objectives pursued by the rule

• 117. 

  In order for the adverse treatment to satisfy the

proportionality test it must be necessary in the sense that there would be no other, less restrictive means to protect the public interest in question. An insight into the case law concerning national rules on direct taxation shows that the Court has enforced the principle of non-discrimination very strictly. In line with general principles developed outside the tax field, the Court has rejected a number of justifications for discriminatory measures advanced by Member States and many of them repeatedly. These include for

example:

• - 

  the need, in the absence of harmonisation, to take account of differences between

national tax rules;

• - 

  the fact that a non-resident could have avoided the discrimination e.g. by setting up a

subsidiary company rather than a branch;

• - 

  economic aims or the protection of tax revenue;
• - 

  the absence of reciprocity;
• - 

  the existence of discretionary or equitable procedures to ensure appropriate fiscal

treatment.

The Court has expressly refused to accept arguments based on the factual overall treatment of a non-national taxpayer as compared with nationals of a Member State. For

example, in avoir fiscal

The Court has generally been reluctant to accept justifications put forward on the basis of the administrative difficulties involved in ensuring efficient fiscal supervision or the prevention of tax avoidance. It has taken the view that Member States should, if need be, provide each other with mutual assistance to overcome such difficulties. In two earlier cases (one of them commonly known as Bachmann ) the Court did hold that a discriminatory provision could be justified by the public interest in preserving the fiscal coherence of a Member State's tax system

• 119. 

  In these cases concerning Belgian tax

rules the proportionality test was considered to be met and the justification was accepted on the ground that there was a need to ensure that a tax deduction granted in respect of pension or life assurance premiums was matched by ultimate taxation of the benefits paid out under the relevant policy.

These two cases have, however, been widely criticised and the Court has indeed shown great reluctance to accept the fiscal coherence type of justification argument ever since. The Court begun to limit the scope of the fiscal coherence as an imperative requirement since 1995

120 and many commentators now question whether the scope of that principle

remains good law. As also more recently demonstrated by the opinions of Advocate

General in Verkooijen

121 and Advocate General in Guipúzcoa 122 the applicability of the

Bachmann defence is limited to situations where a discriminatory rule refusing a deduction for a payment is justified by inability to tax the recipient of the payment. Hence, a justification of a discriminatory measure on the grounds of "fiscal coherence" requires the existence of a direct link between deduction and taxation within the same tax system and as expressed by the Court in Eurowings

123 when referring to Bachmann ,

a "tax disadvantage ...[must be ] compensated for by a corresponding tax advantage for the same person".

In national rules, there is rarely a strict correlation between deductions and benefits. This is even less so if one takes account of bilateral conventions. As the Court noted in

Wielockx

124, "the effect of double-taxation conventions which follow the OECD model

is that the State taxes all pensions received by residents in its territory, whatever the State in which the contributions were paid, but, conversely, waives the right to tax pensions received abroad even if they derive from contributions paid in its territory which it treated as deductible".

State of a permanent establishment of a company established in another Member State is discriminatory and incompatible with the Treaty freedoms

125.

Although much of the case law is concerned with non-residents who are nationals of another Member State, the Treaty also protects individuals from measures adopted by their own Member State which restrict the exercise of Treaty freedoms. Thus, for example, the Court has confirmed that tax measures designed to allow groups of companies to be taxed in broad terms as single entities in a Member State (group reliefs and tax deductible transfer of assets between group companies) may not be construed so as to discriminate against group structures that involve companies having their legal seats in other Member States

126.

Moreover, Member States have, within the framework of Community law, the competence to conclude double tax treaties. However, as demonstrated by the decision

of the Court in Compagnie de Saint-Gobain

127, this competence does not mean that the

Member States would be entitled to impose discriminatory restrictions against nationals

of other Member States.

Against this background, among others the following national measures have been found to be incompatible with the Treaty freedoms:

• - 

  a dividend tax credit granted to companies resident in France but refused to the

branch of a company having its seat in another Member State;

• - 

  a refund of overpaid income tax granted by Luxembourg to permanent residents but

refused to taxpayers moving to another Member State during the tax year;

• - 

  personal reliefs granted by Germany to residents but refused to non-residents even

where they could not benefit from such reliefs in their State of residence;

• - 

  a business relief (a tax deduction for transfers of funds to a pension reserve) granted

by the Netherlands to residents but refused to non-residents;

-

granted in cases where such dividends have been derived on the basis of domestic shareholdings;

• - 

  refusal by Belgium to set off losses from a previous tax year because of profits

arising in a permanent establishment located in another Member States that same year.

• - 

  refusal by Germany of dividends received related double tax treaty benefits (deriving

from treaties with third countries) to a branch of a company having its seat in another Member State.

Beyond discrimination?

The provisions regarding the Treaty freedoms refer generally to "restrictions" to the exercise of the freedoms guaranteed by them. In its non-tax case law the Court has repeatedly held that non-discriminatory restrictions to the free movement of goods, are unlawful unless justified by defined imperative requirements of public interest. As early as in 1974, in its decision in Dassonville

128, the Court held that all trading rules which

are capable of hindering directly or indirectly, actually or potentially, intra-Community trade are in contradiction with Article 28 of the Treaty.

In the widely cited decision in Cassis-de-Dijon 129, the Court qualified the compatibility

of such restrictions with the Treaty freedoms (because of their negative effect on trade)

to situations where they are necessary for the protection of certain public interests, such as fiscal supervision, public health and consumer protection. Thus, for example, domestic product regulations cannot be applied to products imported from other Member States, even though they did not discriminate against imported goods, unless such restrictions can be justified on imperative grounds such as fair trading, consumer protection, environmental protection. In Säger

130, the Court transposed this "restriction

based approach" to cover the free provision of services.

In the direct tax sphere the Court has so far only applied this analysis unequivocally to compliance issues (such as accounting records required of a branch to substantiate losses

It remains to be seen how far the Court is willing to go in adopting such analysis in the field of direct taxation but if it were unequivocally transposed to this area, the removal of the need to show discrimination would certainly broaden the range of obstacles falling within the scope of the Treaty provisions.

Unresolved issues

As discussed above a generally important question to which clarification is still awaited is that of the extent to which non-discriminatory restrictions to the exercise of the Treaty freedoms will be considered unlawful. A particularly delicate area in this respect is the interpretation of the free movement of capital and payments as provided for in Articles 56 and 58 as the latter makes an express reference to permissible non-discriminatory restrictions whilst at the same time it prohibits arbitrary discrimination and disguised restrictions.

More specifically there are a number of questions to which the answers depend on the development of the restriction type analysis in the area of direct taxation. Further guidance could be expected in such areas as restrictions on certain cross-border services, Member States' rights to retain their taxing rights following cross border reorganisations, compatibility of withholding taxes on interest and royalty income and taxation of cross-border dividends.

As to the issue concerning tax treatment of cross-border dividends under systems which integrate the tax treatment of direct investment both at the company and the shareholder levels (but only as regards domestic situations), some light has been shed on by the

recent ruling in Verkooijen

• 133. 

  At issue in this case was the legality of Dutch rules

exempting (up to certain thresholds) dividends received from Dutch companies. The Dutch Government argued that the exemption was intended to reduce economic double taxation and that the limitation of the exemption to dividends from Dutch companies was justified by the necessity of coherence of its tax system.

In the light of the development of the Court's case law since the decision in

Bachmann

134, it was not particularly surprising that the Court did not accept the

establishment provided for in Article 43 of the Treaty, they must be shown to be justified and proportionate to the aims sought by them.

Another set of open questions centre on issues traditionally dealt with in bilateral or multilateral double tax treaties between independent tax jurisdictions. A general question in this area is whether a failure of a Member State to prevent double taxation entails discrimination or restriction contrary to the Treaty freedoms. Another undecided issue is to what extent the Treaty limits the Member States' external competence to conclude double taxation treaties with other member States or third countries. There is as yet no decision concerning the extent to which a Member State can offer differing privileges to nationals of other Member States under its bilateral treaties with other Member States or whether indeed the Treaty imposes an obligation to the Member States to offer nationals of other Member States the most favoured nation treatment as offered under their treaties with third countries.

Some recent case law comes close to these questions135 but it nevertheless remains

unclear whether all differences between tax treaties will be incompatible with the equal treatment principle. In particular it is arguable that the equal treatment principle does not allow reciprocal concessions which go beyond mere allocation of taxing rights, such as differences in concessions to avoid economic double taxation (refunds of imputation credits).

Conclusions

Notwithstanding that direct taxation largely remains part of the national sovereignty of the Member States, national tax provisions may be incompatible with the requirements of the EC Treaty and thereby void. This is because of the impact of provisions of Community law and the general principles of law derived from it and from the national legal systems of the Member States. It is not unprecedented that the Court, whilst being mindful of the consequences of its decisions in respect of established patterns of international tax law, dares to interpret Community law in favour of European integration.

The Court has adopted a very narrow interpretation of derogations to the fundamental principles of free movement of persons and capital, freedom of establishment and free provision of services in the area of direct taxation. The relatively robust approach in the recent case law is reflected in the Court's decisions in which it has defined the limits of the applicability of the "fiscal coherence" justification. The evolution of this view has, however, taken some time and caused considerable uncertainty as to the meaning of the prohibition of discrimination in the area of direct taxation. There is also a suggestion of a tendency towards restriction type analysis by the Court in its case law concerning direct taxation rules of the Member States.

Despite the significant potential of the Court for the removal of existing obstacles to cross-border economic activities it is clear that tackling such obstacles exclusively through judicial process before the Court cannot be sufficient. ECJ rulings are confined to the particular case put to it and may therefore relate solely to individual aspects of a more general issue, the implementation of ECJ rulings is left to Member States, who often fail to draw the more general consequences which flow from them

Moreover, case law in the area of direct taxation has not reached maturity and the full consequences for national tax rules remain uncertain. There are many reasons for this but perhaps the most significant is the limitations inherent in the jurisdiction of the Court. Case law concerning direct taxation results most often from actions begun in a national court from which references for preliminary rulings are made to the Court as provided under Article 234 of the Treaty. Only a few cases have been brought to the Court as direct actions against the Member States. The potential incompatibilities of national tax laws with Community law have therefore not been systematically detected and tested. Equally, the jurisdiction of the Court does not permit it to rule beyond the specific questions of law that have been posed to it.

communications or recommendations. This approach seems to be necessary in order to stop, from a Member States' perspective, the Court from dismantling domestic tax systems. Positive legislation in Member States in advance of Court rulings is therefore the only constructive way forward. Areas of particular importance in this respect are discussed in more detail in the following sections of this part of the study.

THE IMPORTANCE OF FINANCIAL ACCOUNTING FOR REMEDYING TAX

OBSTACLES IN THE INTERNAL MARKET

One basic finding of the analysis of the tax obstacles was that many of them relate to the multiplicity of the rules of 15 tax systems which, in turn, is - to a varying degree - linked to different systems of financial accounting. At the same time, for many non-fiscal reasons accounting harmonisation is still on the international agenda. Before devising tax solutions, it is therefore necessary to consider these links in detail and to examine if and how accounting harmonisation must or can contribute to remedying the tax obstacles, before tax-specific remedies are considered.

In all Member States, the provisions for determining the corporate tax base and financial accounting rules are linked to some extent. Accounting rules are based on the same principles across the EU, but the details vary significantly from one Member State to another. Formal bookkeeping requirements also differ. Since the Ruding report important new developments have taken place in this area.

Generally, it is important to distinguish "financial accounting", i.e. the statutory accounts that companies have to produce and publish regularly for informing the public, and internal "management accounting", i.e. the continuous elaboration of cost information for price-calculation, etc. In practice, both are to a large extent based on identical data and in company practice the compiling of the accounts may be based on similar procedures and carried out by the same persons. "Tax accounting" is required for producing the accounting data that is needed for determining the statutory tax liability of a company. Thus, the basic accounting approach to identical transactions can be very different in different Member States. Transfer prices are a good example.

Independence means that income determination for accounting purposes is in principle independent from income determination for tax purposes. Companies may choose different accounting policies for tax and for financial accounting purposes and the use of special tax facilities is not linked to applying these facilities in the financial accounts. However, the "independence" is never total: Accounting rules will take into account structures which may have been set up for tax purposes and the other way round. The separation of tax accounts and financial accounts is somewhat typical for common-law countries applying an "over-riding principle" of a "true and fair view" in accounting law.

Dependence means that either the financial accounts follow the tax rules, or that income determination for tax purposes is determined by the choices made in financial accounts. This means that the computation of income for tax purposes has generally to follow financial accounting standards and the financial principles of proper bookkeeping

137.

The linkage between accounting and taxation can go even further by insisting that an accounting treatment available under tax law can only be exercised to the extent that the same treatment has been followed in the financial accounts

• 138. 

  The linkage between tax

accounts and financial accounts is somewhat typical for civil-law countries. The underlying idea is to prevent companies presenting different truths to different parties (i.e. a big profit is shown to shareholders and potential investors and a low profit or even a loss is presented to tax authorities), especially where companies have been able to benefit from special tax facilities.

Normally, dependence only exists in individual accounts, because the consolidated accounts are not usually the basis for the assessment of income tax. However, companies may choose to apply the same accounting policies in individual and in consolidated accounts. Consolidated accounts may therefore be influenced by tax rules in those countries where dependence is prevalent. Generally, however, parent companies tend to prepare consolidated accounts in a manner which takes out as much tax influence as possible.

Within the EU, independence between accounting and taxation is prevalent in Denmark, Ireland, the Netherlands and the United Kingdom, while dependence is prevalent in varying degrees in the remaining Member States, with the strongest link between accounting and taxation to be found in Germany. Given the enormous pressure to go to global accounting standards there has recently been a trend in many Member States to loosen this link but there are also recent developments in the opposite direction.

been seriously amended. The changes would have impacted on the tax revenues of many Member States. The basic approach in the 7

th directive on consolidated accounts141 was

similar: Member States have the option to reflect "tax coloration" in the financial accounts

• 142. 

  For instance, accelerated depreciation is accounted for in the parent-

company individual account but normal depreciation in the consolidated account.

Box 47:

How do multinational companies deal with differing requirements for financial accounts

in 15 Member States and beyond?

According to business representatives in the panel of experts, matching the tax accounts and financial accounts in a group of companies is just one practical problem among many in dealing with 15 tax authorities. Solving it however generates considerable compliance cost. ICI, for instance, applies solelyUK accounting standards. Local accounts are then adapted according to local rules. The biggest problem in this respect is not other EU countries (here overall differences are relatively limited) but the USA. AnUK accounting result of 150 may correspond to an US accounting result of +/- 110.

Unilever is a company whose headquarters are in two Member States and consequently faces both the NL and UK accounting system. Being listed on the NYSE, the US accounting arrangements also need to be taken into account, but Unilever still works with a single set of internal financial accounting rules. The financial data are used to compile the accounts according to UK standards. Apart from minor differences (e.g. treasury stock) which are accounted for separately, these accounts are used for both theUK and the Netherlands. For the USA, however, a complete restatement of the balance sheet and the profit&loss account is made.

New developments since the publication of the Ruding report

In the mid-90s, the Commission came forward a new accounting strategy which aimed at incorporating European harmonisation within a broader international harmonisation.

In its Communication on the issue

143, the Commission proposed to the Member States

to make it possible for "global players" to prepare their consolidated accounts in conformity with International Accounting Standards (IAS). In its general communication on accounting matters

144 and the proposal for a regulation on the

the biggest single capital market in the world. This market attracts many EU multinational companies seeking new equity. However, in the USA and in many other English-speaking countries preference is given to an investor-driven conceptual framework of financial accounting which emphasises potential future profits of the enterprise. The stakeholder tax authorities are neglected.

Similarly, the IAS have been drawn up in a manner which is neutral from a tax point of view. The application of those standards in individual accounts would, however, not normally be possible in most Member States as it could result in adverse tax consequences

146.

In the EU, the relation of tax accounting and financial accounting will be significantly influenced by the creation of an EU-wide capital market resulting from the set-up of Economic and Monetary Union. As long as different accounting rules are followed, the publication of financial statements by companies from the euro-zone in the same currency might give a wrong impression of harmonisation. Markets clearly exercise pressure to further harmonise the accounting rules. This trend will be reinforced by the imminent creation of pan-European stock-exchanges for the +/- 7000 stock-listed companies in the EU. There is good reason to believe that these developments will increase the pressure towards a more investor-driven accounting framework (with subsequent effects on tax accounting). However, it must be clearly said that the decision to separate commercial from tax accounting is no foregone conclusion and that the pros and cons are still being weighed.

The Council has recently adopted a directive147 allowing Member States to permit or

impose "fair value accounting" for certain financial assets and liabilities. This new approach will essentially remove the traditional realisation principle (i.e. profits are only accounted for when they are "realised" in a clear transaction). Instead of transactions simple value changes would constitute the accounting base; financial instruments are thus valued at market value (instead of historic cost). Thus, the linkage between financial accounts and tax accounting becomes virtually impossible.

148

In the literature, there is even a growing trend to apply "fair value accounting" to other assets and liabilities. Some scholars suggest that the traditional profit/loss-account does not reflect the true economic operation of the enterprise. Investment properties, for example, are held for investment purposes but are not accounted for at the market value. Investment properties value changes could be mixed with transaction-based income and results in the profit/loss-account or they could be separated but in any event then the question arises, how do you measure the profit - excluding or including the value changes? It is suggested that a new "statement of financial performance" is introduced to

address these problems. These views are of course not undisputed. In particular, reference is made to the volatility of market values, especially for stocks. Moreover, "fair value accounting" is said to influence what should be a purely financial business decision which is generally a negative development.

Conclusions and practical examples

Generally, it is clear that there is no prospect of fully matching tax and financial accounting in the future. To the contrary, "fair value accounting" and recent developments in capital market regulations constitute steps which are likely to alienate the two worlds further. "Fair value accounting" will most probably play a role for consolidated accounts of EU multinationals and there are good reasons to assume that the investor-driven IAS will become the basic standard for these consolidated accounts. Generally, the need for more harmonisation to ensure greater comparability of EU financial statement implies a severing of the link between tax and financial accounting, at least unless Member States were prepared to change drastically their tax accounting. Interestingly enough, however, recent developments in some Member States also show that the separation of financial and tax accounting does not necessarily alienate the two.IAS standards are sometimes also used for tax accounting purposes.

The effects of these developments are first and foremost felt in the consolidated accounts but, at some stage, they will also have consequences for the individual accounts (e.g. through the subsidiaries of multinational companies) which in most Member States also provide to a varying degree tax-driven accounting information. The dependence of financial accounts and tax accounts in many Member States will thus continue to be challenged as the strong linkage between tax and financial accounting makes it extremely difficult to change and modernise the accounting rules (because of the implicit tax consequences).

There are two conclusions to be drawn for the analysis of possible company tax obstacles in the Internal Market. First, inasmuch such obstacles are linked to the tax base, i.e., the various components forming the taxable income, it becomes clear that accounting harmonisation will not contribute towards approximation of company tax bases in the EU. Second, there is good reason to believe that the number of options, including tax options, will increase further in the future. This is because despite the above trend most Member States appear to wish to keep some sort of linkage between tax accounting and financial accounting.

Box 48:

The tax treatment of leasing contracts

The differing tax treatment of intangibles (namely goodwill, trademarks and know-how) is a major obstacle to cross-border business restructuring, in particular in situations in which intangibles are sold/acquired in the framework of an asset deal. In some Member States, tax depreciation of acquired intangibles is allowed, in others it is not. When accepted, different depreciation systems are provided in different Member States. Furthermore, in most Member States capital gains upon the disposal of intangible assets are taxable. As moving an intangible asset from a Member State where a capital gain on an intangible is taxed to one where it cannot be depreciated is sometimes prohibitively expensive, the difference in tax treatment is an issue that needs to be addressed.

Similarly, the tax treatment of leasing is not uniform across the EU, ranging from ordinary assets' depreciation rules to immediate full deduction of payments. Thus, cross-border problems arise essentially because of the differing qualification of contracts as financial leasing or operating leasing in the country of the lessor and the country of the lessee. Thus, localisation as well as financing decisions are affected. Companies may sometimes use the co-existence of differing rules of an identical leasing-situation for tax planning purposes (namely concerning the leasing of aeroplanes or of big plants and machinery in general) but generally leasing is a genuine cross-border tax problem. Leasing is becoming increasingly important for infrastructure and big high-tech equipment and differences in the tax treatment can be decisive for concluding or not a contract.

Another practical example of how financial accounting impacts on company tax issues is the treatment of mergers and acquisitions.

Box 49:

The influence of accounting rules on business decisions: mergers and acquisitions

The accounting treatment of mergers and acquisitions can be decisive for the tax treatment and financial decisions in this area. The choice between the "acquisition method" and the "pooling of interests method"

for business combinations is a good example. This is because the "pooling method" is accepted for tax purposes in some Member States whereas others allow it only in financial accounting. It may also be noted that some EU companies apparently even only list at the NYSE because this allows them to acquireUS companies and apply the "pooling method".

Under the "acquisition method" the subsidiary (i.e. the acquired company) is integrated into the consolidated accounts of the parent company (i.e. the acquiring company) as if the various assets and liabilities of the acquired company had been purchased individually. They are valued at 'fair market value' and the sum is confronted with the own equity according to the book value of the subsidiary. A possible difference is attributed to the various posts of the balance sheet or it is put as "goodwill" or "badwill" on the balance sheet. The cost of the merger are activated on the balance sheet.

Generally speaking, "pooling" is a merger of two companies in which the future parent company exchanges at least 90% of the shares of the subsidiary against (mostly) newly issued own shares. Under the "pooling of interests method" the participation at the subsidiary is valued only at its nominal value. In the consolidation the book-value of the participation of the parent is simply compensated against the equity of the subsidiary. All other assets and liabilities are valued at book value and simply added. A possible difference between the consideration and the book value of the participation changes the reserves. The cost of the merger can be immediately deducted.

As the "hidden reserves" are not made visible under the "pooling" method, it avoids subsequent taxation at market values. Interestingly enough, there appears to be a strong lobbying in the USA to abolish "pooling" domestically and to no longer accept it internationally.

The impending debate on the separation of tax accounting from financial accounting, to be conducted at the European level, presents an opportunity for further approximation of the tax bases of the Member States. To the extent that tax accounting will develop independently from financial accounting, Member States will be obliged to find autonomous rules for tax accounting purposes. In looking for such rules there is an opening for co-ordination and co-operation to start with common base rules, instead of each of the Member States trying to pursue individual solutions.

IV.B

Targeted actions in specific areas

MONITORING OF THE UNIFORM APPLICATION OF EU COMPANY TAX LAW

The analysis in Part III of this study has clearly shown that many of the tax obstacles concerning mergers and acquisitions and in the area of dividend taxation boil down to the differing implementation by Member States of the Merger Directive and the Parent- Subsidiary Directive. Such differences are of course to some extent intrinsic to the legal instrument of a "Directive" but the analysis has revealed a number of issues that merit further consideration.

While a limited degree of divergence of implementation and interpretation of a Directive by Member States in drafting their national laws is inevitable, significant differences in the conditions under which the benefits of the Directive are applicable give rise to problems. Moreover, the substantially different implementation of EU law in Member States increases further the compliance cost resulting from the existence of 15 tax systems within the Internal Market. A more uniform application of (existing and future) EU company tax law could be an important step in order to reduce these compliance costs and to make sure that where EU tax law is relevant comparable situations are treated in a comparable manner. At the same time, the need for litigation would be reduced.

One way of achieving these objectives and to tackle the various problems relating to the divergence of the application of (both existing and future) EU taxation Directives across Member States would be the introduction of some kind of collective monitoring of the implementation of Directives. This would involve the creation of a mechanism for the exchange of best practice and/or some form of peer review. After discussion with Member States, the Commission could issue guidance on the interpretation of important provisions of the Directives.

would be to follow the approach of the 6th VAT directive which explicitly provides for a

VAT Committee as an interpreting body151.

Given the importance of the underlying problems, "collective monitoring" appears to be a particularly appropriate means to deal with these sorts of company tax obstacles in the Internal Market. Some possible elements of such an exercise are considered in more detail below during the discussion of other targeted remedial measures.

REMEDIAL MEASURES IN THE AREA OF DIVIDEND TAXATION

The classical system vs. the imputation system

The analysis in Part III of the obstacles in the area of dividend taxation demonstrated that differences in the tax treatment of resident and non-resident shareholders as well as foreign and domestic investment need to be removed, as such differences entail an incentive towards domestic investment.

The classical system avoids this difference in treatment but leads - for both domestic and foreign shareholders - to economic double taxation of the company profit at company and shareholder level. There are, however, several possible ways to at least significantly mitigate the effects of this double taxation, e.g. the application of a reduced rate on dividend income or partial taxation of dividend income only. In any event, this form of double taxation is not a cross-border tax problem.

The imputation system eliminates, via an appropriate tax credit, the double taxation problem but usually only for resident shareholders. Most commentators maintain that this difference in treatment infringes the free right of establishment and the free movement of capital. Regardless of the future development of the jurisprudence in this respect, the best solution for the Internal Market would be, if imputation systems are maintained, for Member States to extend imputation tax credits also to, on the one hand, non-resident shareholders or, on the other hand, foreign income. Given that most Member States do not apply imputation systems this could, however, not be done on a strict reciprocity basis. Generally, imputation systems tend to become rather the exception than the rule in the EU.

systems which provide a disincentive to cross-border activity or investment may be contrary to the Treaty provisions on the fundamental freedoms. Such rulings raise important issues for the design of Member States' tax systems for which more guidance at EU level would be desirable.

Remedial measures concerning the Parent-Subsidiary Directive

Desirable changes to the Directive

Scope of the Directive

On the basis of the analysis in Part III, the Parent-Subsidiary Directive's double taxation provisions do not need to be radically recast. Though the Directive could be improved, the system for preventing double taxation of dividend payments between associated companies in different Member States works relatively well and enjoys widespread approval. Nevertheless, a number of other improvements would help to remedy the tax obstacles identified in Part III and would assist cross-border business restructuring operations. Given the importance of the underlying problems, it appears that measures to improve the effectiveness of the Directive should be pursued as a matter of urgency.

The Commission remains convinced of the merits of its 1993 proposal for a Directive encompassing all companies subject to corporation tax, whatever their legal form. It would help greatly if the Council were to resume discussions for the adoption of this proposal, which were interrupted in June 1997 to make way for the high-priority "tax package". Admittedly, a majority of Member States would currently rather see the list of company forms covered by the Directive updated than have the Directive extended to all companies subject to corporation tax. But even if this solution were adopted, there would still be a pressing need to resume discussions of the issue. In its meeting of 26- 27 November 2000, the ECOFIN Council cited updating the list as a priority. The agreement reached at the Nice European Council concerning the European Company Statute also entails updating the list, if only to include the new European Company (Societas Europeae).

The first would be to broaden the definition of the holdings taken into account when calculating the 25% threshold to include indirect holdings. This would solve the problems affecting the organisation and reorganisation of groups, but national tax administrations would then have to identify and calculate indirect holdings. The second would be to continue taking account of direct holdings alone but to lower the 25% threshold considerably. This would have the merit of not complicating the Directive's application by the national tax administrations while eliminating most of the difficulties involved in the organisation of groups. In the interests of simplicity, the latter solution seems preferable. It must be borne in mind that several Member States already apply the Parent-Subsidiary Directive with lower thresholds of 5 or 10% of the subsidiary's capital, in order to mitigate the double taxation of dividends received.

Methods for eliminating double taxation

The existence of two methods of eliminating double taxation (exemptions and tax credits) results in much complexity. Abolishing this optional system to leave just one method of avoiding double taxation would simplify the Directive's application within the Community. Many commentators argue that the Internal Market philosophy pleads for the exemption method, which is said to represent Capital Import Neutrality. If theEU were really, from a tax point of view, one Internal Market the CEN-CIN discussion as presented in Part II would become theoretically irrelevant, and turn into an internal debate on the regional allocation of new investments. In this sense, one might argue that the credit method creates an obstacle that is against the spirit of the freedom of establishment. The home state levies additional tax up to its own tax level, as a result of which companies operating in the state by means of a permanent establishment pay more tax than locally operating companies. From this standpoint, "capital import neutrality" as reflected in the exemption method should become the general rule within the Internal Market. Tax would then be levied at the rate of the permanent establishment state and a level playing field created for non-resident companies and local companies as far as taxation is concerned.

However, certain Member States almost always provide for the use of tax credits in their bilateral double taxation treaties, whereas others almost always provide for exemption. This will make it even harder to reach agreement on a common method. What is more, the equations "CEN = credit method" and "CIN = exemption method" seem to be oversimplified and do not correspond to the complex functioning of both methods, which in any event could both be improved.

relatively favourable reception to these particular provisions of the 1993 proposal. Renewed discussion of the proposal should also help settle this point.

Closer monitoring of the implementation of the Directive

The analysis in Part III suggests that the conformity of national implementing legislation needs to be checked in more depth, especially provisions based on the Directive's tax- evasion/avoidance clause. This review would (mainly but not only) focus on countries whose law deems the holding of shares in a Community company by non-Community shareholders to be a presumption of tax evasion or avoidance.

As mentioned in section III 3.3.1 above, in its judgment in Case C-28/95 (A. Leur- Bloem v Inspecteur der Belastingdienst/ Ondernemingen Amsterdam 2 ), the Court of Justice ruled that tax-evasion/avoidance has to be determined on a case-by-case basis. Though this case-law concerns the Merger Directive, the Court's analysis seems applicable to the Parent-Subsidiary Directive.

Firstly, setting aside the Directive where companies paying or receiving dividends are controlled by one or more non-Community companies, without allowing taxpayers to present evidence to the contrary, does not, at first sight, seem consistent with the Directive.

Furthermore, with regard to the creation of a presumption of tax evasion/avoidance, the

Leur-Bloem judgment does allow Member States to consider certain operations to constitute a presumption of evasion or avoidance as long as the taxpayer has the opportunity to present evidence to the contrary. However, a blanket presumption of evasion or avoidance seems excessive. Such a presumption would lead to a situation in which any shareholding in a Community company by non-residents constituted a presumption of evasion or avoidance. Accordingly, the mere fact of having non- Community shareholders cannot be sufficient grounds for a presumption of tax evasion or avoidance.

Remedial measures concerning the Merger Directive

Desirable changes to the Directive

Part III of this study set out a number of shortcomings in the existing Merger Directive. The Commission remains convinced of the merits of its 1993 proposal for a Directive encompassing all companies liable for corporation tax, whatever their legal form. Moreover, although it is not, in the Commission's view, legally imperative that the Directive be amended to cover the conversion of branches into subsidiaries the issue could usefully be clarified in the course of a future amendment of the Merger Directive.

To avoid the risk of double taxation, it would make sense to amend the Merger Directive so that, in the case of the disposal of shares received following a transfer of assets, the taxable capital gain was calculated on the basis of the real value of the shares received on the date of the transfer of assets. This would in particular prevent the taxation of capital gains from effectively precluding such operations. There would be no charge to tax on the disposal if the capital gains on securities received in exchange for a transfer of assets could be calculated on the basis of their cost price on the date of the transfer of assets rather than the book value of the assets transferred.

Likewise, in the case of exchanges of shares, the capital gains on the acquired company's shares received by the acquiring company should be calculated on the basis of the real value of the acquired company's shares on the date of the exchange.

The Directive could usefully be amended to oblige Member States - in the event of mergers, divisions or transfers of assets - to transfer the transferring company's unused losses to the company receiving the assets. In this case, it would need to be considered in more detail whether and to what extent Member States could, on the basis of Article 11(1)(a) of the Directive, refuse, on a case-by-case basis, the ability to transfer losses where the restructuring operation clearly is carried out in order to evade or avoid tax. The Court's judgement of 17 July 1997 on Leur-Bloem (Case C-28/95) provides some arguments in this respect.

ruling requires that "abuse" be determined on the basis of a case-by-case examination. A blanket refusal to apply the Directive to the disposal of shares received before a given period has elapsed, without granting the taxpayer an opportunity to show that the disposal is above board, would not appear to be consistent with the Directive. Nor do particularly long embargoes of five to seven years on the disposal of shares appear to be inherently compatible with the possibility of presuming tax evasion or avoidance when shares received are sold so long after the original transfer of assets or exchange of

shares.

Box 50:

Judgement of the Court of 17 July 1997 (Case C-28/95 Leur-Bloem)

Summary of the issue

Dutch legislation does not apply the Merger Directive's provisions on exchanges of shares where the acquiring company does not itself carry on business or where a natural person who is both the sole shareholder and director of the c