General Report

Pasquale Pistone

Introduction

Tax treaties have developed considerably throughout the twentieth century, based on the

assumption that states need an agreed legal instrument to coordinate the exercise of taxing

powers and thus minimize overlaps and the negative influence of taxation on cross-border

economic activities. In particular, the impetus for international tax coordination grew steadily

in the second half of the twentieth century under the auspices of international organizations,

such as, in particular, the OECD and the UN. The OECD and the UN continued the technical

activity undertaken in the framework of the League of Nations and drafted their Model Tax

Conventions, which are currently the main source of tax treaty clauses around the world. Such

Models are the outcome of technical activities carried out in working parties, which include

representatives from the tax authorities and from business, and are usually regarded as the set

of tax treaty rules that most consistently reflect the international tax policy of the Member

countries. Accordingly, the Models are generally regarded as the best available tax treaty

practice. This reputation has generally enhanced their implementation in bilateral tax treaties

over the past decades, showing that international tax law in fact shares a common substance to

a much greater extent than it may appear in the absence of a proper international customary

tax law. This structural peculiarity makes such Model Tax Conventions the soft source of

international tax rules, which then find their normative dimension in the bilateral tax treaties

that include them.

Over the past decades this structural peculiarity has gradually expanded to a global

dimension. The era of global law and worldwide free trade with decreasing tax barriers for

cross-border activities makes states very concerned about the need to have a global tax

system, which is able to compete with best practices and offer an attractive legal environment

to internationally mobile capital. This is even more the case since when harmful tax regimes

started being dismantled and global fiscal transparency picked up.

This process is beginning to take on features and substance of a multilateralization of tax

treaties, in particular when it comes to clauses that do not affect the allocation of taxing

powers, where the OECD has succeeded in achieving an effective worldwide consensus, as

recent developments on arbitration in taxation and, even more evidently, on global fiscal

transparency show.1

As the national reports indicate, the influence of the OECD Model Tax Convention (OECD

Model) on the general structure and clauses of bilateral tax treaties has gradually gained in

importance so that it now affects the ones concluded with or even between non-OECD

Member countries. Meanwhile, the overall influence of the UN Model Tax Convention (UN

Model) has gradually declined, with its residual role confined only to a limited number of

bilateral tax treaties, or to some specific clauses. In particular, UN Model clauses can still be

found in the treaties between some non-OECD countries which wish to preserve a stronger

taxing sovereignty of the state of source, or of some OECD countries, which maintain a

separate tax treaty policy for their relations with some net capital-importing countries, in

1 See further on this section VIII of this General Report and of the country reports, in particular on the standards for exchanging information set by Article 26 OECD Model.

order to foster the economic development of such countries.2 Several reasons exist for this

phenomenon, including in particular the stronger negotiating powers of OECD countries,

which pursue the defence of allocation rules patterned on the OECD tax treaty policy, but also

the more proactive attitude of the OECD, which is constantly refining its Commentaries

through the activity of its working parties.3 The following analysis is therefore in principle

centred on the influence of the OECD Model on bilateral tax treaties and occasionally

includes some specific remarks that are also relevant for the UN Model.

The tax literature has accompanied this development, by initiating a worldwide technical

debate on legal issues raised by the interpretation of tax treaties, focusing attention in

particular on the clauses included in the Models. By contrast, the technical debate did not

reach the same degree of intensity on two issues, which are central from the perspective of our

analysis. First, clauses not patterned along the schemes of the Models are often relegated to a

regional debate, or only analysed from the perspective of their consistency with the national

tax treaty practice of the states.4 Second, the determination of the actual boundaries of the

influence of (each of the) Models is still significantly underexposed in international tax law.

This often leads to an incorrect a priori general conclusion that tax treaties are what the

Models say they are. Despite the undoubted growing importance of the Models, one may not

ignore the still large grey areas where legal uncertainty rules and international tax planning is

often used as a solution to overcome the mismatches and overlaps that otherwise generate

economic distortions and double taxation.

The approach to such problems requires a detailed analysis of tax treaties based on a common

pattern5 in order to facilitate an immediate comparison among the various tax treaty options

available in the various countries. This chapter will guide the readers through the technical

maze of the bilateral tax treaty provisions analysed in this book and will highlight critical

issues, common points and differences contained in Models and bilateral tax treaties around

the world.6 Together with the opinions and technical information concerning the bilateral tax

treaties covered by the national reports included in the book, this General Report will include

the views of the author, also based on a comparison among the country reports, or of relevant

technical elements put forward by the tax literature.

The Articles of the OECD and UN Models constituted the starting point for the drafting

scheme of this book, which was then further refined by using the subdivisions that have

become of common use in the tax literature.7 More in particular, this resulted in a structure in

which all chapters are divided into eight sections, which focus on (i) the relevance of the

2 A very good example of this can be found in the shared allocation of taxing powers provided for by Article 12 on royalties,

which is perhaps the UN provision exercising the strongest influence on bilateral tax treaties. 3 Nevertheless, there are several clauses of the OECD Model that are either seldom or never included in bilateral tax treaties.

This is for instance the case of Arts. 9(2), 12(1), 24(2) and 27, as well as to a minor extent of the reference to the place of

incorporation as a tie-breaker rule under Arts. 4(3), 18 and 23A(4). Furthermore, peculiar issues arise as to Art. 14, which

was deleted in 2000 from the OECD Model and is instead often kept in post-2000 bilateral tax treaties, in several cases with the wording of the pre-2000 version of the OECD Model. This gives rise to technical uncertainties as to the scope of Arts. 7

and 14, as well as to the full relevance of the corresponding OECD Commentaries to such Arts., only one of which keeps

being updated. See further on this under Section III of this General Report. 4 The author believes instead that in several cases the international tax literature should reconsider the debate on such clauses, especially taking into account their very wide diffusion in bilateral tax treaties. For instance, this is the case of tax sparing

clauses, the entitlement of entities other than individuals and companies to treaty benefits, the tax treatment of cross-border

income from services, but also of the rules for relieving the double taxation in cases not covered by the Model Tax

Conventions (such as, for instance, economic double taxation), the use of most-favoured nation clauses, the boundaries of mutual agreement procedures and types of arbitration in bilateral tax treaties. 5 All chapters were drafted on the basis of a common questionnaire, included in the final section of the book. 6 Accordingly, all tax treaty clauses and specific areas not giving rise to critical issues or major discrepancies at the world

level will be hereby either superficially mentioned or omitted. 7 For this reason distributive rules contained in tax treaties were grouped under the two main umbrella provisions of Arts. 7

and 15, dealing with business income and income from employment, respectively. Likewise, rules contained in the final

chapter of the OECD and UN Model Tax Conventions have been separated in two chapters, focusing on the non-

discrimination principle and the provisions on mutual agreement procedures and international mutual assistance in tax matters.

OECD and UN Models and their Commentaries for the interpretation of tax treaties, (ii) the

personal and material scope of tax treaties, (iii) business profits and income from other

independent activities, (iv) passive income, (v) income from employment and other dependent

activities, (vi) methods for relieving double taxation, (vii) non-discrimination, (viii) mutual

agreement and international mutual assistance in tax matters.

I. The relevance of the OECD and UN Model Conventions and their

Commentaries for the interpretation of tax treaties

Before analysing whether tax treaty clauses contained in bi- and multilateral tax treaties

follow or depart from the Models, this section will preliminarily ascertain the relevance of

such Models for the interpretation of clauses that follow their structure and/or wording.

Although the differences in understanding the scope of Article 3(2) of the OECD and UN

Models are acknowledged, the section will not specifically address them.

The influence of the Models on the interpretation of tax treaties is perhaps one of the most

controversial issues in the international tax literature, leading to a significant volume of

judicial decisions around the world, and is based on the assumption that one of the Models

influenced the drafting of a bilateral treaty. Unless specified otherwise, we will assume that

the wording of tax treaty clauses matching those of either Model is the intentional outcome of

tax treaty negotiations that took the Model into account.8 Accordingly, the technically correct

interpretation of the Model should be relevant in order to determine the interpretation of such

bi- and multilateral treaties.

A good way to approach the influence of the Models on the interpretation of tax treaties is to

look at whether legal elements external to a tax system can affect the interpretation of its

rules. An IFA seminar analysed this issue in international taxation,9 comparing the influence

of Models with that of foreign court decisions on the interpretation of tax treaties. Enormous

differences could be seen around the world, ranging from countries that consider the

interpretation of OECD-like tax treaty clauses to be technically correct only when complying

with the criteria provided by the OECD, to the ones that instead regard the OECD

interpretation as one of the many relevant elements to be taken into account when interpreting

tax treaties.

The national reports confirm this, in particular as to the differences between the influence of

the Models and that of clarifications contained in their Commentaries, dynamic and static

interpretation of tax treaties,10

the value of reservations and observations, often including, on

a more nuanced basis, the positions held by the administrative and judicial practice. Some

general trends can be detected and are now briefly reported, together with some relevant

elements arising from the national reports.

Countries like Canada,11

the Netherlands12

and the United Kingdom (UK)13

give the OECD

Model and its Commentaries an almost binding value. Other countries, such as Australia,14

8 The author acknowledges that only in some countries is this explicitly acknowledged. However, the empirical approach of

this report to the problem suggests that even in the countries that do not specifically accept this, a perfect match with the Model clauses is almost never a mere coincidence. 9 IFA, 2008 Congress, Seminar F ‘The Use of Foreign Court Rulings for Tax Treaty Interpretation Purposes’, unpublished. 10 The importance of this dichotomy has exponentially grown since when the OECD has significantly increased the frequency

of amendments to the Commentaries on the Model Tax Convention. The author believes, however, that the key issue in such difference is not just finding out which one is technically more in line with the obligations undertaken by a State, since

technical arguments duly back up both theories. The issue is rather whether OECD Commentaries actually interpret or, by

contrast, revise the substance of clauses contained in the Model. 11 As the Canadian report indicates, in the Crown Forest Industries case the Supreme Court of Canada acknowledged the high persuasive value not only of the OECD Model, but also of its Commentaries, when interpreting the definition of

residence under the tax treaty with the US (see p. 152). More recently, the Federal Court of Appeal, in the Prévost Car case,

defined the OECD Model and its Commentaries as a widely-accepted guide to the notion of beneficial ownership in tax

treaties (see p. 152). Further Canadian decisions also compare the wording of the UN and OECD Models, to empirically show that in general the former does not have the same influence on Canadian tax treaties (see p. 153).

the Czech Republic and Germany,15

do so to a lesser extent, often regarding both rather as a

supplementary means of interpretation.16

Other countries, such as France,17

Italy18

and

Slovenia19

in particular, ascribe a more limited importance to the OECD interpretation, in

some judgments even considering it to have a similar value to that of authoritative tax

academics. This brief overview should also mention that in a growing number of countries,

including a likewise growing number of non-OECD countries such as Brazil, Colombia, Peru,

Russia and Serbia among the ones covered in this book, there is a general awareness of the

technical interpretation based on the (in particular that of the OECD) Model(s), which permits

their use among the arguments discussed in court, though without a decisive influence. A

stronger technical consideration of the OECD Model can instead be noted in a limited number

of non-OECD countries, such as in particular India. Besides such differences, the author sees

a growing trend to use the OECD Model as a vehicle to aggregate the rules of all tax treaties

around some consolidated and homogeneous legal standards, thus facilitating a voluntary

building-up of internationally accepted standards. This approach minimizes the relevance on

domestic law and mismatches in tax treaty interpretation, achieving in fact consistency across

bilateral tax treaties and it secures legal certainty while preventing interpretative disputes. The

author positively regards the fact that OECD Member States put reservations on the Articles

or observations on the interpretation given by the Commentaries when they do not agree with

the content, since this achieves some transparency as to how such states will carry out their

own tax treaty practice at the bilateral level.20

All such desirable results should, however, remain subject to an effective correspondence

between the text of bilateral tax treaties and the ones of the Models, as interpreted in their

Commentaries. National reports confirm that this problem is traditionally approached by the

tax literature within the dichotomy between dynamic and static interpretation. However, the

author believes that it should more properly presuppose the truly interpretative function of the

Commentaries, thus carving out all cases in which the Commentaries in fact do not clarify the

meaning of a provision contained in the Model, but rather make innovations to it. Although no

clear general trend can be regarded around the world, tax authorities are often keener in

supporting a dynamic interpretation, with a view to securing that tax treaties are always fine-

tuned to the development of international taxation and technically up to best-practice

12 Since 1992 the Netherlands Supreme Court considers the OECD Model and its Commentaries of significant relevance for

the interpretation of tax treaties based upon them, generally acknowledging the dynamic theory of interpretation. This view

was expanded in 2003 to treaties with non-OECD countries (Brazil and Nigeria), as well as a fortiori to more recent tax

treaties (such as the one concluded in 2006 with Barbados) that specifically indicate that both Contracting States are bound to follow the OECD Commentary. See further on this the Netherlands national report on p 153. The same type of clause is

included in the protocol to the tax treaty between Austria and New Zealand, as the New Zealand report indicates on p. 570. 13 The UK national report indicates that this is in particular the position of the Revenue on the OECD Model (reaching the

opposite conclusion for the UN Model and its Commentaries), whereby courts and academics have taken a more nuanced view on the matter, but affirming in general the reliance from 1984 onwards (see p. 882). The Hong Kong report suggests

that judicial decisions of common law countries, in particular of the UK, may also indirectly affect the position that Hong

Kong courts will take in the future on the matter (see p. 367). 14 As the Australian reporters indicate, in the Thiel case the High Court of Australia concluded that the OECD Model

provides a guide to the current usage of terms used by the parties in the tax treaty with Switzerland and, in the Lamesa Holdings case the Federal Court of Australia, reached a similar conclusion on the one with the Netherlands. 15 Nevertheless, this judicial view is criticised by tax literature, which supports a stronger relevance of the OECD Model and

its Commentaries, as the German report indicates. 16 For the signatory countries of the Vienna Convention on the Law of Treaties (VCLT) this means relevance within the

framework of the conditions listed in Art. 32 VCLT. 17 See further on this the report from France. 18 The Italian report also suggests that a different position has been put forward in Italian tax literature, more inclined to

support a technical value of the Models and their Commentaries, in particular of the OECD. 19 See further the report from Slovenia. 20 The author does not consider such reservations in their strict meaning under public international law, since the Models are

not international treaties, but mere non-binding models used to coordinate the exercise of tax treaty practice, whose clauses

only obtain their a proper legal and binding dimension once transposed in the actual treaties signed by the states. However, in

the author’s view, this should not prevent the use of the VCLT for interpreting the two-tier system of Models and bilateral treaties patterned on their clauses.

standards. However, some countries21

accept this theory only in general and, by contrast,

others22

are stronger supporters of the static interpretation. The foundations of the dynamic

interpretation are often criticized by the tax literature,23

based on various arguments, including

the need to make the interpretation correspond to the intended meaning of the Contracting

State at the time when the treaty was signed. However, considering the technical relevance of

the Model on the interpretation of tax treaties patterned along its clauses, the author feels that

later versions of the Commentaries cannot be completely ignored when interpreting clauses of

previously signed bilateral treaties, provided that they do not in fact change the plain wording

of the clause. The reports confirm that this corresponds to the current evolution worldwide.

Whether this relevance is equivalent or weaker to that of versions of the Commentaries

predating the bilateral tax treaty is a matter that depends on the way in which a given country

interprets tax treaties and how it applies the provisions of Arts. 31 and 32 of the VCLT to

them. From such a perspective, it also seems appropriate, however, not to ignore the fact that

even a clause with the same wording as the OECD Model can have different implications if

framed in a different context. This is often also a reason why the author suggests that, despite

the OECD Model and its Commentaries being the most important indicators of the

internationally accepted tax treaty standards, they should have a very limited indirect

influence on treaties that are not patterned on the schemes provided by the OECD.24

Observations on the Commentaries and reservations to the Model can raise more critical

issues in their practical application than in their general value and effects.

Observations on the Commentaries are used to reject an interpretation accepted by all other

Member countries. Accordingly, they are an instrument to narrow down in advance the scope

of interpretation of tax treaty clauses patterned on the Model. However, this should not

necessarily imply that in the absence of an observation, a state should be bound by the

interpretation contained in the Commentaries. Even in the countries that more closely follow

the technical indications of the Commentaries, other instruments can be used to achieve the

same result. Furthermore, a different conclusion would deprive non-Member countries of the

possibility of achieving a similar limitation in the scope of a clause included in a bilateral tax

treaty and patterned along the Model.25

Reservations on the Models generally achieve an equivalent result to the one they have in a

binding multilateral tax treaty, since they are almost always reflected in a different wording of

clauses (or in different clauses) included in the bilateral tax treaties of such a country. The

existence of a reservation, however, should also affect the interpretation of all other clauses

included in the treaty, which must comply with the different context created by such a

reservation. However, reservations to the Model have no legal value when the state includes

in its bilateral treaties clauses that correspond to the ones on which the reservation was made.

A separate issue arises for the position of non-Member countries that have been included as

annexes to the Commentaries on the OECD Model since 2000. Their function being that of

enhancing transparency in the interpretation and application of tax treaties, they are neither to

be equated with observations nor with reservations and have a mere informative function as to

the current tax treaty practice of a country.

21 See in particular the reports from the Netherlands, Peru and Spain. 22 See in particular the reports from Argentina, Italy and Liechtenstein. 23 See M. Lang, Introduction to the Law of Double Taxation Conventions, (Vienna: Linde, 2010), pp. 45 et seq.; M. Lang and F. Brugger, ‘The role of the OECD Commentary in tax treaty interpretation’, Australian Tax Forum 23, (2008), 95-108 (at

101 et seq.); M. Lang, ‘Die Bedeutung des Musterabkommens und des Kommentars des OECD-Steuerausschusses für die

Auslegung von Doppelbesteuerungsabkommen’ in W. Gassner, M. Lang and E. Lechner (eds.), Aktuelle Entwicklungen im

Internationalen Steuerrecht, (Vienna: Linde, 1994), pp. 11-41 (at pp. 24 et seq.); J. F. Avery Jones, ‘The Effect of Changes in the OECD Commentary after a Treaty is concluded’, 56 Bulletin for International Fiscal Documentation 3 (2002), 102-9 (at

103 et seq.). 24 However, the European Court of Justice (ECJ), when interpreting cases involving tax treaty issues, sometimes regards to

the OECD Model as almost an equivalent source to international customary law in tax treaty matters. 25 See further on this the UK report.

II. The personal and material scope of tax treaties

II.1 The personal scope

The OECD and UN Models do not present major differences as to their personal scope, which

is defined by Articles 1 and 4. Some old bilateral treaties still in force do not include an

equivalent clause to Article 1 OECD/UN Models,26

whereas others still follow the 1963

version of the OECD Model and do not provide for a limitation to liable-to-tax persons in

either contracting state.27

Save for the specific issues related to the United States (US) taxing

on the basis of citizenship and Russia distinguishing between citizenship and nationality,

similar criteria are used to determine the personal entitlement to treaty benefits, taking into

account the residence of individuals28

and basically using the same tie-breaker rules contained

in both Models.29

Perhaps, the most relevant difference between the UN and the OECD Models as to the

personal scope is the reference to the place of incorporation of companies, which is among the

criteria used by the UN Model to determine the residence of companies, but which can also be

relevant for the OECD Model under the other criteria of a similar nature to the place of

management. A rather high number of bilateral tax treaties around the world30

currently

include the place of incorporation, which, in several treaties, unlike in the UN Model, is also

used as a tie-breaker rule.31

Another relevant difference concerning the tie-breaker rules can

often be noted with respect to the fact that several treaties include mutual agreement

procedures for persons other than individuals32

and further treaties do not include reference to

the place of effective management at all.33

In some cases additional criteria are used to

determine the residence of companies.34

Several more recent bilateral treaties contain specific rules on the entitlement to treaty

benefits for entities other than companies and individuals, such as in the case of pension

funds,35

trusts,36

investment vehicles,37

partnerships,38

dormant inheritance39

and non-profit

organizations.40

26 See for instance the Brazil-Japan tax treaty and the reports from India and the Netherlands. 27 A classic example of this category is Serbia. However, see also the reports from Australia, Canada, the Netherlands, New

Zealand and the US. Belgian treaties with countries not levying taxes contain special clauses in this respect. Special clauses are also contained in the Hong Kong treaties. Lebanese and Ugandan treaties include the requirement but also consider it met

in case of fully exempted entities. 28 The reports from Canada and India illustrate the relevance of concept of ordinary residence under the tax treaties of such

countries. The Chinese report addresses special problems of residence under Chinese domestic law that also affect the personal entitlement to tax treaties. The French treaty with Monte Carlo contains a special presumption of residence. 29 The report from Australia indicates the more limited use of habitual abode as a tie-breaker rule. 30 See the reports from Argentina, Brazil, Canada, Chile, Croatia, Estonia, Finland, France, India, Lebanon, Liechtenstein,

Norway, Peru, Portugal, Russia, Serbia, Slovenia, Sweden and the US. The UKreport suggests that the increasing use of this criterion is not related to the importance of the UN Model, but rather is to be seen as a consequence of domestic law. 31 See the reports from Chile, Croatia, Russia and Spain. 32 See the reports from Argentina, Belgium, Brazil, Chile (which in some cases can even lead to denial of treaty benefits),

Croatia, Hong Kong, the Netherlands, Slovakia, Spain, Sweden, the United Kingdom (which indicates that mutual agreement procedure is gradually replacing the place of effective management as tie-breaker rules for persons other than individuals)

and the United States. 33 See the reports from Australia (which refers to domestic law instead), Brazil, Colombia (which indicates that the place of

effective management has no meaning in the domestic tax system), Estonia and Spain. There seems to be a diffused awareness among the national reports of the uncertainties related to the use of this criterion, which is losing importance in

bilateral treaties, or requires a further specification of its meaning. The latter option is followed by Croatia in its treaty with

Armenia or is linked to domestic law, as in the New Zealand treaties. For the purpose of countering such problems Estonia

uses it instead only within the framework of mutual agreement procedures. 34 Australian tax treaties refer to domestic law. German treaties include reference to the place of the statutory seat, which is in

substance equivalent to the place of registration under the Hong Kong treaty with Indonesia and some UK treaties. The

Hungarian treaty with Japan refers to the location of the head office. 35 See the reports from Argentina, Belgium, the Netherlands and Portugal. 36 See the reports from Hong Kong and Portugal.

Unlike in the UN and OECD Models, the personal entitlement to the benefits of several

bilateral treaties is often limited for the purpose of countering abusive practices. The US has

supported the introduction of the so-called limitation-of-benefits clauses, which restrict the

personal entitlement for residents of the contracting states in the absence of a sufficient

genuine link with their taxing jurisdiction.41

However, other types of clauses affecting the

personal entitlement to treaty benefits are included as well in tax treaties to counter abusive

practices.42

II.2 The material scope

The material scope of both Models has an almost identical wording and exercises a notable

impact on bilateral treaties around the world. Most bilateral treaties cover income and capital;

others narrow down their material scope to the taxes levied on income.43

This is generally the

case when taxes on capital are not levied in a state.44

Generally, separate treaties exist for

inheritance, estate and gift taxes. However, such treaties have been concluded to a much more

limited extent. Rarely does the material scope of tax treaties on income and capital cover

further levies, such as social security contributions45

or exit taxes,46

or otherwise expand the

national sovereignty.47

Further specific problems have occurred in the material scope of tax

treaties of some countries.48

Nevertheless, two main types of divergence from the pattern of the Models seem recurrent in

the national reports.

First, not all treaties in all countries include taxes levied at government levels other than the

central one.49

In some cases, this is the obvious consequence of the absence or secondary

nature of taxing powers at lower levels of government; in some others it is a tax treaty policy

decision,50

often related to the division of powers within a given country.51

Second, several treaties opt for omitting a definition of the material scope52

and go straight to

the list of taxes on income and, where included, on capital, following the approach envisaged

by the OECD Commentary to Article 2.53

III. Business profits and income from other independent activities

37 See the Liechtenstein treaty with Uruguay. 38 This type of clause is not necessarily needed in countries that treat partnerships as opaque for tax purposes. For clauses

specifically mentioning the entitlement of partnerships to treaty benefits, see the reports from France, Hong Kong;

Kazakhstan, Portugal, Russia, Slovakia and Sweden. 39 See the report from Liechtenstein. 40 See the reports from Liechtenstein and the Netherlands. Estonia makes the entitlement to treaty benefits contingent on the

submission of a certificate of residence. 41 However, LOB clauses have spread well beyond the US treaties. See further on this the reports from Colombia, the Czech

Republic, Estonia, Finland, France, Norway and Russia. 42 See the reports from Belgium, Chile, Colombia, the Czech Republic (which uses beneficial ownership as a general anti-

treaty shopping clause), Portugal and Spain. 43 Further treaties carve certain taxes out, such as in the case of Norway and Sweden for the wage tax paid by the enterprise. 44 However, Kazakhstan levies taxes on property and does not normally include them under the scope of its tax treaties. 45 This is a highly disputed issue in Brazilian tax treaties, but specific clauses have been included for this purpose elsewhere,

as for instance in the Russia-Slovenia tax treaty. 46 See the reports from Australia, Canada, France, the Netherlands and Sweden. 47 See the report from Norway on the continental shelf. 48 See, among other reports, the Chinese, Hungarian and Italian ones on the business tax, the Estonian report on the

introduction of the corporation tax levied at time of distribution of the dividend. 49 For countries only including taxes levied by the central level of government, see the reports from Argentina, Australia,

Brazil, Canada, Chile, Croatia, the Czech Republic, France, Hong Kong, the Netherlands, New Zealand and the US. 50 As, for instance, in the case of France. 51 A good example of this category is the US. 52 In such cases the bilateral tax treaty does not include clauses of the type provided for in Art. 2(1) and (2) of the Models.

See further on this, the reports from Australia, Brazil, Estonia, France, India, Serbia, Slovakia, Slovenia, Sweden and the UK. 53 Commentary to Art. 2, para. 6.1, OECD Model.

In tax treaties following the UN and OECD Models Article 7 is the distributive rule

immediately applicable to profits falling within the scope of a tax treaty, due to the provision

included in the last paragraph of this Article.54

No exception to this principle is reported in the

chapters of this book. The residual function of Article 7 makes more specific tax treaty

clauses prevail over it, as is generally the case for passive income. However, the analysis of

bilateral tax treaties notes two further significant exceptions, which show important deviations

from this principle otherwise regarded as the expression of a generally accepted international

tax treaty practice. First, India strongly pursues the conclusion of treaties including a specific

article for fees for technical services, which significantly reduces the scope of the article on

business profits and more strongly protects the taxing powers of the source state.55

This is also

achieved to a minor extent furthering other countries, which include the so-called Service PE

provision along the pattern provided for by Article 5(3)(b) UN Model, especially when linked

to short-term time thresholds.56

Second, Brazil pursues an equivalent result by allowing for a

concurrent application of the business profits Article with the other income Article, which

would never be possible under the UN and OECD Models (where the latter provision has a

residual function).57

Further techniques affect the scope of Article 7 at the level of allocation

of income in the relations between head office and permanent establishment or between

associated enterprises.58

The predominant influence of the OECD Model on bilateral tax

treaties around the world finds some consistent exceptions, usually in relations with, or

between, non-OECD countries, which will be seen together with the additional types of

variations reflecting the tax treaty policy of single countries.

For the purpose of analysing more precisely the issues related to business profits, this section

will be divided into five sub-sections, which focus on i) permanent establishments; ii)

business profits under Article 7; iii) income from transport, iv) business profits in relations

between associated enterprises and v) income from independent personal services.59

III.1 Permanent establishment

Permanent establishment is a legal creation of tax law, aimed at isolating a stronger

connection of business income with a territory other than that where the main establishment is

situated and at determining a different way of exercising taxing powers on the corresponding

profits. The PE concept prospered in tax treaties, which were its true legal carriers, and was

then introduced in domestic law of virtually all countries around the world, including the ones

having a territorial system. Various national reports indicate the existence of different national

and tax treaty concepts of permanent establishment, pointing out difficulties at the level of

interpretation and application that may result from this combination. The author believes that

the solutions to such problems must comply with international standards, since this would

best remove the obstacles to cross-border activities. In some areas of the world, particularly in

the relations with developing countries, the permanent establishment concept has been

traditionally regarded as a possible instrument to strengthen the taxing powers of the source

country. Accordingly, the national reports indicate a consistent trend of treaties with

developing countries toward including rules, mainly patterned on the schemes provided by the

UN, which facilitate the existence of permanent establishments and the attribution of income

54 Article 7(6) UN Model and Article 7(4) (of the 2010 version, or 7(7) of the 2008 version of the) OECD Model. 55 See further in the report from India. 56 An example of this type of clause can be found in the report from Chile. 57 See further the report from Brazil. A concurrent application of the business profits clause with the residual one can also be found in the Germany-Luxembourg treaty, as indicated by the report from Germany. 58 Examples of this technique have been reported in the reports from Brazil and China, concerning deemed profit allocation,

in particular affecting transfer pricing. 59 Articles 16 (Directors’ Fees) and 17 (Artistes and Sportsmen) will be addressed in Section IV of the General Report, but have been usually included in this section of the national reports.

to them. We will focus our attention on a selection of the clauses of this type that have been

most frequently reported in the country chapters of this book.

Article 5(3)(a) UN Model provides for a broader notion of construction PE,60

combined with a

shorter term for its existence (six months), if compared to that of (twelve months contained

in) Article 5(3) OECD Model. This clause can be found very frequently in treaties between

non-OECD countries,61

but also in some OECD countries62

more frequently than in others, in

particular in their treaties with non-OECD countries. In other cases, the Construction PE

provision is patterned along the OECD scheme, but applied with a shorter time limit.63

Some

reports also mention the existence of longer Construction PE time limits, indicating that they

were designed in order to attract international capital.64

Interestingly, some countries include

ad hoc measures to counter avoidance of the time thresholds on Construction PEs.65

Article 5(3)(b) UN MC regulates the so-called Services PE, according to which a permanent

establishment may exist insofar as this type of activity continues for a period or period of

more than six months within any twelve-month period in the aggregate. This type of clause is

rather frequent in bilateral tax treaties,66

also among OECD countries.67

Further treaty clauses, such as the ones in Article 5(6) UN Model on insurance companies,68

and Article 5(7) UN Model on the independent agent not dealing at arm’s length69

(or

equivalent clauses with a different drafting)70

go in the same direction, but are to be found in

bilateral tax treaties with a somewhat lower frequency.

Finally, various country reports indicate that the boundaries of permanent establishment are

flexibly determined with a view to reaching out for the peculiar problems arising in bilateral

relations.71

Some clauses, however, can be found in more cases than others. The clearest

example of this type of provision is the PE clause concerning the exploitation of natural

resources.72

III.2 Business profits73

60 The UN construction PE notion also includes assembly or...supervisory activities in connection with the building site if lasting more than six months. 61 See the reports from Hong Kong, India, Lebanon and Uganda. 62 See the reports from Australia, Austria, Chile (also with other OECD countries), Estonia, Finland, Italy, Norway, Slovenia

and Sweden. 63 See the reports from Brazil, Colombia, Germany, New Zealand and Portugal. 64 See the reports from Hungary, Poland and Russia. 65 See the reports from New Zealand (including an ad hoc clause in New Zealand tax treaties) and Spain (for the single-

project interpretation by the courts). Under the Nordic Convention more enterprises may be pooled up for the purpose of the PE threshold, as the report from Finland indicates. A substantially equivalent result was achieved in the Philip Morris case

with the multiple permanent establishment, as the Italian report indicates. 66 See the reports from Argentina (all treaties from 1990), Colombia, Hong Kong, India (also with a one-day service PE

threshold in the tax treaty with Australia), Kazakhstan, Russia, Serbia and Uganda. 67 See the reports from Austria, Chile (with a deemed existence rule driven by a 183-days rule not excluding its possible

existence in shorter periods of time for recurrent activities), Czech Republic, Hungary, the Netherlands (in tax treaties with

developing countries), Portugal and Slovenia. 68 See the reports from Argentina, Austria, Chile, Colombia, the Czech Republic, Germany, Hungary, India, Lebanon, Liechtenstein, the Netherlands, Portugal, Romania and Slovakia. 69 See the reports from Argentina, the Czech Republic, Hong Kong, India, the Netherlands, Portugal, Slovenia and Spain. 70 See the report from Australia. 71 An interesting type of PE clause can be found in Australia (see reservation in para. 86 of the 2010 version of the OECD Commentary to Article 7) and New Zealand treaties (in compliance with the reservation indicated in para. 76 of the 2008

version of the OECD Commentary to Article 7, still applicable due to the reservation contained in para. 95 of the 2010

version of the OECD Commentary to Article 7) on taxation of business profits to which residents are beneficially entitled

from a trustee or a trust estate. See further on this the reports from Australia and New Zealand. 72 See the report from Argentina (also referring to fishing vessels), Austria, Estonia, Germany (recently enlarged, in order to

cover also renewable energies), Kazakhstan, New Zealand, Norway (whose treaties also apply to the continental shelf),

Portugal, Serbia, Spain and the UK. 73 Since this book puts the emphasis on the relation between bilateral treaties and the Models, national reports include only limited information on the 2010 OECD Model version of Art. 7 and will therefore not be addressed in this chapter.

The diffusion of Article 7 OECD Model (in its pre-2010 version) is fairly high in bilateral

treaties from around the world. However, two exceptions based on the UN Model are rather

frequent, mainly in bilateral relations with, or between, non-OECD countries.

The first and most notable one is what is known as the limited force of attraction, according to

which, in the presence of a permanent establishment, the country where the latter is situated is

allowed to tax not only income that is specifically attributable to such a permanent

establishment, but also more in general all income sourced there.74

This clause is the rule in

some non-OECD countries,75

where, especially in the case of non-sophisticated tax

administrations, it can also play an important role in streamlining the taxation of profits of

non-resident enterprises. Accordingly, it can also be seen in several treaties of OECD

countries with non-OECD (especially developing countries)76

and, exceptionally, even in a

relation between two OECD countries.77

The second exception is the prohibition on deducting payments in the head office-PE relations

on services (other than in the banking sector), following the pattern of Article 7(3) UN Model.

Nevertheless, this provision is to be found in a more limited number of treaties, usually

involving at least one non-OECD country.78

Further deviations from the patterns of both Models exist, but are more related to specific

bilateral needs.79

III.3 Income from transport

Although the UN Model provides for two alternative clauses on income from shipping, inland

waterways transport and air transport, one of them matches the clause contained in Article 8

of the OECD Model. Such clauses are most frequently seen in the bilateral treaties around the

world. However, there are treaties following the alternative B of Article 8 UN Model,

according to which exclusive taxation in the country of the place of effective management can

be replaced by a mechanism of shared taxing rights based on a fixed percentage.80

National

reports also mention a third option, giving primary taxing rights to the state of residence of the

enterprise, as under the US Model,81

or to the enterprise in general.82

Another recurrent deviation from the pattern above concerns the scope of such a clause, which

in several treaties, especially those of European countries,83

include all four types of transport,

i.e. by road, rail, air and sea, often including special provisions on containers as well.84

III.4 Associated enterprises

74 The author believes that the presence of this clause in a bilateral treaty requires the interpretation of Art. 7 to follow the

rationale of the UN Model for the attribution of taxing powers even where the remaining wording of the clause follows the

pattern of the OECD. This, rather frequent, mixture of clauses may increase the incidence of inconsistencies in interpretation and application. 75 See the reports from Argentina, Colombia, Kazakhstan and Peru. It is also present in several treaties from Romania and

Russia. 76 See the reports from Australia (especially in cases of tax avoidance), Canada, Chile, the Czech Republic, Estonia, France, Germany, Hungary, Italy, New Zealand, Poland, Portugal, Slovakia and the UK. 77 See the report from France. 78 See the reports from Australia, Austria, Brazil, Canada, Croatia, the Czech Republic, France, Germany, Hong Kong,

Hungary, Portugal, Romania, Russia, Serbia, Slovakia, Spain and Uganda. 79 An example of this type of deviation can be noted for silent partnerships. See further on this the reports from Austria and

Slovenia. 80 See the reports from Australia, Chile, the Czech Republic, the Netherlands and Poland. 81 See the reports from Estonia, Germany, Hong Kong, Lebanon, Portugal, Sweden and the UK. 82 See the reports from New Zealand and Portugal. 83 See the reports from the Czech Republic, Hong Kong, Liechtenstein, the Netherlands, Poland, Romania, Russia and Serbia.

Treaties having Art. 8 with a broader scope usually also ensure consistency with the treatment of income from transport in

other relevant provisions of the bilateral treaties. 84 See the report from Russia.

The clause on associated enterprises plays a particularly important role within the Model for

being the only one aimed at countering problems of economic double taxation. From this

perspective, the presence of a secondary adjustment under Article 9(2) was the outcome of

activities of OECD and UN aimed at strengthening the approach to this type of problem.

Accordingly, one would expect an adequate presence of this clause in bilateral treaties. This

expectation is, however, wrongly placed, since the presence of Article 9(2) is more the

exception than the rule for a large number of bilateral tax treaties. Even more difficult is to

find clauses patterned on Article 9(3) UN Model within bilateral tax treaties, which introduce

restrictions on Article 9(2) in cases of fraud, gross negligence, or wilful default.85

III.5 Independent personal services

The taxation of income from independent personal services shows a basic difference in the

approach of the OECD and UN Models from the moment in which the former, in the year

2000, decided to eliminate Article 14 becasue significant differences with business profits no

longer existed from a tax treaty perspective. From a theoretical perspective, the author

welcomes this change, which contributes to streamlining tax treaty allocation rules, removing

an unnecessary duplication that often creates difficulties in the dividing line between the

scope of Articles 7 and 14. However, if one looks at the impact of this reform on bilateral

treaties, a completely different pattern seems to arise. Ten years after the elimination of

Article 14 from the OECD Model, a separate provision on independent personal services is

generally still the rule,86

or is at least frequent87

in bilateral tax treaties concluded by various

countries after 2000. The national reports indicate that states have generally opted either to

follow Article 14 UN Model,88

or to keep their treaties aligned with the last available wording

of Article 14 OECD Model.89

Interpretative difficulties occur in the latter situation, since the

deliberate intention of parties is to follow a tax treaty clause of the Model that no longer exists

and that is meant to be equated to Article 7, which instead the contracting states opted not to

apply to regulate the exercise of their taxing powers on income from independent personal

services.

IV. Passive income

Distributive rules on dividends, interest, royalties and capital gains are often grouped together

under the general label of passive income concerning the homogeneous cluster of the

economic returns from a capital investment. The UN and OECD Models achieve consistency

in the allocation of taxing powers among such a group of rules along different schemes, which

give a stronger protection to the source and residence countries, respectively. The consistency

of rules in bilateral tax treaties generally shows a technical syncretism between the two

Models, which is more intricate than what can be perceived in respect of other groups of

distributive rules. Despite the general predominance of the OECD Model as mould for

bilateral tax treaties around the world, the author perceives that rules on passive income often

do not entirely follow the latest OECD standards even in the very relations between OECD

countries.90

This is in particular true for the taxation of royalties.

85 These clauses are recorded only in the reports from Argentina, Australia, Belgium, Chile, Colombia, the Czech Republic,

Hong Kong, Kazakhstan, Peru, Portugal, Romania, Serbia, 86 See the reports from Argentina, Austria, Brazil, Chile (narrowed down to individuals), Croatia, Estonia (narrowed down to

individuals), Poland, Portugal, Romania, Serbia and Spain. 87 See the reports from Australia, the Czech Republic, Hungary, Lebanon, the Netherlands, Russia and Slovenia. 88 See the reports from Germany, Hong Kong, Peru and Slovakia. 89 See the report from Italy. 90 Considering this general trend, various national reports from non-OECD countries show a closer full compliance with the UN Model in respect of this category of distributive rules.

The current section will now give more precise evidence of this perception with reference to

each tax treaty clause on passive income, following the numbering of tax treaty articles, i.e. i)

dividends; ii) interest; iii) royalties91

and iv) capital gains.

IV.1 Dividends

The scope of the dividends Article is homogeneously defined in Article 10(1) and 10(3) of

both the UN and the OECD Models. However, bilateral treaties, also when following the

general definition92

and relevant features of dividend payments93

eligible to the benefits of

this Article, frequently show the need for carve-outs94

and specifications95

which affect the

scope of this provision and its relations to the other ones governing passive income.96

The shared allocation of rights, capped in the state of source, is rather common in bilateral tax

treaties around the world.97

As the fixed withholding rates in the OECD Model are indicated

as the maximum amounts that countries should use,98

this report considers that a true

difference with the flexible rates of the UN Model only occurs in bilateral treaties that exceed

the rates indicated (either) in Articles 10(2)(a) and (or) in 10(2)(b).99

Some bilateral treaties

apply voting rights to differentiate between the two withholding tax rates,100

whereby other

treaties omit the difference between the latter two provisions and apply a single withholding

tax rate instead.101

Furthermore, a growing number of bilateral tax treaties apply zero rating in

the state of source, in particular to intercompany dividends.102

The latter phenomenon may be

related to the new policy of intercompany dividends in some major areas of the world,

including the countries of the European Union (which apply it under the Parent-Subsidiary

Directive)103

and the US. The EU relief on intercompany dividends may also be seen as the

possible reason for relieving economic double taxation in some bilateral treaties,104

whereby

the US Model is likely to be the reason for including branch profits tax clauses in other

91 This cluster will also include some special tax treaty clauses that can be found in bilateral treaties of some countries, such

as the ones on technical services or on commissions. 92 Some treaties do not include a definition of dividends. See further on this the reports from Australia and Poland. Others

achieve the same results by leaving the definition to be completed by domestic law, which in fact implies the use of Art. 3(2) and thus that no full treaty definition exists. See in this case the reports from Belgium, Canada, Germany, India and Poland. 93 See for instance the reports from India (whose treaties in some cases also include deemed dividends), Liechtenstein, the

Netherlands and Portugal. 94 Several countries carve out a specific reference to jouissance rights. See further on this the reports from India and Serbia. Others omit a specific reference to them. See further on this the reports from Australia, New Zealand and the UK. 95 The reports from France and the UK indicate a trend to provide a streamlined definition of dividends if compared to that

contained in the Models. The reports from Germany, Russia and the UK indicate that interest recharacterized as dividends

specifically falls under Art. 10 of several treaties. In Germany this also applies to income from silent partnerships. Some Portuguese tax treaties exclude some dividends from the scope of the Article or have further limitations. The latter case is

also common to some Romanian treaties. Further treaties include a specific reference to partnerships (see the reports from

Australia, Austria and Hong Kong), investment funds (see the report from Russia) and pension funds (see the report from

Belgium). 96 See for instance in the UK report, where the broader drafting of the scope of the dividend article has determined a residual

function of the interest article in the more recent tax treaties. 97 However, the opposite principle (i.e. the exclusive allocation of taxing rights to the state of source) applies under the

Andean Model (see further on this the reports from Argentina and Chile). 98 See OECD Commentary to Art. 10, para. 13. 99 This situation is not uncommon, especially in OECD treaties with or non-OECD countries or in the ones between them.

See further on this the reports from Australia, Austria, Canada, Chile, Peru, Russia and Uganda. 100 See the reports from Australia and Sweden. 101 This approach is perfectly compatible with the UN dividend tax treaty clause. See further on this the reports from

Argentina, Brazil, Canada, Chile, Colombia, the Czech Republic, New Zealand, Portugal and Serbia. 102 See further on this the reports from Austria, Belgium, Estonia, Finland, Hong Kong, Hungary, India, the Netherlands,

Norway, Romania, Spain, Sweden, the UK and the US. 103 Some countries like Belgium, however, also report zero-rating in the relations with non-EU countries. 104 See further on this the reports from Estonia, Italy, Portugal and Spain. Relief for economic double taxation in respect of

intercompany dividends is also provided by tax treaties in other areas of the world, such as Hong Kong, India, Kazakhstan,

Russia and Singapore. Also Chinese tax treaties provide for it, though for other reasons, mainly related to the fact that juridical double taxation is already relieved through domestic law provisions.

bilateral tax treaties.105

Further specific clauses are included in bilateral tax treaties in order to

pursue specific tax treaty needs of single countries,106

or are the outcome of certain single

package deals.107

Finally, there seems to be a growing consensus for including ad hoc anti-abuse clauses108

within the tax treaty clauses on this and other clauses on passive income in bilateral tax

treaties (even if some treaties are still lacking a beneficial ownership clause)109

and, to a

minor extent, in limiting the use of mutual agreement procedures on the mode of application

of such distributive rules.110

IV.2 Interest

Despite the specifications added to the scope of the interest clause in the 1977 OECD Model,

the drafting of bilateral treaties of some countries still reflects that of the 1963 OECD Model,

or otherwise has a more general scope.111

The allocation of taxing powers is usually shared along the lines of both Models,112

though a

general zero-rating by the state of source is no longer infrequent in bilateral tax treaties113

and

some of them provide for other types of allocation.114

Frequently, ad hoc carve-outs grant

source-state exemption115

or different rates116

to specific categories of interest, such as more

frequently on government bonds and on loans between banks and financial institutions, but

also on long-term loans.117

Furthermore, several treaties include specific rules to achieve certainty on the identification of

the source of interest,118

to counter abusive practices,119

to accommodate specific problems of

the contracting states,120

or still include features of previous versions of the Models.121

IV.3 Royalties

105 See further on this the reports from Argentina, Austria, Canada, Estonia, India, Italy, Kazakhstan, Romania and the US. 106 See for instance the report from Chile for what is known as the Chile clause, the report from France for the direct

application of the other income Article for dividends not falling under Art. 7 and that from Germany for the suspension clause. 107 This is for instance the case of tax treaty clauses including the right to reduce the exercise of taxing rights by the state of

source under the MFN treatment (see the reports from Argentina, Brazil and Colombia). A peculiar clause gives a one-way-

only treaty protection to dividends under the Bulgaria-Finland tax treaty (see further on this the report from Finland). 108 See further on this the reports from Brazil, Canada, Estonia, France, the Netherlands, Romania and the UK. 109 See the reports from Finland, India and Romania. By contrast, in some treaties beneficial ownership is a condition for the

application of the Article, rather than for the application of its reduced rates in the state of source (see on this in the report

from Australia). 110 See further on this (also in respect of other distributive clauses on passive income) the reports from Austria, Chile,

Colombia, Estonia, India, Italy, Serbia and Sweden. 111 See further on this the reports from Australia, Belgium, Brazil, Germany, New Zealand, Portugal and the UK. 112 As indicated earlier in this General Report under the dividend Article, the difference between the two Models only arises insofar as the levying of withholding tax rates exceeds the maximum amount indicated in the OECD Model. 113 See the reports from Belgium, Canada, Finland, Germany, Hong Kong, Hungary, Italy, the Netherlands, Norway, Poland,

Russia, Serbia, Sweden and the UK (in older treaties with subject-to-tax clauses). 114 See the reports from Argentina and Chile (on the Andean treaty, which provides for exclusive taxation in the state of the payor) and the equivalent rule reported from Poland on the treaty with Pakistan. Furthermore, an MFN treatment is applicable

as the outcome of specific negotiation deals. See further on this the reports from Argentina, Colombia and Estonia. 115 See the reports from Argentina, Australia, Austria, Brazil, Canada, Chile, the Czech Republic, Hong Kong, Hungary,

Italy, Lebanon, New Zealand (this carve-out is usually based on a dedicated paragraph within Art. 11), Norway, Portugal, Romania, Russia and Slovenia. 116 See the reports from Argentina, Colombia, Germany, India, Norway and Portugal. 117 See the report from Brazil. 118 See the reports from Argentina, Australia and the Netherlands. 119 See the reports from Australia, Brazil, Canada, Italy, Romania, Slovenia and the UK. 120 See for instance the report from Brazil (on the clause in the treaty with Japan). 121 Some tax treaties (see the reports from India and Romania) still do not include beneficial ownership clauses (the

Australian treaties, by contrast, generally include such a clause in paragraph 1 of the interest Article). Other treaties follow a limited-force-of-attraction principle (see on this the reports from Australia and Italy).

Bilateral tax treaties around the world undoubtedly show that the influence of the OECD

Model royalties clause is more the exception than the rule.122

The UN royalties clause is

instead the main point of reference for bilateral tax treaty clauses on royalties, which often are

accompanied by additional dedicated provisions.123

The consistency with the UN Model goes well beyond the shared allocation of rights, also

including the source of royalties,124

the scope of the clause, which, often includes ‘film or

tapes used for radio or television broadcasting’,125

‘the use or the right to use, industrial ,

commercial or scientific equipment’,126

or including a specific reference to software.127

In

further cases the provisions differentiate between literary and industrial royalties, also for the

purpose of determining the applicable rate.128

Furthermore, there is a growing trend to include anti-abuse provisions in the royalties clause

of bilateral tax treaties129

and additional clauses are the outcome of package negotiations.130

IV.4 Capital gains

Bilateral tax treaties present a slightly more homogeneous pattern in respect of the taxation of

capital gains with both Models. Besides some clauses to address some specific bilateral

problems,131

the main differences are to be noted as regarding i) the allocation of taxing

powers on the residual clause and its scope, ii) the presence and drafting rules on the

alienation of shares in companies mainly holding immovable property and iii) the specific

liaison with exit and trailing taxes applicable under the domestic tax law of either contracting

state.

In particular, several bilateral tax treaties show a shared allocation of taxing powers under

their residual scope as under the UN Model,132

whereas almost all others provide for an

exclusive allocation to the country of residence of the alienator.133

Furthermore, in several

treaties Article 13(4) on the alienation of shares in companies mainly holding immovable

property follows the shared allocation rules as under the UN Model134

and a growing number

of specific clauses are included in bilateral tax treaties135

for the purpose of countering

122 Following the OECD Models, some countries mainly include bilateral tax treaties with exclusive taxation of royalties in

the country of residence of the recipient. See further on this in particular the reports from Hungary, Sweden and the UK. By

contrast, the Andean Model provides for exclusive taxation in the country of use of the royalties (see on this the report from

Argentina). 123 Indian tax treaties usually include a separate clause on fees for technical services, which more closely reflects the tax

treaty policy of that country concerning the narrowing down of the business profits provision. Technical services are

otherwise frequently included in the royalties Article (in particular with developing countries: see the reports from Colombia,

Finland, France, Germany, Slovenia and Spain). Furthermore, Romanian tax treaties usually provide for a clause on commissions, possibly due to reasons related to domestic law. 124 This provision reflects Art. 12(5) UN Model and is frequently included in bilateral tax treaties to regard the payor of the

royalties as their actual geographical source for the purpose of applying the distributive rule of the bilateral tax treaty. See

further on this the reports from Argentina, Australia, Brazil and Canada. Italy also includes a provision on the source of royalties in tax treaties with other EU Member States. 125 See the reports from Australia, Austria, Brazil, France and Portugal. 126 See the reports from Australia, Austria, China, Finland, France, Germany, Hong Kong, Italy, Kazakhstan, New Zealand,

Poland, Portugal, Romania, Russia, Slovakia, Slovenia and Spain. 127 See the reports from Canada, Kazakhstan, Slovakia and Spain. 128 See the reports from Argentina, Austria, Chile, the Czech Republic and Portugal. 129 See the reports from Austria, Canada, Chile, Estonia, Hong Kong, New Zealand, Romania, Russia and the UK. 130 See for instance the MFN treatment clauses in the reports from Argentina and Brazil. 131 The report from Argentina indicates the specific MFN treatment clause applicable. Specific issues can be found in the

reports from Australia (where capital gains were taxed only at a later moment) and New Zealand (which does not include

capital gains taxation yet and therefore includes tax treaty clauses with a different wording). Furthermore, tax treaties from

Poland and Slovakia often include alienation of trucks within the scope of Art. 13(3). 132 See the reports from Argentina, Brazil, France, Kazakhstan, Peru, Portugal, Russia, Serbia and Uganda. 133 An exception is the Andean Treaty, which gives exclusive taxing powers to the state of the assets (see on this the report

from Argentina). 134 See the reports from Australia, Austria, Brazil, Germany, Hong Kong, India, Kazakhstan, Peru, Portugal and Romania. 135 See the reports from Austria, Canada, Finland, the Netherlands, Norway, Portugal, Russia, Spain, Sweden and Uganda.

abusive practices,136

or preserving the sovereignty of the state of emigration of the alienator

(though for a number of years after the emigration) and allow for the application of exit and

trailing taxes,137

more frequently in respect of individuals who are nationals of the country of

emigration.

V. Income from employment and other dependent activities

This section completes the analysis of tax treaty clauses of bilateral tax treaties (Articles 15

OECD and UN Models) that regulate the exercise of taxing powers on income from

employment. It includes all further clauses that directly interact with Article 15, namely

Articles 16, 17, 18, 19 and 20.138

For the purpose of achieving a systematic framework of the

relations with all distributive clauses, Article 21, on other income, has also been included in

this section. Preliminarily, the point should be made that certain countries (such as, for

instance, the Netherlands) conceive Article 15 as the gateway to Articles 16-19, thus

excluding in principle that the failure to apply either of them could open the way to the

residual clause of Article 21. Accordingly, the failure to apply either lex specialis would

either facilitate the application of another lex specialis within the same cluster of provisions,

or to its lex generalis, i.e. Article 15. However, this view is not shared by other countries,139

where the relations among such provisions is articulated according to a different pattern,

which can even conceive of a concurrent application of Article 21 with other distributive

rules.140

V.1 Income from employment

Despite Article 15 being almost identically drafted in the OECD and UN Models, the national

reports present a fairly diversified number of clauses, designed to fit the peculiar needs that

arise concretely. Generally, it seems as if most countries closely follow the general pattern

provided by the Models, adapting their bilateral treaties to the latest available standard. This

seems evident if one looks at the way in which the short-term employment carve-out clause of

183 days in Article 15(2)(a) is seen to occur in the bilateral treaties.141

Rules on income from employment in bilateral treaties are often more specifically drafted

than the ones contained in the Models, and take into account the influence of domestic law as

well.142

Four examples can be given. First, the provision of Article 15(2)(c) is in some cases

linked to the place where the tax on labour is deductible,143

rather than, as in the Models, to

where is borne. Second, various countries have adopted provisions on the international hiring

out of labour that also allow determining an appropriate notion of employer for tax treaty

purposes.144

Third, some countries extend the provision of Article 15(3) to truck and railway

136 See the reports from Hong Kong, Italy, Romania and Spain. 137 Following the general terminology in use in international taxation, exit taxes are levied on the accrued value of gains upon

emigration, whereby trailing taxes are levied also on the post-emigration value of gains for a number of years after

emigration. 138 Considering that in particular Articles 16 and 17 also act as leges speciales in respect of Article 7, various national reports include the analysis of their relevant issues in Section 3. 139 See for instance the Brazilian report. Also see the Hong Kong report on the treaty with China. 140 This would instead be unacceptable under the OECD Model, due to the residual function of Art. 21. 141 See for instance the reports from the Czech Republic, Finland, the Netherlands and Romania. An exception is Estonia, which keeps treaties following the 1963 Model even where the newer clause was available. Other countries, like Australia,

include a shorter period in Art. 15(2)(a). Further countries, instead, such as Colombia, apply a different concept for allocating

taxing powers. 142 Chile is a good example of this, since it approximates the notion of employer on the basis of the links with labour law. 143 See the reports from Australia and New Zealand. 144 See the reports from Canada, China, Finland, Germany, Kazakhstan, the Netherlands, Norway and Romania. The Czech

report indicates that in some cases a definition of employer is directly included in Art. 15(4). An alternative solution can be

found in the Peruvian treaties with Canada and Chile, which do not use the term ‘employer’ at all. Finally, the Norwegian report includes a broader geographical notion, in order to include activities related to the exploitation of offshore activities.

transport.145

Finally, several Western and Central European countries include specific

provisions for frontier workers, allowing for exclusive taxation in the country of residence,146

but in some cases also for a mechanism of taxation at source with revenue sharing.147

However, especially within the European Union, there are signs that could perhaps lead one to

consider such clauses to be an outdated tool.148

V.2 Directors’ fees

The need for a special provision on directors’ fees is mainly related to the difficulties in

determining where cross-border services have been performed. This clause is not only lex

specialis to Article 15, but also to Article 7.

The influence of the UN Model on this Article in bilateral tax treaties is frequently reported in

the national chapters contained in this book,149

but it seems to entail a much higher level of

complexity than the one that can be imagined on the basis of the Models.

A clear example of this complexity arises as to the determination of its scope. Besides the

well-known differences between the French150

and English wording of the OECD Model, and

the absence of uniform criteria in company law,151

negotiations have yielded a very

fragmented pattern, which was simply not perceived at the international level until now. Some

countries also apply Article 16 to day-to-day managerial activities;152

others consider all

income of directors to fall under Article 16;153

further countries exclude benefits in kind154

and in yet some other countries it is simply impossible to determine a consistent scope of this

provision, which significantly changes according to the applicable bilateral treaty.155

V.3 Artistes and sportsmen

Article 17 secures effective taxation in the country of source for income derived from artistic

or sports activities performed in public.156

The reports indicate a rather high degree of

consistency in the bilateral treaties around the world that reflects the common OECD/UN

pattern157

and very often adds Article 17(3) to carve out income from publicly-funded

events.158

145 See the reports from Belgium and Poland. 146 See in particular the reports from France, Germany, the Netherlands, Portugal, Spain and Sweden. 147 This mechanism is generally preferred by the Swiss treaties’ provisions on frontier workers. 148 Belgium is abandoning this type of clause. Besides, the problem in the European Union would rather be to see whether, in

the light of the increased mobility, the different treatment is truly justified by a different substance of these situations and how to isolate them from the remaining ones. 149 See the reports from Austria, Belgium, Canada, the Czech Republic, Germany, Hungary, India, the Netherlands and

Romania. 150 Membre du conseil d’administration ou de surveillance. This scope is similar to that reported from Austria, Brazil, Estonia and Germany. 151 This is the reason for having some special structure of the provision in the clauses on directors’ fees contained in German

and Netherlands tax treaties. 152 See the reports from Australia and the Netherlands. 153 See the reports from China and the Netherlands. 154 See the report from Peru. 155 See the reports from Poland and Romania. 156 Although it is nowadays extremely rare not to find this Article in bilateral tax treaties, some countries in the European Union are adding some specific nuances to it, in order to make it compatible with the Gerritse decision of the ECJ. See

further on this the report from Belgium. 157 Some countries, like Chile, accept a kind of force of attraction of the income derived by artistes and sportsmen. This is

rather questionable from the perspective of a correct application of the Models, which instead requires a performance in public for the application of this Article. In the France-Poland tax treaty the professional status of sportsmen is required for

the purpose of applying Art. 17. The Andean Treaty applies instead an exemption with a subject-to-tax requirement. 158 This type of clause generally reflects the one indicated in para. 14 of the Commentary to Art. 17 OECD Model. See

further in the reports from Argentina, Austria, Brazil, Colombia, the Czech Republic, Germany, Hong Kong, India, Lebanon, Poland, Romania, Russia, Serbia, Slovakia, Sweden and the UK (which accepts it upon request of the counterpart).

V.4 Pensions

The national reports indicate that the influence of the OECD Model on bilateral tax treaty

clauses concerning pensions is rather limited at present.

Three main trends can currently be recorded at the worldwide level.

First, a large number of bilateral treaties depart from the pattern of exclusive taxation, in the

state of residence,159

of pensions paid in respect of past employment, to reach out for various

goals. Some provisions are broader,160

others are narrower,161

further clauses more closely

resemble the options provided by Articles 18A162

and 18B163

of the UN Model, also bringing

social security pensions under the scope, and yet further provisions specifically mention

annuities,164

alimony and maintenance payments165

and for such reasons are no longer

necessarily linked with income from past employment.

Second, there seems to be a trend to lump together all types of pension for tax treaty

purposes.166

Third, some countries want to preserve their taxing rights and discourage post-retirement tax-

motivated transfers of residence. For such purposes they either keep taxing powers with the

state of the fund,167

or apply equivalent mechanisms at the time of making the payment of the

contribution to the pension scheme.

V.5 Government services

Article 19 is a provision that found its initial justification in the need to prevent one state from

levying taxes in respect of functions exercised by the other state on its territory. However, this

initial function has gradually been lost from the moment in which the local workforce

exceptions were added to this provision by both Models. Currently it creates a different tax

treaty regime, which contributes to the complexity of tax treaties (including in particular its

clause concerning pensions) without necessarily being entirely justified. Besides some minor

variations in the allocation of taxing powers,168

the national reports do not record particular

deviations from the common mould of the Models and they are missing in some relations

between Latin American countries.169

V.6 Students and professors

Article 20 represents a provision with structural and functional peculiarities within a tax treaty

because it is a clause that does not in fact allocate taxing powers, but rather merely limits their

exercise in the contracting state of sojourn. The almost identical wording of the provision in

the UN and OECD Models170

allows making a joint assessment of its impact on bilateral

treaties across the jurisdictions covered by this book.

159 Some Indian treaties apply the opposite principle. Some Colombian and French treaties provide instead for shared taxing

rights. 160 For instance, dismissal payments could be included, as in the case of several Netherlands tax treaties. 161 See for instance, the Germany and Hong Kong reports. 162 See the report from Brazil, Croatia, the Czech Republic, Estonia, Germany, Hong Kong, Hungary, India, Lebanon, Russia,

Slovenia and Slovakia. 163 See the reports from Argentina, Austria, Brazil, Canada, Germany, India and Slovenia. 164 See the reports from the Czech Republic, Estonia, Hungary, the Netherlands, Poland, Portugal and Uganda. 165 See the reports from the Czech Republic, Estonia and Germany. 166 This trend is fairly clear from the UK report. 167 See the reports from Croatia, Finland, the Netherlands and Sweden. 168 See the report from the Netherlands. 169 See the report from Argentina. 170 The UN Model defines the subjective scope of the clause by also making reference to business trainees besides students and business apprentices, which are instead the only two categories explicitly mentioned in the OECD Model.

Two elements are particularly important. First, a very large number of countries applies the

students clause with some modifications (very often as a separate article) to allow short-term

visiting professors and researchers to be fiscally carved out from the territory of the other

contracting state in which they sojourn.171

Second, virtually all students clauses share the core

elements of the Models, combined with further specifications that are the outcome of joint

negotiations and of tax treaty policy standards.

V.7 Other income

The residual distributive rule has a different importance and concrete scope according to

whether or not its scope is interpreted narrowly. Besides the relations with other provisions,

which have already been indicated earlier in this section, the clause presents different features

in both Models, since the UN Model retains shared taxing powers (Article 21(3)), whereby

the OECD clause exclusively allocates them to the state of residence. Once more the influence

of the UN Model can be clearly perceived in the relations with non-OECD countries,172

but

also in general in OECD countries like Australia.

VI. Methods for relieving double taxation

The differences between the UN and the OECD Models on the methods for relieving double

taxation are rather limited and mainly concern Article 23A, which mentions Article 12 in

paragraph 2 of the UN Model and prevents double non-taxation through the paragraph 4

added in 2000 to the OECD Model. Accordingly, insofar as this common mould is reflected in

a bilateral treaty, both the OECD and UN Commentaries can have an influence in a more

flexible way over the entire network of bilateral tax treaties.

Nevertheless, the analysis of bilateral tax treaties reported in this book shows quite a large

number of treaties that are only partly consistent with the Models and provide for alternative

solutions. Such treaties can be grouped in the following three categories, having regard to the

issues of consistency that they raise.

The first issue arises as to the method chosen for relieving double taxation. Some countries

tend to include either the exemption,173

or the credit174

method in their treaty network. Either

method affects the exercise of the taxing jurisdiction, thus also the treatment of foreign

losses.175

However, usually no complete consistency is achieved, since the national policy

often changes across time176

and, more frequently, includes exceptions for various reasons.177

In some cases this concern may lead countries to apply different relief methods even in

171 See the reports from Argentina, Australia, Austria, Brazil, Canada, the Czech Republic, Estonia, Finland (though only in

old treaties), Hong Kong, India, Liechtenstein, the Netherlands, Portugal, Serbia, Slovakia, Spain and the US. 172 See the reports from Argentina, Austria, Brazil, Chile, Colombia, Estonia, Finland, India, Lebanon, Portugal, Russia and

Slovakia. 173 See the reports from Austria, Belgium, Germany, Liechtenstein, the Netherlands (credit for passive income), Poland,

Romania (also credit), Slovenia (also credit) and Spain (also credit). 174 See the reports from Argentina, Australia, Brazil, Canada, Chile, China, Colombia, Croatia, the Czech Republic, Estonia,

Finland, France, Hong Kong, Hungary, India, Italy, Kazakhstan, Lebanon, New Zealand, Norway, Peru, Portugal, Romania

(also exemption), Russia, Serbia, Slovakia, Slovenia (also exemption with progression), Spain (also exemption), Sweden,

Uganda, the UK and the US. 175 Accordingly, when, as for the exemption method, the taxing jurisdiction of the country of residence is not exercised on

foreign-source income, losses are generally not considered by this country. However, some countries using the exemption

method, as the report from Austria indicates, apply a temporary deduction of foreign losses with a later recapture. This seems

particularly to be an issue in the European Union for the purpose of countering cash-flow disadvantages. 176 See in particular (though not exclusively) the reports from Argentina, Brazil, Croatia, the Czech Republic, France,

Norway, Poland, Serbia and Slovenia. 177 This can be due to the presence of territorial systems or treaties, such as for the Andean Convention (on which, see the

reports from Argentina or Chile), or the stronger concern for double non-taxation in the relations with some countries (see the reports from Austria, Brazil, Estonia, France, Germany, the Netherlands and Poland).

respect of different categories of income within the very same treaty,178

also in further cases

than the ones falling under the scope of Article 23A(2) of the OECD and UN Models.179

The second issue arises as to the influence of relief methods on cross-border investment and

the need to preserve the tax policy goals of a given country, which is simply not covered by

either Model, but very diffused in bilateral treaties, in particular180

with developing countries.

The clash here mostly appears between a paternalistic theory, developed in the US several

decades ago,181

countering tax sparing clauses and the opposite view, mainly advocated by

Brazil182

and India, which supports such clauses with a view to protecting the tax policy

prerogatives of capital-importing countries.183

The third issue arises with the boundaries of relief, differentiating between countries that limit

it to juridical double taxation (as under the Models) and those that apply it also to economic

double taxation,184

which mainly occurs in respect of intercompany dividends.

VII. The non-discrimination principle

The non-discrimination principle finds a rather common basis in both Models, and is the only

difference related to the specification of different treatment with respect to residence, which is

included (since 1992) in the OECD Model only, and is not always caught by bilateral tax

treaties around the world. Exception being made for such a difference and for the non-

discrimination provisions treaty clauses used by some countries,185

bilateral tax treaties show

a fairly high level of consistency with the nationality clause of Article 24(1), a frequent use of

the Article 24(3) (PE non-discrimination provision), a rather frequent use (after 1992) of the

clause that applies the non-discrimination principle regardless of the subjective and objective

scope of a tax treaty, a more moderate use of the deduction (Article 24(4)) and ownership186

(Article 24(5)) non-discrimination clauses,187

as well as a very frequent absence of the

stateless non-discrimination clause of Article 24(2).188

Two further important points should be mentioned before a more specific analysis of some

peculiarities of the single non-discrimination clauses in bilateral tax treaties is provided. First,

178 See further on this the reports from Canada (foreign affiliate rules), the Czech Republic (employment income), Finland (various provisions), Hong Kong (various provisions), Portugal (various provisions), Romania (various provisions) and Spain

(various provisions). 179 A typical example of this may be the presence of subject-to-tax, switchover, activity clauses and further anti-avoidance

(including anti-rule shopping) clauses as limits to the application of a given relief method in respect of specific categories of income. Such clauses do not necessarily apply only in respect of tainted abusive practices, but also in order to achieve a

consistent exercise of taxing powers, as their use in the Nordic Convention indicates (see further on this in particular the

reports from Finland and Sweden). 180 Some countries, however, use tax sparing also in tax treaties with OECD countries. See further on this the report from the UK. 181 This theory was developed by Stanley Surrey and presented at the hearings before the US Senate on the treaty with

Pakistan. See further on this the report from the US. 182 The report from Brazil clearly explains that the rationale of tax sparing clauses is related to keeping the allocation of taxing powers between the two contracting states and differentiates them from matching credit clauses, also frequently

included in bilateral tax treaties with developing countries, which have instead a clearer tax incentive function. 183 The author believes that the criticism of notional tax credits should be reconsidered in an open economy, where

multinationals still repatriate lower taxed income through complex international tax planning schemes. See further on this P. Pistone, ‘Tax Treaties with Developing Countries: A Plea for New Allocation Rules and a Combined Legal and Economic

Approach’, in M. Lang, P. Pistone, J. Schuch, C. Staringer, A. Storck and M. Zagler (eds.), Tax Treaties: Building Bridges

between Law and Economics, (Vienna: IBFD, 2010), pp. 413-439 (at pp. 420 et seq.). 184 See further the reports from Canada, China, Estonia, Hong Kong, India, Italy, Kazakhstan, Portugal, Russia and Spain. 185 See further on this the reports from India and New Zealand. 186 The reports from Argentina, Austria, Belgium, Canada, Colombia, Germany, Italy, the Netherlands, New Zealand, Peru

and Russia indicate that the clause is not always included in the bilateral tax treaties. 187 The reports from Austria, Belgium, Colombia, the Czech Republic, Italy and New Zealand indicate that the clause is not always included in the bilateral tax treaties. 188 Although this is generally not perceived as a problem in the countries reported on in this book, the author believes that the

failure to include this clause in some countries with a higher number of stateless persons (such as, in particular, in the case of

Estonia after the dissolution of the Soviet Union) could easily give rise to possible problems, including from an EU law perspective.

the existence of a corresponding, but much broader, principle under European Union law has

considerably reduced the importance of this principle within Europe, also in the relations with

non-EU countries. The importance of this is not always clearly perceived by tax treaty experts

who focus their attention only on the wording of the Models. Second, some countries

deliberately discriminate against non-residents and non-nationals and for this reason have

either long refrained from pursuing the inclusion of non-discrimination clauses in their tax

treaties,189

or have designed carve-out clauses that allow them to discriminate in cross-border

situations without giving rise to possible friction with this principle under their tax treaties.190

Among the various peculiarities arising from the analysis of the single non-discrimination

clauses contained in bilateral tax treaties, three have been singled out and are now briefly

discussed. First, the scope of Article 24 in some bilateral tax treaties is limited to the

subjective191

and objective192

scope of a tax treaty and in others to individuals only,193

whereby in others it is expanded even beyond its boundaries or otherwise includes peculiar

categories of income, persons, taxes or rules.194

Second, some tax treaties include a right to

MFN treatment connected to the scope of the non-discrimination clause.195

Third, access to

the non-discrimination provisions in some bilateral treaties is precluded in cases of abuse or

fraud.196

VIII. The mutual agreement procedure and international mutual assistance in tax

matters

Tax treaties do not contain proper procedural rules, which are instead left to the national law

of the contracting states. However, various clauses contained in tax treaties allow tax

authorities of the contracting states to cooperate with each other with a view to solving cases

(Article 25) of taxation not in accordance with the treaty by means of a mutual agreement (or

also, in the case of Article 26, with domestic law of the contracting states), exchanging

information (Article 26) and cooperating in the collection of taxes (Article 27).197

For all other

purposes there seems to be a very wide consensus throughout bilateral tax treaties along the

common pattern provided by the OECD and UN Models, which the author regards as

potentially suitable for replacing the current bilateral tax treaties with multilateral ones in a

not too remote future. Possibly, this has already occurred to some extent, such as in the case

of the multilateral convention provided by the OECD and the Council of Europe, which

several countries have already signed/ratified.198

189 The clearest example of this category is Australia and, even more, New Zealand. However, also other countries report (see also in the reports from Argentina, Austria, Belgium, India, the Netherlands, Norway and Spain) the absence of the non-

discrimination clause in some treaties. 190 Typical examples of such provisions are the ones aimed at considering branch profits taxes (see the reports from Canada,

Chile, Kazakhstan, Uganda and the US), thin cap rules applicable only in the cross-border situations (see the reports from Argentina, Australia, the Czech Republic and Russia), or other discriminatory provisions (see the reports from Estonia and

India) as compatible with the non-discrimination clauses. 191 See further on this the reports from Argentina, Australia, Brazil, Canada, Chile, the Czech Republic, Hong Kong and the

UK. 192 See further on this the reports from Austria, Canada, Chile, Colombia, Estonia, Finland, France, Hungary, India, the

Netherlands, Peru, Russia, Slovakia, Slovenia and the UK. 193 See further on this the reports from Austria, France, Germany and Serbia. 194 See further on this the report from Belgium (including a provision on pension funds and a rule on equal treatment as residents of the other contracting state in case more than 50 per cent of income is sourced there), France (including social

security contributions), Hungary (including rules to avoid an MFN treatment), the Netherlands (including pro-rata deduction

rules, but also contribution to pension schemes and charities) and Slovenia (including contributions to pension schemes). 195 See further on this the report from Austria, the Czech Republic (mentioning a treatment in fact similar to an MFN clause), France, New Zealand, Poland, Portugal, Russia (not a proper MFN treatment, but rather an equivalent one) and Spain. 196 See further on this the report from Australia, Austria and Romania. 197 This provision is not included in the UN Model. 198 See further on this the reports from Belgium, Canada (though not yet ratified), Finland, the Netherlands, Portugal, Sweden and the UK.

Some more specific information is now provided on bilateral treaties reported on in this book

with respect to i) mutual agreement procedures and arbitration, ii) exchange of information

clauses and iii) mutual assistance in collection of taxes.

VIII.1 Mutual agreement procedures and arbitration

The OECD and the UN Models present a fairly homogeneous wording and structure on the

procedures for cooperation between competent authorities, aimed at solving cases of taxation

not in accordance with the treaty. The main differences arise from the fact that UN Model

currently contains a different provision on implementation of such procedures in its Article

25(4) and lacks an arbitration clause at the end of a mutual agreement procedure, as Article

25(5) OECD Model does.

Bilateral tax treaties show various differences from the common pattern. The main ones can

be grouped in the following six categories. First, some treaties have either a shorter time

limit199

for presenting the case before the competent authorities than the Model, or no time

limit at all.200

Second, some treaties do not include the clause usually contained in the second

sentence of Article 25(2) of the Models and thus do not waive the time limits for the

implementation of the mutual agreement into the domestic tax systems of either contracting

state.201

Third, various treaties do not include the clause allowing for interpretative mutual

agreement procedures,202

or that for eliminating double taxation in cases not provided for by

the treaty,203

both included in Article 25(3) of the UN and OECD Models. Fourth, although

bilateral tax treaties usually allow competent authorities to communicate with each other, they

do not always provide for the possibility of setting up a joint commission as indicated under

Article 25(4) of the Models,204

or of developing appropriate methodologies for the

implementation of the mutual agreement along the scheme of the second and third sentences

of Article 25(4) UN Model.205

Fifth, some treaties do not allow for mutual agreement

procedures when a judicial procedure is pending.206

Sixth, and by far the most controversial

issue, some countries reject arbitration in their tax treaties concluded after 2008207

and other

countries apply them on the basis of a different pattern from that provided by Article 25(5)

OECD Model.208

Although it may still be too early to draw some conclusions, since the

OECD only included it in its 2008 update to the Model, the author perceives some difficulties

in spreading the arbitration culture into tax treaties and believes that such a dispute settlement

199 The shorter time limit for raising the case is usually of two years. See further the reports from Argentina, Australia, Brazil,

Canada, Croatia, France (in some treaties even as short as three months), Hungary, India, Italy, Norway, Poland, Portugal,

Russia, Slovakia, Slovenia and Uganda. 200 This is usually the case of treaties with Sweden and the UK. See further on this the reports from Australia, Chile,

Hungary, India, Poland, Portugal, Slovakia and Slovenia. 201 See the reports from Argentina, Australia (in one case also indicating the actual time limits for implementation of the

agreement), Canada, Chile, Croatia, Estonia, India, Poland, Slovakia and Slovenia. 202 See the reports from Brazil, Canada, Colombia, Estonia and Slovenia. 203 See the report from Argentina. The US report indicates that in US tax treaties the list of issues that can be object of mutual

agreement procedure is specifically indicated. 204 Several countries reported the absence of such a clause. See further on this the reports from Argentina, Australia, Brazil, Canada, Colombia, Estonia, France, Germany, India, Portugal, Romania, Russia, Slovenia and Sweden. 205 Nevertheless, this clause is included in some treaties. See further on this the reports from France and India. 206 See further on this the reports from Italy and Spain. 207 See further on this the reports from Argentina, Brazil, China, Colombia, India, Lebanon, Norway, Serbia, Spain, Uganda. Other countries only rarely use arbitration, such as Belgium, Croatia, the Czech Republic, Italy, Poland, Romania and

Slovakia. This is also the case of Australia and New Zealand, the only treaty including such a clause for both countries and

for factual disputes only. 208 In some treaties the arbitration is carried out along the lines of GATS (see the reports from Australia, Canada, Chile, Finland and New Zealand). The Austria-Germany tax treaty contains a clause requiring arbitration before the ECJ (see further

on this the report from Austria, which also indicates a more general right of being heard under the Austrian tax treaty

arbitration procedures). Other tax treaties provide different types of arbitration. See further on this the reports from Austria,

Estonia, Finland, France, Italy, Kazakhstan, Poland, Russia, Sweden and the US. In particular, the US includes the so-called baseball arbitration clauses.

mechanism can more properly be used in respect of factual disputes than those concerning

legal interpretation.

VIII.2 Exchange of information clauses

Due to the very effective action undertaken by the OECD in the recent years, exchange of

information clauses have experienced during the past two years the most dramatic evolution

ever recorded in tax treaties around the world.209

All country reports contained in this book

clearly witness this evolution and indicate the firm commitment of each state to secure an

effective and broad exchange of information on tax treaty and domestic law issues, along the

pattern provided by the OECD in the framework of its attempt to pursue global fiscal

transparency.210

Accordingly, even countries with narrow information clauses, i.e. limited by

Articles 1 or 2 of tax treaties, or allowing for information that is necessary to secure the

correct application of the tax treaty only, are now being gradually upgraded in order to

remove all such limitations and make it possible for the tax authorities to obtain any

information that is foreseeably relevant for the correct application of a tax treaty or of

domestic law of the requesting state as well. This radical change will certainly imply an

exponential growth in the volumes of information exchanged, which can easily cause critical

situations, especially if the current major brake to such an evolution, generally known as the

prohibition of relying on exchange of information clauses in tax treaties in respect of fishing

expeditions, is still largely perceived by the national reports as a terra incognita. Such critical

situations can arise particularly when countries try to outsource their auditing to the other

contracting state or, more in general, in situations of asymmetrical flows of information.211

Some major progress still seems required in three major areas, such as i) the update of Article

26(4) and (5) into all bilateral tax treaties, ii) the diffusion of spontaneous and automatic

exchange of information clauses, and iii) the techniques and clauses for sharing best tax

practices. Nevertheless, interesting clauses are already available in some treaties around the

world.212

VIII.3 Mutual assistance in collection of taxes

The OECD may have already perceived the problems and implications of a tax treaty clause

on the assistance in collection of taxes when including in the 2003 update of the OECD

Model,213

since it explicitly suggested negotiating it in the presence of similar tax systems as

well. This perception is confirmed by the country reports included in this book, which show

that the clause is not used by all countries214

and that the ones that do use it, only include it in

a selected number of their tax treaties.215

209 The starting moment is in early 2009, when the existing reservations on Article 26(5) were withdrawn. 210 The author believes in the possible synergies between the OECD and the EU, especially considering the long-standing experience of the latter in some areas, such as spontaneous and automatic exchange of information, also on the basis of

special agreements, such as the one on the taxation of savings, and the recently updated EU directive on exchange of

information. 211 This can for instance be the case with some capital-importing countries, which have underdeveloped structures for tax administration and can thus find it more difficult to supply technical information requested by the other contracting state, as

well as have a more limited number of cases in which they seek information from the other state. However, the author

perceives that problems of asymmetry can also arise in relations with countries, such as Switzerland, which have long kept

tight secrecy rules and are now naturally exposed to a strong request by all states that have concluded protocols or treaties including clauses on exchange of information with such a country. 212 See further on this the reports from Australia (treaty with India), Finland and Sweden (both for the Nordic Convention,

which provides for the servicing of documents, simultaneous auditing, single taxation of employees with revenue sharing and

collection of taxes) and the Netherlands (agreements for automatic exchange of information and joint tax examinations). 213 This clause is not included in the UN Model. 214 Some countries have not included clauses on mutual assistance in collection of taxes in their bilateral treaties. See further

on this the reports from Argentina, Brazil, Chile, China, Italy, Liechtenstein, Peru, Portugal, Serbia and the US. 215 See further on this the reports from Australia, Canada, Colombia, Croatia, the Czech Republic, Germany (often with a different structure from the OECD clause), Lebanon (not with the wording of the OECD clause), New Zealand (not with the

wording of the OECD clause), Norway, Poland, Russia (usually not with the wording of the OECD clause), Slovakia, Slovenia, Spain and the UK.