Income tax and capital gains tax: capital gains in pre-CGT tax treaties

This Ruling sets out the ATO's position on the extent and manner in which capital gains are dealt with in Australia's tax treaties negotiated before the enactment of the capital gains tax by the insertion of Part IIIA into the ITAA 1936.

This Ruling is not about the treatment of income gains. Thus the one-off transaction with business characteristics held by the High Court to constitute a business profit in Thiel v. FCT [F1] is not covered in this Ruling. It is also noted that in Lamesa v. FCT [F2] the ATO did not take the view that the gains on the sale of the shares were of a capital nature. The ATO's view was based on the fact that the controlling minds behind Lamesa were in the business of acquiring companies, building them up and then selling them at a profit. On this basis the ATO considered the gain was of an income nature.

This ruling assumes that certain receipts on the borderline of the income/capital gain distinction and assimilated to income under Australian law (discussed below) are treated as income in Australia's tax treaties. Apart from that observation, the Ruling does not deal with the question of what amounts to a capital gain for the purpose of a double tax treaty.

Australia's right to tax gains taxable in Australia exclusively under the capital gains tax regime (that is, Part IIIA of the ITAA 1936, or Part 3-1 of the ITAA 1997) is not limited by pre-CGT treaties. This is because: (a) from Australia's perspective these treaties do not distribute taxing rights over capital gains; and (b) with the exception of the Australia/Austria DTA, under relevant Taxes Covered articles, Australia's taxes on capital gains are not taxes to which pre-CGT treaties apply.

Pre-CGT treaties apply, however, to gains which at the time of signature of the relevant treaty were assessable under provisions other than the comprehensive capital gains regime. These treaties can apply, for instance, to income according to ordinary concepts arising on the alienation of property and certain borderline gains assimilated to income (for example, gains assessable under sections 25A, 26(a), 26AAA, 36, 47 or 59 of the ITAA 1936) which were taxable prior to the introduction of the comprehensive capital gains tax regime.

The following tax treaties, as negotiated before Australia introduced comprehensive taxation of capital gains are referred to as 'pre-CGT treaties': United Kingdom (1967); Japan (1969); Singapore (1969); Germany (1972); New Zealand (1972); France (1976); Netherlands (1976); Belgium (1977); Philippines (1979); Canada (1980); Switzerland (1980); Malaysia (1981); Sweden (1981); Denmark (1981); Italy (1982); Korea (1982); Norway (1982); USA (1983); Ireland (1983); Malta (1984); Finland (1984); and Austria (1986).

Although the treaty with Austria was signed shortly after CGT legislation was assented to in 1986, the final text was negotiated in 1984. The signed text remained unchanged from the time of negotiation, so for the purposes of this Ruling it is a 'pre-CGT treaty'. Note also that the Singapore (1969), New Zealand (1972) and Malaysia (1981) treaties have since been renegotiated. (Of course, over time the other pre-CGT treaties will be renegotiated.) A second Protocol to the Netherlands treaty (1986) and a protocol to the French treaty (1989) [F3] do not address alienation of property.

Treaties negotiated after the Australia/Austria DTA are referred to as 'post-CGT treaties'.

An important category of pre-CGT treaties concerns those negotiated prior to Australia's membership of the OECD: United Kingdom (1967); Japan (1969); Singapore (1969); Germany (1972); New Zealand (1972). These treaties contain some marked departures from the OECD Model. In this Ruling, to distinguish them from other pre-CGT treaties, these five treaties are referred to as 'pre-OECD treaties'.

This ruling refers to 'borderline gains'. These are gains which are on the borderline of the income/capital distinction. Technically these gains may not be income according to ordinary concepts, but are included in the income tax base as assessable income. In the current law, such gains include: profit making undertakings or schemes; sale of trading stock as part of the sale of a total business; lump sum payments on termination of employment; return to work payments; realisation of traditional securities; certain foreign exchange gains; bounties and subsidies; compensation for loss of trading stock or profits; certain liquidator's distributions; depreciation recapture; and income received after death. In past law which applied when various of the pre-CGT treaties were negotiated, they included gains on property acquired for the purpose of profit-making by sale or purchased and sold within 12 months. In reality all these measures address borderline issues between income and capital such as income replacement, deduction recapture, clarifying the borderline or administrative rule of thumb solutions where the income/capital distinction is very fine or difficult to apply. Unless the context requires otherwise, 'income' when used in double tax treaties clearly requires a broader meaning than 'income according to ordinary concepts' - otherwise these borderline gains could be outside the scope of Australia's tax treaties.

Gains representing depreciation recapture and gains assessable under the former sections 26(a), 25A and 26AAA were specifically referred to in various negotiations as examples of income from the alienation of property. The ATO considers these to be the most obvious examples on the borderline of the income/capital distinction but other cases in the previous paragraph could also involve alienation of property and would be treated similarly.

Other abbreviations and terms used in this Ruling are listed below: 1963 Draft Convention Draft Double Tax Convention on Income and Capital , OECD, Paris. A commentary also accompanied the Draft Convention. Now replaced by the OECD Model (discussed below). Agreements Act International Tax Agreements Act 1953 ATO Australian Taxation Office CGT capital gains tax credit article Methods of Elimination of Double Taxation Article. Australia uses the credit method to eliminate double taxation of items where, under the distributive rules, taxing rights are shared. distributive rules Tax treaty provisions where the Contracting States agree to limit their taxing rights are referred to as 'distributive rules' in this Ruling. In the OECD Model, these are the rules contained in Chapter III. In most Australian treaties they are in Articles 6 to 21. DTA Double Tax Agreement DTC Double Tax Convention ITAA 1936 Income Tax Assessment Act 1936 ITAA 1997 Income Tax Assessment Act 1997 OECD Model Model Tax Convention on Income and on Capital , OECD, Paris. This model (and Commentary) was originally published in 1977, effectively replacing the 1963 Draft Convention. It was revised in 1992 and periodically updated since then. residual article Income Not Expressly Mentioned Article or Other Income Article. In the OECD Model this is found at Article 21. The first expression was used in the 1963 Draft Convention and the latter adopted in the 1977 OECD Model. undefined terms provision General rule of interpretation for terms not defined in tax treaties. In the OECD Model this is found at Article 3.2. UN Manual Manual for the Negotiation of Bilateral Tax Treaties between Developed and Developing Countries , United Nations New York, 1979. Essentially, this was an 'unofficial' version published in advance of the UN Model. UN Model United Nations Model Double Tax Convention Between Developed and Developing Countries United Nations, New York, 1980. Vienna Convention Vienna Convention on the Law of Treaties , done at Vienna on 23 May 1969.

This ruling, when finalised, will apply to years commencing both before and after its date of issue. However, the Ruling will not apply to taxpayers to the extent that it conflicts with the terms of a settlement of a dispute agreed to before the date of issue of the final Ruling (see paragraphs 21 and 22 of Taxation Ruling TR 92/20).

This ruling responds to views expressed by some commentators that capital gains are relieved from liability to capital gains tax by the operation of Australia's pre-CGT tax treaties. Under these treaties they argue that, in certain circumstances, taxing rights over a non-resident's capital gains are allocated exclusively to the country of residence thereby eliminating liability to Australian tax.

While noting there are alternative arguments, the ATO adheres to the view that Australia's right to tax capital gains is not limited by pre-CGT treaties. The ATO considers that taxes on capital gains are not taxes to which pre-CGT treaties apply. Even if this is not the case, the distributive rules of pre-CGT treaties do not limit domestic law taxing rights over capital gains.

Context is central to the ATO understanding of these treaty issues. For example, context is relevant to deciding whether under the Taxes Covered Article, CGT is substantially similar to the 'Australian income tax' existing when pre-CGT treaties were signed. Also the undefined terms provision may be used to enliven the meaning of key expressions such as 'income from alienation' and 'profits of an enterprise' by using the meaning under domestic tax law - so long as the treaty context permits.

The ATO position is based on a holistic view of the contextual matrix of pre-CGT treaties. It does not rest on any particular point but a number of indicators collectively indicate that capital gains were not covered in pre-CGT treaties (except Austria) and distributive rules in these treaties did not limit taxing rights over capital gains.

The ATO considers the context of these treaties shows that Australia attempted to limit the application of the Alienation of Property Article in all of Australia's pre-CGT treaties to income gains and did not intend that capital gains were to be covered by these treaties. These contextual factors include: * Australia's 1976 reservation to the OECD Model retained the right to propose changes to the Capital Gains Article, because Australia did not levy a capital gains tax. The reservation was removed from the 1992 Model - the first Model published after CGT was introduced. * Australian pre-CGT treaty practice compared to the prevailing OECD Model (Table 1, columns 1 and 2) consistently reflects this reservation. * Drafting changes to Australia's tax treaties made after the introduction of CGT (Table 1, columns 2 and 3) also confirm an intention to not deal with capital gains in pre-CGT treaties. * Comparing the pre-CGT treaties with the tax treaty policies of pre-CGT tax treaty partners reveals that in its negotiations Australia sought and obtained significant changes to the international models used by treaty partners. Most of Australia's pre-CGT treaty partners consistently followed the OECD Model and comprehensively dealt with all capital gains from the alienation of property. Departures from the OECD Model were clearly at Australia's behest - implying that the treaty partners would not interpret these Australian treaties like their other tax treaties. Furthermore treaty partners had comprehensive CGT regimes in their domestic law and would be expected to seek to deal with capital gains. Pre-CGT treaties reflect compromises generally negotiated at Australia's insistence. While these changes reflect a consistent Australian position, treaty partners were able to agree to this position because it did not restrict their right to impose their own capital gains taxes and, because there was no Australian CGT, there was little prospect of double taxation occurring. * Australia's economic and political interests favoured not dealing with capital gains in the pre-CGT treaties. CGT was a contentious domestic issue. Further, (as Australia's post-CGT treaty practice clearly shows) Australia's economic interests were not perceived to align with allocation of taxing rights under the OECD Model's Capital Gains Article. By not dealing with capital gains in pre-CGT treaties, Australia avoided referring to this contentious domestic issue while preserving its freedom of action to subsequently negotiate appropriate taxing rights over capital gains if a CGT was introduced.

The above observations are made independently of the evidence of the actual negotiations. Nevertheless (while there is some debate as to the admissibility of these materials into evidence [F7] ) Australia's records of the negotiations with pre-CGT treaty partners [F8] also reflect the view taken by the ATO.

Australia's pre-CGT practice was against a background in which Australia did not tax capital gains and the introduction of such a tax was unlikely. Thus there was little possibility of double taxation. The question is how did the parties intend pre-CGT treaties to deal with capital gains in the event Australia introduced such a tax?

One interpretation might be that Australia simply sought to remove references to capital gains for presentational purposes - possibly prompted by domestic political concerns. If Australia subsequently introduced a CGT, in the pre-CGT treaties the result, broadly, would be that taxing rights over capital gains would be allocated under the business profits rules, yielding a similar result to the OECD Capital Gains Article. It might also be argued that the chances of an Australian CGT being introduced were so remote that in practice it was only a minor matter for Australia to effectively cede many CGT taxing rights to the residence country.

The ATO considers the better view is that these changes were intended to remove capital gains entirely from the scope of pre-CGT treaties. In so doing Australia, as a source country retained its freedom of action to levy a CGT if it was subsequently introduced. Only when a CGT was introduced would the domestic political processes be able to give proper consideration to its cross-border application. In any event, because of Australia's strong disposition to preservation of source country taxing rights (evidenced by its reservations to the OECD Model and post-CGT treaty practice) Australia is very unlikely to have intended taxing rights over capital gains in pre-CGT treaties to be allocated similarly to the OECD Model Capital Gains Article - which has very restricted source country taxing rights. Australia's records of negotiations confirm Australia had intended not to deal with capital gains in pre-CGT treaties.

A number of authors have written on this issue in recent years. Many have argued that pre-CGT treaties deal with capital gains. Clearly there are differences of emphasis between authors and differences between treaties. The following synthesises some of the key themes emerging from these commentaries.

The ATO does not agree with these arguments, often because they give insufficient weight to the context of these treaties as outlined above. The ATO response is provided in the detailed Explanation.

The alternative view often emphasises the requirement in Article 31 of the Vienna Convention that treaty language is to be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty and in the light of its object and purposes. As one of the purposes of tax treaties is to avoid double taxation such treaties should be interpreted liberally to achieve this purpose: a point enunciated by McHugh J in Thiel . [F9]

In considering whether CGT is a tax to which a pre-CGT treaty applies, a key requirement is that the CGT be considered one of the taxes existing at the time of signature (usually by Australia as the 'Australian income tax') or that the tax be at least substantially similar to an existing tax. The alternative view argues that an ambulatory approach is to be applied to the meaning of 'Australian income tax' and that CGT is similar to the existing taxes.

The alternative view argues 'Australian income tax' effectively included a capital gains tax when most of the pre-CGT treaties were signed. Section 26AAA, for example, was always a 'pure' tax on capital gains. The CGT merely extends the range of transactions subject to the tax. Therefore CGT is an 'existing tax' or alternatively it is substantially similar to an existing tax and is therefore a tax to which pre-CGT treaties apply. The views of international tax authors and decisions of a foreign tribunal are quoted to support this view.

There is also an argument that if one country's list of existing taxes encompasses capital gains, then a tax on capital gains introduced by the other state will fall with the description of taxes covered by the treaty. Where the 'existing taxes' of a treaty partner include CGT, it is argued the treaty will cover Australia's CGT. Views of international tax authors and decisions of foreign courts are cited to support this view.

The central issue here is whether the expression 'profits of an enterprise' used in this article, encompasses capital gains. This expression is not defined in tax treaties.

A principal argument advanced by commentators is that under subsection 3(2) of the Agreements Act, 'profits' is read as meaning a reference to any 'taxable income' from commercial activity. As taxable income includes net capital gains, capital gains are included within the meaning of 'profits' in this Article.

Furthermore, the undefined terms provision permits recourse to domestic tax law to give meaning to 'profits of an enterprise'. Several Australian cases dealing with taxation of dividends have found that 'profits' can include capital gains. The advocates of the alternative view have also identified several instances in post-CGT treaty practice where the term 'profits' appears to have been used interchangeably with 'gains'. It is therefore argued that the expression 'profits of an enterprise' can include capital gains.

The Alienation of Property Article in pre-CGT treaties generally deals with the income (and occasionally gains) from the alienation of property. The ATO considers that in this context the word 'income' does not embrace capital gains and where 'gains' are referred to, they relate only to revenue gains.

An alternative view is that 'income' and 'gains' should be read expansively to deal with capital gains. In particular it is argued that one of the distributive rules in this article deals with 'income from the alienation of capital assets' and that expression as a whole must be read as dealing with capital gains.

If a capital gain is not dealt with by the Alienation of Property or Business Profits Articles, the alternative view argues that potentially the residual article allocates taxing rights on a source basis.

To further demonstrate that the distributive articles of pre-CGT treaties are capable of covering capital gains reference is made to the residual article in the US treaty in particular. In general it is argued that the term 'income' should be interpreted as including 'capital gains', because in Australian treaty practice the terms 'income', 'profit' and 'gains' are used interchangeably. But in relation to the Australia/United States DTC, commentators refer to material produced by the United States as evidence of an understanding by the US that capital gains may be dealt with in this Article.

The pre-OECD treaties raise slightly different considerations because they do not have an Alienation of Property or residual article. The ATO and counter views are briefly discussed below.

The ATO considers the absence of an Alienation of Property Article strongly suggests an intention not to deal with capital gains in these treaties. The United Kingdom and Japanese treaties do not have OECD type business profits article and instead define 'industrial and commercial profits'. [F10] For a capital gain to come within this definition it is necessary to consider whether the capital gains can be considered as 'income' (United Kingdom) or 'profits' (Japan). The ATO considers that the context of these treaties, including relevant Explanatory Memorandums, does not permit such an interpretation. The alternative view argues that treaties must be interpreted liberally and the income/capital distinction is unjustified in these cases. Reference is also made to Australian tax case law where the expressions 'income' and 'profits' have been taken to include capital gains.

The analysis of this issue raises the following considerations: * To what extent do pre-CGT treaties cover taxes on capital gains? * Does the Business Profits Article deal with capital gains? That is, are capital gains included within the expression 'profits of an enterprise' or (in the case of earlier treaties) 'industrial and commercial profits'? * Are capital gains dealt with by the Alienation of Property Article of pre-CGT treaties? Is a capital gain 'income' from the alienation of the types of property described in this article? * Are capital gains dealt with by the residual article of pre-CGT treaties as income not dealt with in the other distributive rules? * Even if capital gains are not dealt with in the distributive rules is there an obligation to give credit for source country taxation under the Methods of Elimination of Double Taxation Article of pre-CGT treaties?

The ATO argues that the context of pre-CGT treaties is relevant to the above issues. After a detailed analysis of the context of the pre-CGT treaties in relation to capital gains, this Explanation then examines each of the above issues in detail. For the purposes of this analysis, pre-CGT treaties have been examined within two broad categories: * Pre-OECD treaties: that is the earliest pre-CGT treaties entered into before Australia joined the OECD (United Kingdom, Japan, Singapore, New Zealand and Germany). The last issue listed above (application of the credit rules) is particularly relevant to the pre-OECD treaties. * Other pre-CGT treaties (which are examined first).

For the purposes of Article 31 of the Vienna Convention, context includes, in addition to the text, any instrument which was made by one or more of the parties in connection with the conclusion of the treaty and accepted by the parties as an instrument in relation to the treaty. In Thiel [F11] , Dawson J was prepared to consider the OECD Model and Commentaries as part of this context, however he noted some doubts expressed in an article by Avery Jones and others on this point. [F12]

Avery Jones agreed the 'Vienna Convention context' is too narrow for the purposes of the undefined terms provision. While it provides a starting point in determining the meaning of 'context' in the undefined terms provision, it is clear that it cannot be definitive. As the authors state: Applying the Vienna context definition to the expression 'unless the context otherwise requires' would make no sense because the Vienna Context was not meant to be used in isolation from ... other factors. [F13]

The authors suggest a wider range of 'external' context is available in applying the undefined terms provision although exactly how far the ambit of 'context' should extend is difficult to say.

In relation to statutory interpretation the current view of the courts has been to consider context in its broadest sense. In Consolidated Press Holdings Ltd & Anor v. FC of T [F14] Hill J stated: Although judicial views on the principles of construction of taxation statutes have differed over time (see cases referred to in the article 'A Judicial Perspective on Tax Law Reform' (1998) 72 Australian Law Journal 685), the modern view, at least generally, would seem to be that the task of construing a taxation statute, like the task of construing any other statute, requires the Court to ascertain the meaning of the words used in the context in which they appear and so as to give effect to the purpose of the legislature to be found in the language which it has used, but aided by extrinsic materials to which regard is directed to be had by virtue of s15AB of the Acts Interpretation Act 1901. Context is used in a broad sense to encompass such matters as the existing state of the law and the mischief, if any, which the legislature sought to remedy: CIC Insurance Ltd v. Bankstown Football Club Ltd (1997) 187 CLR 384 and see FC of T v. Australia and New Zealand Savings Bank Ltd (1998) 156 ALR 570 at 577.

The ATO considers treaty context should be taken in its broadest sense to have regard to the full fabric of matters that may be considered , including the historical and political matrix and the matters referred to in the Avery Jones discussion. In particular the following discussion has regard to: * the practices and policies of Australia and the treaty partners in negotiating treaties at that time and subsequently; * given the usual treaty object of avoidance of double taxation, the domestic taxation environments of the two countries when the treaty was negotiated; and * the political, economic and diplomatic background to the treaty.

Australia joined the OECD in 1971. In 1976 Australia entered reservations against the new OECD Model published in 1977 (the first since the 1963 Draft Convention). Among these was a reservation to the OECD's Capital Gains Article (Article 13): [F16] Australia reserves the right to propose changes to reflect the fact that Australia does not levy a capital gains tax and that the terms 'movable property' and 'immovable property' are terms not used in Australian law.

By reserving the 'right to propose changes' to the OECD Model Australia clearly signalled it was not going to follow the OECD Model approach - which was to include a comprehensive article dealing with capital gains.

This reservation was withdrawn with the 1992 OECD Model revision - the first publication of an updated version of the OECD Model following the introduction of capital gains in Australia.

Comparing Australia's pre-CGT treaties with prevailing international model treaties points to Australia's treaty practices and policies. Australia's tax treaties are broadly based on the prevailing OECD Model. However, Australia's treaties vary from the OECD standard in many respects. In particular, there are significant variations in relation to alienation of property (both before and after the introduction of CGT).

The scheme of the OECD Capital Gains Article is described in the following table:

The OECD Model allocates taxing rights to the source country only for alienation of immovable property (on the basis of situs ) and permanent establishment property.

Consistent with an underlying philosophy of the OECD Model that trade between Member Countries as a whole would be in balance, taxing rights over the alienation of residual property are allocated to the residence country. Given Australia's position as a net capital importer and consequent disposition to protecting source country taxing rights (see discussion on the economic background below) the OECD Model would have been difficult to adopt without modification. Following Australia's membership of the OECD it was necessary to develop and defend a position on the OECD Model Capital Gains Article for negotiation with other Member Countries. [F17]

Developing a treaty source country position in an environment in which Australia did not tax capital gains would have been a difficult abstract [F18] exercise, potentially limiting the cross-border aspects of any future CGT regime. Australia's response was to reserve its position on the OECD Capital Gains Article. In negotiations it then proposed modifications to that Article that effectively continued the previous policy of not dealing with capital gains in its tax treaties. Thus Australia preserved its right to negotiate appropriate distributive rules for capital gains in the event a comprehensive CGT was introduced.

Pre-CGT treaties consistently modified the OECD Capital Gains Article. The Australian Model Article 13 for mid 1980 (see Annexure A) is reflected in pre-CGT treaty practice. This Model: (a) dealt with 'income' - not 'gains' as in the OECD Model (b) changed the heading to ' Alienation of Property' and (c) dealt only with real property interests.

Alienation of Property Articles in most of Australia's pre-CGT treaties dealt only with 'income' from the alienation of property. By contrast the OECD Model deals with (capital) gains. Pre-CGT treaties were drafted deliberately to exclude capital gains. The use of 'income' demonstrates a consistent intention by Australia not to deal with capital gains as such in this article. [F19]

Likewise, Australia consistently avoided using the OECD heading of ' Capital Gains'. Where a specific article deals with alienation of property, Australia's pre-CGT treaties use the broader ' Alienation of Property' as the title. [F20] Again, this reflects the fact that distributive rules in the Alienation of Property Articles of Australia's pre-CGT treaties are generally expressed to apply only in relation to income from alienation of certain property. From the Australian perspective it would be inappropriate to use the OECD heading in pre-CGT treaties, because capital gains were not to be covered by the Article.

Australia's pre-CGT Model Alienation of Property Article was less comprehensive than the OECD Model, applying only to real property interests. Annexure B, which summarises Australia's pre-CGT treaties as negotiated, reflects this more restricted approach. Alienation of Property Articles usually had limited coverage, frequently only dealing with real property interests. Alienation of business property is dealt with in only 7 [F21] of the 17 pre-CGT treaties containing an Alienation of Property Article. Because these treaties did not include a residual provision (for example OECD Model Article 13.4) even where there are rules for alienation of business property, important areas outside the context of an enterprise (such as alienation of intellectual property) were not comprehensively dealt with.

Following the introduction of Australian CGT, a number of changes were made to Australian treaty practice in relation to the Alienation of Property Article: * While the heading ' Alienation of Property' was retained, the article now dealt with 'income, profits or gains' on the alienation of certain property (formerly 'income' only). Post-CGT treaties continue the Alienation of Property heading because they continue to deal with income, as well as profits and gains (whereas other countries use the OECD Model 'gains'). * The post-CGT treaty article is more comprehensive, always explicitly dealing with alienation of real property interests, business property, ships and aircraft. * Post-CGT treaties are comprehensive with respect to capital gains, as a 'sweep up' provision reserving domestic law taxing rights to gains of a capital nature not dealt with elsewhere in the Article is also included. This provision is not found in the OECD Model, but is similar to the UN Model 'alternative' Capital Gains Article. (Income gains not dealt with in the article were likely to be dealt with as business profits or in the residual article.)

In the following discussion Australia/Korea DTC and the Australia/China DTA are referred to as representative of the 'pre' and 'post' CGT treaties respectively.

Source rules in Australia's tax treaties also changed to coincide with the introduction of CGT. [F22] All the source rules in pre-CGT treaties deal only with 'income', but source rules for Australia's post-CGT treaties changed to also deal with 'profits and gains'. Compare the source rules in Article 23 of the Australia/Korea DTC which deal only with 'income', to the rules for the Australia/China DTA (Article 23.8 of that treaty and subsection 11S(2) of the Agreements Act) which deal with 'income, profits or gains'. The expression 'income, profits or gains' is used in the source rules in all post-CGT treaties and the Agreements Act - except for the Australia/Thailand DTA, where the source rule deals only with 'income'.

The Entry into Force articles in pre-CGT treaties also change terminology following the introduction of CGT - although the changeover was not so rigorous. Pre-CGT treaties, such as the Australia/Korea DTC, refer to entry into force in Australia of (a) withholding tax on income and (b) other Australian tax, in relation to income . The language in the Australia/China DTA is unchanged. However, the next concluded treaty (with Papua New Guinea) refers to other Australian tax 'in relation to income or gains'. The following two treaties (with Thailand and Sri Lanka) revert to the pre-CGT formulation. But the following treaty with Fiji refers to 'income, profits or gains' - a form of words used consistently in all Australia's later tax treaties. Regarding the China, Sri Lanka and Thailand treaties, the ATO considers that given the context of these treaties, they have entered into force in respect of capital gains.

Drafting changes to the 'force of law' provisions of the Agreements Act further underline the view that pre-CGT treaties were not considered to deal with capital gains. Australia's method of giving legislative effect to its tax treaties is by reproducing each treaty in its entirety as a schedule to the Agreements Act. Currently, sections 5 to 11ZJ of the Agreements Act give each of the schedules (i.e., the tax treaties) force of law in Australia. In the case of the Australia/Korea DTC the relevant provision is section 11L which, among other things, provides that the provisions of the treaty 'so far as those provisions affect Australian tax, have and shall be deemed to have had, the force of law'. Importantly, the section gives the treaty the force of law in relation to tax (other than withholding tax) 'in respect of income' only. The force of law provisions of the subsequent post-CGT treaty with China, reflect new coverage of capital gains by omitting the 'in respect of income' requirement.

Some commentators have suggested instances where they believe the words 'income' and 'gains' have been used interchangeably. This is addressed below in relation to the Business Profits Article.

The Table at Annexure C shows that Australia's treaty partners consistently followed the OECD Model in contemporaneous treaties with other countries. By implication, the departures from the OECD Model in Australia's treaties were at the insistence of Australia.

In the case of the Australia/United States DTC, the Australian negotiators substantially altered the then preferred US Model to exclude capital gains from the distributive rules. The US Treasury Department's draft Model Income Tax Convention , which was published in 1981, [F23] contained a comprehensive Gains Article. This article is set out in column 2 of Annexure D. The US Model reflected US domestic law - especially in relation to alienation of real property. The changes made during the course of negotiations to both the OECD Model and the US Model were significant. Article 13 of the Australia/United States DTC is at column 3 of Annexure D. The willingness of the USA to extensively modify this Article, contrary to its treaty practice and Model position, supports a presumption that it was the intention of the negotiating parties to exclude capital gains from the coverage of the final text.

Many of Australia's tax pre-CGT treaty partners imposed a domestic capital gains tax. Australia did not. Yet Australia consistently limited the application of the Alienation of Property Article. For example, the United Kingdom had introduced a comprehensive capital gains tax regime in the 1965, just before the negotiations with Australia commenced, but no article expressly dealt with capital gains. Canada had a comprehensive capital gains tax regime and would have been expected to deal with more than real property interests in its treaty with Australia. The United States had also long taxed capital gains, but Australia limited the treaty coverage of gains in that Convention to the extent of its 1981 Foreign Investment in Real Property Tax Act and certain transport assets.

Australia's economic interests did not favour dealing with capital gains in pre-CGT treaties. Australia had long held a strong disposition to the protection of source country taxing rights. [F24]

The OECD Model assumes capital flows between Member countries are roughly equal and therefore allocates taxing rights over activities not clearly attributable to the source country to the country of residence. But Australia was a large net capital importer when the pre-CGT treaties were negotiated. [F25] Had Australia adopted the OECD Model on capital gains, in the event that a capital gains tax was introduced there would be a significant revenue impact.

Without a fully developed Australian position on cross-border taxation of capital gains (which would have been politically difficult to develop - see below) and the lack of an international source country benchmark against which to negotiate the Capital Gains Article, [F26] Australia preserved its freedom of action in the event a CGT was introduced, by generally excluding capital gains from treaty coverage.

Of course when pre-CGT treaties were negotiated, the absence of a comprehensive Australian CGT would have contributed to a perception for Australia's treaty partners that it was unnecessary to limit source country taxing rights over capital gains. Further any Australian objective to limit source country taxation of capital gains on the relatively small amount of outbound Australian investment at the time, would have been balanced against the fact that such provisions would effectively lock in the limits of the international aspects of any future capital gains tax regime.

It was also politically inexpedient to deal with this issue in tax treaties: capital gains taxation was a politically sensitive issue in this period. [F27] For most of the time which pre-CGT treaties were negotiated (1967- 1984) [F28] it would be expected that reference to capital gains in a treaty would have been, at least, unnecessary, and inconvenient to Australian Governments.

Context has not been comprehensively addressed by many of those proposing an alternative view. On the other hand the ATO sees the context of these treaties as an important aspect of the analysis.

The ATO considers Australian treaty policy, driven by its economic interests had consistently emphasised source country taxing rights to a greater extent than the OECD Model. Post-CGT treaty practice in relation to capital gains on alienation of property clearly confirms this. However there were political sensitivities in dealing with this issue in pre-CGT treaties. Moreover, in the absence a domestic capital gains regime, identifying the potential cross-border limits of such a regime would have been a difficult and abstract exercise. Australia addressed this issue by not dealing with capital gains in its pre-CGT treaties: it removed capital gains entirely from the scope of pre-CGT treaties. In so doing Australia retained its freedom of action to levy a CGT if it was subsequently introduced. Australia's records of negotiations confirm Australia had intended not to deal with capital gains in pre-CGT treaties.

The taxes to which a treaty applies are set out in the Taxes Covered Article (usually Article 2). This is intended to limit the application of the distributive rules of the treaty to the taxes described in the Taxes Covered Article. [F29]

Like many countries Australia departs from the OECD Model, by deleting the opening two paragraphs of the OECD Model Taxes Covered Article which contain generalised descriptions of the taxes to be covered. A typical Australian Taxes Covered Article and the OECD Model equivalent are set out in columns 1 and 2 respectively of Annexure E.

The Taxes Covered Article of pre-CGT treaties generally lists the 'existing taxes' to which the tax treaty applies in the opening paragraph. Australian taxes are listed as 'the Australian income tax' and most include a reference to the undistributed profits tax. [F30] The second paragraph (the 'extension provision') extends the taxes covered to identical or substantially similar taxes.

First, taxes on 'borderline gains' assimilated to income and taxed under Australian tax laws prevailing at the time of signature are 'existing taxes' for the purposes of the Taxes Covered Article of pre-CGT treaties. (Note that legislation relating to some borderline gains has been repealed - for example, sections 26(a), 25A - in relation to property acquired for the purpose of profit-making by sale - and 26AAA.)

Secondly, an important consequence flows from deleting the general descriptions of taxes covered in paragraphs 1 and 2 of the OECD Model Taxes Covered Article. Under the OECD approach, because of these general descriptions, all income and capital taxes are dealt with by the treaty and the distributive rules will apply to both groups of 'existing taxes' listed by the two countries. But where the two countries delete the OECD's paragraphs 1 and 2 and separately list the taxes covered, some distributive rules may be relevant for one country but not the other. This is starkly demonstrated in Article 21 of the Australia/Germany DTA, which allocates taxing rights over capital, but only Germany includes a capital tax in the list of capital taxes at Article 2. Clearly Article 21 of that treaty will apply only for German capital taxes and has no application to Australia.

In other words, distributive rules need not necessarily apply bilaterally. As discussed later, some provisions dealing with alienation of property were important for treaty partners particularly for presentational reasons, but from the Australian perspective were unnecessary. [F31]

A third important point emerges from a comparison with the OECD Model. The deletion of OECD model paragraphs 2.1 and 2.2 and 'in particular' from paragraph 3 imply that some caution is to be exercised in applying the extension provision. Where OECD Model paragraphs 1 and 2 are present, OECD Model paragraph 3 is generally treated as a detailed but not necessarily complete list of taxes which only illustrate the general principles in paragraphs 1 and 2. [F32] However, where only OECD paragraphs 3 and 4 are used (as paragraphs 1 and 2 in Australia's treaties) then OECD paragraph 3 (i.e., Australian Model paragraph 1) is interpreted as an exhaustive list. Countries may restrict the Article to OECD paragraphs 3 and 4 only, because they consider the treaty should exhaustively identify the taxes to be covered.

A recent revision to the OECD Model Commentary on the Taxes Covered Article (in publication) confirms that OECD Members hold this view. Referring to the practice of Member countries which do not include OECD paragraphs 1 and 2 in their treaties the Commentary notes: These countries prefer simply to list exhaustively the taxes in each country to which the convention will apply and clarify that the convention will also apply to subsequent taxes that are similar to those listed.

Because taxes are exhaustively listed, new taxes (such as CGT) need to be very similar to the existing taxes before the extension paragraph makes the new tax one of the taxes covered. The OECD suggests that bilateral clarification of the application of new taxes will be required before new taxes are adopted in the treaty. This has not occurred in the case of Australian CGT.

The ATO considers the introduction of CGT fundamentally increased Australia's tax base and the context of the treaty suggests that there was no intention to cover capital gains taxation within the expression 'Australian income tax' in paragraph 1 or by extension. The overall context of pre-CGT treaties and additional caution (implicit in the omission of the OECD's paragraphs 1 and 2) to be applied in relation to extension provision requires that a tax on capital gains is not included in the Taxes Covered Article in pre-CGT treaties. The one exception to this is the Austrian treaty discussed below.

The central expression in defining Australia's existing taxes is 'Australian income tax.' A counter view [F33] suggests that an ambulatory approach should be adopted in interpreting tax treaties: an ambulatory approach is now embodied in OECD Model Article 3(2) ('undefined terms provision') and the views of international writers generally support this approach. [F34] It is argued the undefined terms provision in tax treaties gives a domestic tax law meaning to undefined terms and that provision operates in an ambulatory way. Because net capital gains are now included in assessable income, it is argued that the expression 'Australian income tax' has evolved to cover capital gains.

As a matter of general principle the ATO agrees with the ambulatory approach in applying the undefined terms provision. [F35] However, as the Taxes Covered Article has specific extension provision designed to accommodate changes in taxes covered, it is clearly inappropriate to use the undefined terms provision to accommodate the evolution of 'Australian income tax'. The extension provision has that role.

The fact that there is no specific reference to capital gains tax as existing taxes in post-CGT treaties (which clearly apply to capital gains) and similar expressions are used to describe Australian income tax in pre and post-CGT treaties is also regarded as evidence that pre-CGT treaties must extend to capital gains. The ATO response to this is that when pre-CGT treaties were signed, the only taxes which could be 'existing' taxes did not include a comprehensive CGT.

A technical analysis of aspects of sections 26(a), 26AAA, [F36] and 25A indicates that at the time of signature, existing taxes may have extended to some gains which were not income according to ordinary concepts. One counter view is that the extension provision would also permit the taxes covered to extend to the CGT, because it is substantially similar to the one of the existing taxes, namely section 26AAA. In comparing section 26AAA with the CGT, what is needed is 'a comparison of the taxes concerned and not an analysis of underlying transaction'. [F37] The argument appears to be that section 26AAA taxes pure capital gains and the CGT regime merely extended the range of capital gains in the Australian income tax.

However there were many 'borderline gains' of this kind. While the ATO agrees these items may be included within the existing taxes, the question is whether CGT is identical or substantially similar to these provisions.

The Taxes Covered Article is designed to avoid the necessity of concluding a new treaty whenever the domestic tax laws of treaty partners are modified. The extension provision achieves this, although there are limits - new taxes must be at least substantially similar. The ATO considers the question of whether a tax is substantially similar is a matter of proportion, not requiring in this case an overly technical analysis of the capital/income distinction. Significantly Australia's international position (set out in its 1976 OECD Model reservation, discussed at paragraph 45) when most pre-CGT treaties were negotiated, was that 'Australia does not levy a capital gains tax'. This position was clearly not based on a technical approach to the issue. 'Borderline gains' such as those assessable under the former section 26AAA were not considered a CGT for tax treaty purposes.

Academic writings are said to confirm the counter view that where capital gains are normally dealt with in income tax laws, any new CGT will, for treaty purposes, be at least similar to income tax: Taxation of capital gains is normally dealt with in income tax laws, although in some circumstances separate legislation is devoted to that subject (see national reports in LXIb CDFI 129ff (1976)). Consequently any new capital gains tax will for treaty purposes, normally have to be considered as being at least similar to income tax; the Danish Landskatteret ('Danish Tax Court') 26 ET 114 (1986): DTC Denmark / France , differs however. . .'. [F38]

The ATO does not agree that authorities cited by Professor Vogel support his conclusion. The discussion of a decision of the Danish Tax Court in 26 European Taxation 114 (1986) actually concludes that the Court did not address the issue of whether a subsequently introduced CGT was similar to an 'ordinary' income tax. The decision was based on an agreed interpretation by the French and Danish tax authorities. Furthermore, Vogel's observation that the 'DTC Denmark / France , differs however . . .', [F39] - which seems to imply a decision contrary to the decision discussed in European Taxation - is puzzling, because this 'contrary decision' is in fact referring to the same decision of the Danish Tax Court.

Authors [F40] have also referred to the decision of Gadsden v MNR . [F41] In that case the Canadian Tax Review Board considered that the failure of the Canadian CGT to be mentioned in the Taxes Covered Article of the 1966 Canada/United Kingdom DTA did not prevent the subsequent introduction of a capital gains tax in 1972 from being picked up within the concept of 'income tax,' as Canada (similarly to Australia) taxes capital gains as part of its general income tax legislation. However, while the case may appear to be similar to the present issue, it can be distinguished. It would be necessary to take into account the complete contextual matrix and particularly in this case there were significant drafting differences between Australian treaty practice and the Canada/United Kingdom DTA. In this treaty a specific article was entitled Capital Gains and it comprehensively dealt with capital gains.

A reference has also been made to Baker [F42] who believes the United Kingdom Inland Revenue considers that where capital gains tax is introduced after a treaty was concluded capital gains taxes are drawn in by the extension for substantially similar taxes. The ATO notes that this statement is not definitive, nor is it confidently asserted. Further it is obvious, given departures from the OECD Model by many countries in relation to the Taxes Covered Article, that broad assertions, which do not take into account the context of each treaty cannot be made in this area.

It has also been argued that if the list of taxes enumerated by a treaty partner in paragraph 1 of the Taxes Covered Article includes a capital gains tax, paragraph 2 of that Article will operate to include within the taxes covered, any substantially similar taxes levied by the other treaty partner. Thus in the case of the Australia/United Kingdom DTA (where the United Kingdom lists 'the capital gains tax' in the list of taxes covered) Australia's subsequently introduced CGT could be a substantially similar tax.

This issue has been discussed mainly in the context of capital taxes, such as wealth taxes, which are often dealt with in double tax treaties of other countries. Vogel, Shannon, Doernberg and van Raad advocate this view in relation to capital taxes: The express enumeration of taxes is cumulative in this context. For example, if a tax on capital in the other contracting state is included in the express enumeration, a tax on capital introduced in the United States after the date of signature of the treaty would fall within the scope of the treaty provided it is substantially similar to the foreign tax. Similarly, taxes of the other contracting state imposed after the date of signature of the treaty that are substantially similar to enumerated United States taxes fall within the scope of the treaty. [F43]

However, the ATO considers this cannot be the case in Australian pre-CGT treaties containing distributive rules over capital. For example, the Australia/Germany DTA does not contain a provision for 'entry into force' in relation to future Australian capital taxes. If Australia was to subsequently introduce such a tax at the federal level there is no mechanism within the treaty for the treaty as a whole to take effect in relation to such taxes. Compare this also with Article 24(5) of the Australia/Norway DTC, which clearly does not accord with the view proposed by these authors, because Article 24(5) requires further negotiations if an Australian capital tax were to be introduced.

While these authors make this statement without supporting authority, support for this view is claimed [F44] in a 1985 decision of the Paris Tribunal de Grande Instance ( Rohrmoser No. 10669/84), concerning the France/Austria DTC (1959) and a 1992 decision of the Cour de Cassation (No 572 P) concerning the France/Switzerland DTC (1996) and capital taxes. Both cases dealt with the question of whether a French wealth tax introduced long after the conclusion of the relevant tax treaty was covered by that treaty, even though France had not included a wealth tax in its list of existing taxes. It was found that France's wealth tax was similar to the wealth taxes listed by Austria and Switzerland, and by applying the extension provision the subsequently introduced French wealth tax was covered by the treaty.

However, the ATO considers these cases are not relevant to the treatment of capital gains in Australian tax treaties. Both French treaties contained the equivalent of OECD Model Article 2.1, which specifically provided that those treaties would apply to taxes on capital. Furthermore those treaties both contain an article equivalent to the Capital article in the OECD Model. By contrast, in relation to capital gains, Australia's pre-CGT treaties omit OECD Model Article 2.1 and do not contain a specific article allocating taxing rights over capital gains.

It needs to be remembered that Australian pre-CGT treaties exclude paragraphs 1 and 2 of the OECD Model (as does the US and Canada). The approach of such countries is to separately list their taxes. The natural meaning of the extension provision as applicable to these jurisdictions is that the extension provision applies separately.

Capital gains taxes are covered by Article 2 of the Australia/Austria DTA. Legislation enacting the Part IIIA of the ITAA 1936 was assented to on 24 June 1986 and the Australia/Austria DTA was signed on 8 July 1986. Therefore, Part IIIA was one of the existing taxes at the time of signature. While capital gains tax is now assessed under Parts 3-1 and 3-3 of the ITAA 1997 this is regarded as substantially similar to the capital gains tax existing at signature.

However, for the contextual reasons already discussed the ATO does not consider that the distributive rules of the treaty limit taxing rights over capital gains.

Australia has a consistent policy of not including OECD Model Articles 2.1 and 2.2 in its tax treaties. The Australia/Italy DTC is the only exception in all of Australia's tax treaties (pre or post CGT), and was clearly inserted at the insistence of Italy. Article 2.1 provides that 'the Convention shall apply only to taxes on income imposed on behalf of each Contracting State irrespective of the manner in which they are levied'. There are major changes from the OECD analogue, relating to sub-national taxes (see Australia's reservation to Article 2) and capital taxes. Importantly this provision is possibly more restrictive than the OECD Model, because it applies only to income taxes. Therefore, if in the context of that Convention income does not include capital gains, by the operation of Article 2.1, the Convention does not deal with such gains.

The Business Profits Article deals with business profits of an enterprise of one Contracting State, which carries on business in the other State. The key provision is Article 7.1, which in pre-CGT treaties generally follows Article 7.1 of the OECD Model: The profits of an enterprise of one of the Contracting States shall be taxable only in that State unless the enterprise carries on business in the other Contracting State through a permanent establishment situated therein. If the enterprise carries on business as aforesaid, the profits of the enterprise may be taxed in the other State but only so much of them as is attributable to that permanent establishment.

Does the expression 'profits of an enterprise' encompass capital gains? Under the undefined terms provision, this expression can take the meaning it has under the domestic law, where the context permits.

In Thiel the High Court held that the word 'profits' could not be held to have a domestic law meaning. [F45]

Where a country distinguishes between income and capital gains, the Business Profits Article would not be expected to deal with capital gains. The OECD Model Commentary on Article 13 implies [F46] that if the domestic law of a State taxes 'capital gains' such gains will not be dealt with in the Business Profits Article, but in the 'article on capital gains'. Relevantly, Australia's reservations on capital gains were made against OECD Model Article 13 - not the Business Profits Article. While the analogy with the OECD Model cannot directly be made (as the Alienation of Property articles in pre-CGT treaties differ from the OECD's Capital Gains Article) against this international benchmark and given Australia's recognition of the income/capital gain distinction, it would not be expected that the Business Profits Article would deal with capital gains.

Further the ATO considers that taking into account the context of Australia's pre-CGT treaties discussed in detail above, capital gains are outside the scope of pre-CGT treaties and 'profits of an enterprise' would not cover capital gains.

Subsection 3(2) of the Agreements Act which equates business profits to taxable income is frequently cited in support of the view that business profits include capital gains. It provides: For the purposes of this Act and the Assessment Act, a reference in an agreement to profits of an activity or business shall, in relation to Australian tax be read, where the context so permits, as a reference to taxable income derived from that activity or business.

It is argued that subsection 3(2), operates so the term 'profits' is to be read as meaning a reference to any 'taxable income' from commercial activity. Arguably, as taxable income includes net capital gains, this could mean that capital gains are included in the definition of 'profits' in the Business Profits Article.

However the subsection requires that the link between the term 'profits' and 'taxable income' can only be made where the context so permits. [F47] The ATO does not consider the context of pre-CGT treaties permits the term 'profits' in the Business Profits Article, to be read as including capital gains - even though net capital gains are included in taxable income.

Furthermore the ATO considers the purpose of subsection 3(2) of the Agreements Act is not to define the term profits as used in treaties. The legislative history and related explanatory material demonstrate this subsection is strictly a mechanism to implement a treaty term which is foreign to Australia's taxation laws within the technical language of Australia's domestic legislation [F48] - rather than as an aid in defining a tax treaty term. Confirming this purpose subsection 3(2) is drafted to apply 'in relation to Australian tax' - not in relation to the agreement.

Alternative views may also observe that the OECD Commentary states that the word 'profits' should be understood as having a broad meaning including all income derived in carrying on an enterprise. [F49]

Against this background FCT v Slater Holdings Ltd ( No 2 ), [F50] Kenneth A Summons Pty Ltd & Ors v FCT , [F51] and MacFarlane v FCT [F52] may be cited as providing evidence that the term 'profits' is to be read as including capital gains. However the ATO notes that these cases deal with the meaning of profits in relation to dividend payments and address the interface between the taxation law and corporations law concepts. Spanish Prospecting Company Ltd , [F53] might also be advanced as authority for the view that profits could apply to both income or capital gains, but again the context of this case, involving payments to creditors, differs from the treaty context. The ATO does not see these decisions as having application to this issue.

Advocates of the alternative view also argue that the concepts of income, profits and gains have been used interchangeably in Australia's pre and post-CGT treaties. Gains includes capital gains. Therefore it is argued that profits of an enterprise should be read as including capital gains. Examples cited in support of this are: * The credit article in both pre and post-CGT treaties gives credit, on the Australian side, only for taxes on 'income'. There is no mention of 'profits or gains' pre or post-CGT. Clearly credit is to be given for tax on capital gains in post-CGT treaties. For the credit rules to work post-CGT, the concept of 'income' must include 'gains'. Therefore in all Australian tax treaties the word 'income' can extend to 'gains'. (The technical aspects of this argument are questionable - see footnote.) [F54] * In post-CGT treaties the expression 'gain of a capital nature' is used in the residual rule in the Alienation of Property Article and other parts of the article refer to 'income, profits or gains'. It is argued that, because it is necessary to make the 'gains of a capital nature' qualification the earlier use of the word 'gain' cannot be used to contrast with 'income'. [F55] * Some pre-CGT treaties deal with income from the alienation of capital assets. It is argued [F56] that it is difficult to see how one could derive 'income' from capital assets unless 'income' encompassed capital gains.

The ATO agrees that the words 'income', 'profits' and 'gains' may be used interchangeably within a treaty. [F57] The ATO considers that the meaning of these expressions in particular treaties is influenced by their context. Even if the first example was technically correct, in say a post-CGT treaty, given the context of that treaty it may possibly be the case that a reference to 'income' could include 'capital gains'. However for a pre-CGT treaty the context may well require that the word 'income' be read as not including capital gains.

As for the second example again the ATO accepts that the words 'income', 'profits' and 'gains' may be used interchangeably. However, the reason for the qualification in the sweep-up provision arises from the fact that the gains dealt with by articles in post-CGT treaties do have a wider coverage than in pre-CGT treaties. The sweep-up is only dealing with capital gains (and not income gains) and so it was necessary to describe that class of gains accordingly.

The third example is discussed in detail below. At this stage it is noted that from the Australian perspective depreciation recapture is an example of 'income from alienation of capital assets'.

In the Australia/Netherlands DTA, Item 6(a) of the First Protocol to that agreement provides: Where one of the States is entitled to tax the profits of an enterprise, that State may treat as profits of an enterprise , profits from the alienation of capital assets of the enterprise , not being profits that consist of income to which paragraph (1) of Article 13 applies. [Emphasis added]

Whether a Contracting State can tax the 'profits of an enterprise' is determined under the Business Profits Article. The Contracting State in which the enterprise or the person conducting the enterprise is resident always has a taxing right under the Business Profits Article over the profits of that enterprise. The other Contracting State acquires a primary taxing right over such profits where the enterprise has a permanent establishment in that State.

It is sometimes argued [F58] that this provision explicitly demonstrates that profits from the alienation of capital assets (which might be considered to encompass capital gains) fall naturally within the concept of 'profits of an enterprise'. On the other hand, the provision could equally be interpreted to the effect that without the provision 'profits of an enterprise' would not include profits from the alienation of capital assets. Furthermore, in Lamesa [F59] the Federal Court commented that this treaty was concerned only with taxes on income, having no direct concern with capital gains.

In any event, the ATO does not consider that in this context 'profits' should be read as widely as meaning all 'capital gains' from the alienation of capital assets. The context of the treaty and the scope of the taxes covered under the treaty need to be taken into account. At the time the Netherlands treaty was concluded, sections 26(a), 26AAA and 59 of the ITAA 1936 potentially assessed gains on and recoupment of depreciation on certain alienations. Therefore the Protocol provision is dealing only with those 'borderline' classes of income assessable at the time the Protocol was concluded.

Most of Australia's pre-CGT treaties only deal with 'income' from the alienation of property. The ATO considers this expression was carefully chosen, to exclude capital gains. By contrast the OECD Model deals with (capital) gains. A number of pre-CGT treaties also mention 'gains' and the ATO considers that against the contextual backdrop of Australian pre-CGT treaty practice, this is a reference only to revenue gains - it does not extend to capital gains.

The use of 'income' demonstrates a consistent intention by Australia to deal with items that were income or assimilated to income and not to deal with capital gains as such in this article.

As noted earlier Australia consistently avoided using the OECD heading of ' Capital Gains'. In pre-CGT treaties the broader expression ' Alienation of Property' was used because the distributive rules of this article are generally expressed to apply only in relation to income from alienation of certain property. By contrast the distributive rules of the OECD Model Capital Gains Article apply to (capital) gains arising from the alienation of all property.

The question arises as to why Australia agreed to a separate article on income from alienation of property if the business profits or residual article already dealt with such income?

Many treaty partners (especially OECD members) would have wished to have a Capital Gains Article similar to OECD Model. Australia could accommodate this objective (with some reluctance as the treaty practice shows) by referring only to income in an Alienation of Property Article. From the Australian perspective some alienations which were treated as income by Australian tax law (see for example sections 26(a), 26AAA and 59 of the ITAA 1936) would potentially be dealt with under this article. With the exception of alienation of real property discussed below, taxing rights would be the same as if these particular alienations had been dealt with under the Business Profits Article. For presentation reasons Australia could agree to Alienation of Property articles dealing with income .

For treaty partners, on the other hand, the 'presentational' integrity of using text similar to the OECD Model had been maintained. Furthermore, jurisdictions which do not strongly adhere to the income/capital distinction would have been accepting of the use of the word 'income' because from their perspective it could potentially embrace a wider class of gains from the alienation of property. [F60]

Also there were advantages to Australia's source country interests in dealing with revenue gains on alienation of real property on the same basis as the OECD's Capital Gains Article. [F61] Given the significance to the Australian economy of real property (as defined) there would have been a strong imperative to secure source country taxing rights over such alienations. Australia's drafting approach fulfilled this objective by expressly dealing with income from alienation of real property in this provision, because it better protects source country taxing rights over real property than would the Business Profits Article. [F62] Furthermore, this provision expands the meaning of real property and defines the situs of certain interests in land.

There are some isolated departures from usual Australian treaty practice at the time, regarding the use of the terms 'income or gains' is used. Clearly these changes were at the instigation of the treaty partner and give rise to an important question as to whether the term 'gains' means income gains or capital gains or both. The following table summarises these departures.

The use of the word 'gain' was at the instigation of treaty partners for the treaty partner's purposes. Significantly, except in the case of 'unilateral' provisions dealing with disposal of shares in subsidiaries (Netherlands, Norway and Italian treaties) and alienation of ships and aircraft in the Norwegian Convention (discussed below), the expression 'gains' is never used without also referring to 'income'. In the case of Ireland, gains are defined (at Article 14(2)(a)) only in relation to Ireland.

The ATO considers that for the contextual reasons discussed at the outset, from the Australian perspective 'gains' as used in these provisions does not refer to capital gains, but is confined to income gains, or gains assimilated to income - such as those assessable under section 26(a), 26AAA and 59 of the ITAA 1936. In particular it is noted that under the credit articles of pre-CGT treaties, Australia relieves double taxation only in respect of 'income'. Revenue gains readily fall within that description. However, if from the Australian perspective, 'gains' in the Alienation of Property Article had been intended to extend to 'capital gains' it would be expected that the credit article would have specifically referred to such gains. (Although the credit articles of post-CGT treaties refer only to 'income', post-CGT changes to the source rules and Australia's foreign tax credit rules had the effect of dealing with capital gains in the credit article - see footnote 54.)

Article 21 of the Australia/Norway DTA is the only case where the word 'gain' alone is used in a bilateral way. Records of the negotiations for the Norwegian treaty show Norway was concerned to ensure that taxing rights over alienation of shipping was expressly dealt with. (Given Norway's extensive maritime interests this is not surprising.) From the Australian perspective the records indicate the gains were seen as dealing with depreciation recapture or section 26(a) type profits.

On one occasion (the Australia/Malaysia DTA) the relevant article, which deals only with real property interests, refers to 'income or profit'. For contextual reasons the ATO considers that profit does not extend to capital gains.

The alternative view proceeds on a basis that the word income is used interchangeably with capital gains. [F63] Other aspects of this issue are discussed in relation to Alienation of Property .

Some pre-CGT treaties limit taxing rights over 'income' (and 'gains' in the case of the Australia/Ireland DTA) from the alienation of capital assets. The alternative view argues that it is difficult to see how income could be derived from the alienation of capital assets unless the term income included capital gains.

However, the ATO considers this overlooks the context in which these treaties were negotiated and particularly the fact that under sections 26(a), 26AAA and 59, 'income' on the borderline of the income/capital distinction could be derived from the alienation of capital assets. On this basis, Dawson J's comments in Thiel , can be reconciled with both views. After concluding that the taxpayer's activities confirmed what he did was by way of an adventure in the nature of trade and had the requisite business character he states: [F64] This conclusion makes it apparent that the applicable article of the Swiss Agreement is Art 7 rather than Art 13. Having regard to the nature of the appellant's activity, it would clearly be inappropriate to regard his gain as being by way of income from the alienation of capital assets. Necessarily, the nature of the enterprise upon which the appellant was engaged did not involve the acquisition of capital assets.

It has been argued that the 'residual income rules' in the Income Not Expressly Mentioned Article [F65] apply to capital gains. This residual article provides a general rule relating to income not dealt with in the previous articles of the treaty. It deals both (a) with types of income not dealt with elsewhere and (b) with income arising in third states. Not all pre-CGT treaties contain this provision.

Article 21 of the Australia/United States DTC is typical: (1) Items of income of a resident of one of the Contracting States which are not expressly mentioned in the foregoing Articles of this Convention shall be taxable only in that State. (2) However, if such income is derived by a resident of one of the Contracting States from sources in the other Contracting State, such income may also be taxed in the State in which it has its source. (3) The provisions of paragraph (1) shall not apply to income derived by a resident of one of the Contracting States which is effectively connected with a permanent establishment situated in the other Contracting State. In such a case, the provisions of Article 7 (Business Profits) shall apply.

The ATO considers, for the contextual reasons discussed at the outset, that it was not intended to deal with capital gains at all in pre-CGT treaties. Specifically it means that Australia did not intend the expression 'items of income' in the residual income rule to include capital gains.

An alternative view might be that the residual article in the pre-CGT treaties that include such a provision was intended to act as a 'catch-all' provision and should be read widely - arguably to cover those capital gains not dealt with elsewhere in these pre-CGT treaties as 'items of income'.

Vogel appears to take this view. In a discussion relevant to the Australia/Germany DTA (see later discussion on pre-OECD treaties) Vogel observes that in the absence of an express capital gains rule, such gains could be dealt with by the residual article. [F66]

Furthermore for similar reasons discussed in relation to business profits, it could be said that the terms 'income', 'profits' or 'gains' have been used interchangeably and on that basis income includes capital gains.

In the case of the Australia/United States DTC, the relevant USA authorities appear to agree that this Article is capable of dealing with capital gains. The US Treasury Technical Explanation of the DTC, for example, states that: . . . [g]ains with respect to other property are covered by Article 21 (Income Not Expressly Mentioned), which provides that gains effectively connected with a permanent establishment are taxable where the permanent establishment is located, in accordance with Article 7 (Business profits) and that other gains may be taxed by both the State of source of the gain and the State of residence of the owner. Double taxation is avoided under the provisions of Article 22 (Relief from Double Taxation). [F67]

However, this was not the view adopted by Australia. The Explanatory Memorandum to the legislation which enacted the Australia/United States DTC, [F68] did not make a similar observation.

It is interesting to note that US Letter Ruling LTR199918047 which deals with the treatment of gains from the sale of a foreign subsidiary, touches upon the US authorities' views on the treatment of gains on the sale of shares in an Australian subsidiary under the Australia/United States DTC. [F69] Essentially the Letter Ruling implies that if the gain is dealt with by the residual article of the treaty, then in the particular circumstances outlined in the Letter Ruling, the source of the gain is in Australia.

While the US Technical Explanation is a relevant consideration in the contextual matrix, against the background of the overall context of Australia's pre-CGT treaties and in the absence of any agreement on the Australian side, less weight should be given to the Technical Explanation. Nevertheless, in accordance with Article 24 of the US DTC, the Australian competent authorities will undertake mutual agreement procedures with the United States.

Australia's earliest treaties with the United Kingdom (1946), United States (1953), Canada (1958), and New Zealand (1960) were strongly influenced by the 'Colonial Model' used by the British Empire and Commonwealth. [F70] Elements of this Model can still be found in the drafting of the United Kingdom (1967); Japan (1969); Singapore (1969); Germany (1972); and New Zealand (1972). All of these treaties have since been renegotiated except United Kingdom (1967); Japan (1969); and Germany (1972). Australia's treaties concluded in the late 1960s and early 1970 were clearly moving closer to the OECD Model, but important aspects of the Colonial Model remain.

The Colonial Model defined a concept of 'industrial and commercial profits', (the analogue to the OECD's undefined expression 'profits of an enterprise'). This definition varies slightly from treaty to treaty. It generally excluded certain specific items which were often (but not always) expressly dealt with under other distributive rules. Following this practice, Australia's current treaties with the United Kingdom and Japan define 'industrial and commercial profits'.

Under the Colonial Model it was common to delete items from distributive rules if the source country retained taxing rights over the item. [F71] There was no OECD Capital Gains Article - even though the United Kingdom and Japan were OECD members. (Likewise, these treaties omitted the prevailing OECD Immovable Property , Capital [F72] and Income Not Expressly Mentioned [F73] Articles.)

The Australia/Germany DTA signed shortly after Australia joined the OECD, but negotiated before Australia's membership omits the OECD Capital Gains , and residual articles. An abbreviated version of the OECD Immovable Property Article is included.

Obviously, issues relating to whether capital gains are dealt with by the Capital Gains / Alienation of Property and residual articles are not relevant to these treaties.

The Australia/United Kingdom DTA uses the expression 'industrial or commercial profits' which was defined in Article 5(7) to mean ' income derived by an enterprise from the conduct of a trade or business, including income from the furnishing of services of employees or other personnel . . .'. In this context does 'income' include 'capital gains'? [F75]

The alternative view argues that since the title and preamble to the United Kingdom treaty is expressed to apply to capital gains and the taxes covered include UK capital gains tax, and there is no Capital Gains or residual article, then industrial or commercial profits should include capital gains. It is argued that income and gains are used interchangeably in pre-CGT treaties and should be given a broad interpretation. [F76]

As was discussed above, the ATO considers that in the pre-CGT treaties, the term 'income' was used to specifically exclude capital gains taxation from the scope of pre-CGT treaties. The contextual support for this is especially strong in the case of the United Kingdom, because of the omission of an Alienation of Property / Capital Gains Article. [F77]

Given the common legal backgrounds of Australia and the United Kingdom it is even more likely the term was carefully chosen, so as not to embrace capital gains. Evidence that the two territories adhered to the income/capital distinction in the treaty is in the title and preamble, which refer to both 'income and capital gains'. Consequently, the specific definition of the term 'industrial or commercial profits' to mean 'income' in this treaty is an express indication that the CGT is not intended to be dealt with as industrial and commercial profits Also arguments relying on the application of subsection 3(2) of the Agreements Act are difficult to apply to this definition, because subsection 3(2) refers to 'profits' of an activity or business - not 'income'. There are other indicators concerning the application of the credit rules (discussed below) that confirm that it is not appropriate to regard capital gains as included in this definition.

The Australia/Japan DTA also uses the expression 'industrial or commercial profits', but defined it in Article 4(5) to mean ' profits derived by an enterprise from the conduct of a trade or business [but excluding dividends, interest, royalties, shipping and aircraft income and personal services income]'. Again the preliminary issue is whether the profit has been derived by an enterprise from the conduct of a trade or business. But in any event the ATO does not consider capital gains are included within the meaning of the word 'profits' for reasons similar to those mentioned in the earlier discussion in relation to 'profits of an enterprise' as used in post-OECD treaties.

The Australia/Germany DTA uses an abbreviated OECD style Business Profits Article and the above discussion in relation to post-OECD applies similarly. For similar contextual reasons the ATO does not consider that capital gains are dealt with in this provision. In this case, the context is arguably more emphatic, because of the absence of a Capital Gains / Alienation of Property Article in the treaty. Significantly, the German author, Vogel confirms no distributive rule applies to capital gains. He considers that 'taxation of capital gains . . . continues to lack a specific rule and the domestic law of the contracting States consequently applies in this respect without any restriction'. [F78]

The following discussion suggests the credit article of the United Kingdom treaty could potentially deal with capital gains. However the ATO considers that for reasons set out in Australia's Explanatory Memorandums to the enacting legislation, that the credit article was designed to deal only with United Kingdom chargeable gains which were then assessable income in Australia (for example, under the former section 26(a)).

It appears that the Methods of Elimination of Double Taxation Article (often referred to as the 'credit article') was designed to allocate taxing rights over income not expressly mentioned in the distributive rules of the treaty. [F79] Article 19 of the 1967 Australia/United Kingdom DTA reads: (1) . . . (a) Australian tax payable under the laws of Australia and in accordance with this Agreement, whether directly or by deduction, on profits, income or chargeable gains from sources within Australia . . . shall be allowed as a credit against any United Kingdom tax computed by reference to the same profits, income or chargeable gains by reference to which the Australian tax is computed; and (b) . . . (2) (a) . . . United Kingdom tax payable under the laws of the United Kingdom and in accordance with this Agreement . . . whether directly or by deduction, on income derived by a resident of Australia from sources in the United Kingdom . . . shall be allowed as a credit against the Australian tax assessed by reference to the same income by reference to which the United Kingdom tax is payable.

For Article 19 to apply to an item of income, the only requirement is that the income (also profits or 'chargeable gains' in the case of the United Kingdom) is from sources [F80] within the relevant territory. It is not strictly necessary to find a distributive rule specifically allocating profits to the country of source. [F81]

A similar approach has been used in all Australian treaties since the 1946 United Kingdom treaty. The Explanatory Memorandum accompanying the enactment of Australia's this treaty and subsequent treaties (including the 1967 Australia/United Kingdom DTA) confirms it was commonly understood at that time that the credit rules were to deal with income not dealt with by the distributive rules. [F82]

This Article requires the country of residence to give credit for source country taxes on income not specifically dealt with in the distributive rules. For example, because real property income is not covered by the distributive rules, under Article 19 the residence country will credit source country taxation - a similar result if a standard Real Property Article was specifically included. Thus, it was unnecessary to include such a provision in the United Kingdom (and Japanese) treaties.

However, as drafted this credit rule does not adequately distribute taxing rights if (say) real property income was derived from third countries. As the credit rules do not do this, the negotiators developed a unique Dual Resident Article that allocates taxing rights over the third country income of dual residents. In modern OECD style treaties, some specific distributive rules (such as the Business Profits Article) deal with third country income with the residual article dealing with all other third country income. The Dual Resident Article is only used in Australian treaties where there is no residual article. Inclusion of this article confirms the Australian negotiators recognised the credit article was to deal with the balance of residual income not specifically mentioned in the distributive rules.

However, Article 19 contains a requirement contained in the prevailing OECD Model which first entered Australian tax treaty practice with the 1967 United Kingdom treaty. It requires that a credit must be provided by both the United Kingdom and Australia 'in accordance with this Agreement'. Strictly speaking the text of the 1967 UK treaty does not prescribe the treatment of real property income or other residual income and it might be argued therefore that credit could not be required in such circumstances. But against the contextual background set out above, this qualification should be read more in keeping with the purpose of the provision as 'not contrary to this Agreement'.

Accordingly, it might be argued that capital gains, which are not dealt with by the distributive rules, are dealt with in Article 19 of the treaty. If a United Kingdom resident makes an Australian-sourced capital gain, it might be argued that Article 19(1)(a) provides an unrestricted source country taxing right, and requires the United Kingdom to allow a credit in respect of such 'chargeable gains'.

However, the language of Article 19(1)(a) and 19(2)(a) and the accompanying Explanatory Memorandum suggests that at the time of conclusion of the treaty it was intended that only 'chargeable gains' that were also treated as assessable income by Australia when the treaty was concluded (eg under the former section 26(a)) would come within the ambit of Article 19. First, under paragraph 19 (a), the UK was to provide credit for Australian tax or 'profits, income or chargeable gains'. On the other hand, Australia was to provide credit only in respect of 'income'. [F83] There is no reference to 'capital gains' on the Australian gains.

Secondly, the relevant Explanatory Memorandum confirms the Australian Government considered the provision would only apply to income or gains assimilated to income. It was to apply only for the purpose of giving credit for United Kingdom capital gains taxes in relation to receipts that would be treated as income by Australia. The Explanatory Memorandum says: As the agreement is drafted, it will in practice have little effect in relation to the taxes imposed on capital gains by the United Kingdom. The principal effect will be in relation to article 19 [the credit article]- see notes on that provision. On the Australian side, the Agreement applies only to income tax imposed by the Commonwealth, the withholding tax on dividends and interest being within the description 'income tax.' [F84] again, at p 52: A credit for Australian income tax would be allowable against United Kingdom capital gains tax in any case in which a particular gain is treated as income in Australia (for example, profits from the sale of Australian real estate purchased for purposes of resale at a profit) but for United Kingdom tax purposes is treated as a capital gain and charged to capital gains tax.

Accordingly the ATO does not consider that the treaty generally covers capital gains. Note this is consistent with the ATO's view of the operation of all pre-CGT treaties regarding capital gains.

The same principles apply in interpreting the Australia/Japan DTA. There is a clear structural relationship between the United Kingdom treaty and the Japanese treaty concluded two years later. The Explanatory Memorandum to the Income Tax (International Treaties) Bill 1969 confirms the common approach of this treaty with the United Kingdom and other earlier treaties. In common with the Singapore Agreement, the agreement with Japan is broadly along the lines of the treaties negotiated with the United Kingdom, the United States, Canada and New Zealand. [F85]

Technically it is also arguable that the credit rules in the German treaty are capable of dealing with all residual income although the accompanying explanatory material does not supply the contextual support for this interpretation. While technically the same may be true of all subsequent treaties as they all use a similar credit rule, for the contextual reasons discussed at the outset, the ATO does not consider that the credit rules in the German and later pre-CGT treaties generally deal with capital gains.

A related issue is the title and preamble to the Australia/United Kingdom DTA which specifically directs that their purpose is the avoidance of double taxation and the prevention of fiscal evasion in respect of 'income and capital gains'(emphasis added). The same issues arise with the Australia/Ireland DTA.

The Federal Court in Lamesa may have been referring to this aspect of the Australia/United Kingdom DTA when it said: [u]nlike more recent treaties, the [Netherlands] Agreement is concerned only with taxes on income. It has no direct concern with capital gains: cf the double tax agreement between the United Kingdom and Australia which refers specifically both to taxes on income and capital gains. [F86]

However, notwithstanding the reference to capital gains in the title and the preamble, the ATO considers that this reference does not imply that these treaties generally deal with capital gains.

It is a United Kingdom speciality, found in its treaties generally, to include 'capital gains' in the title and preamble to deal with the specific description of taxes covered and the nature of the its capital gains tax rules. The United Kingdom probably refers to capital gains because for individuals it imposes a separate tax on capital gains. As explained there are arrangements in the credit rules to cover those items, which at the time were included in assessable income in Australia but would be chargeable gains in the United Kingdom. For this reason and consistently with its treaty practice there is a reference to capital gains in the title and preamble.

There is no similar explanation in the Explanatory Memorandum for the Australia/Ireland DTA, but the similar construction of the credit rule and also the definition of the credit article in that treaty, coupled with a definition of chargeable gains only for Irish purposes suggests similar reasoning can be applied to the Australia/Ireland DTA.

Below is a detailed table of contents for this draft Taxation Ruling: Paragraph What this Ruling is about 1 Ruling 4 Definitions 6 Pre - CGT treaties and post - CGT treaties 6 Pre - OECD Treaties 9 Borderline gains 10 Glossary 12 Date of effect 13 Overview 14 Overview of ATO Position 15 Alternative Views 23 Taxes Covered Article 26 Business Profits Article 29 Alienation of Property Article 32 Residual Article 34 Pre - OECD Treaties 36 Explanations 38 The Issues 38 Context 40 Australia's OECD Reservation 45 OECD Model and Australian Treaty Practices and Policies 48 Post CGT : Changes to the Alienation of Property Article 57 Post CGT : other changes 58 Treaty Partners : treaty practice and domestic law 63 Political , economic and diplomatic background 66 Context - Conclusion 71 Taxes Covered 73 ATO View 76 Alternative View - existing taxes include capital gains 83 Alternative view: capital gains substantially similar to existing taxes 86 Alternative view: new tax substantially similar to the existing taxes of other State 93 Austria 99 Italy 101 Business Profits Article 102 ATO view - capital gains not ' profits of an enterprise' 104 Alternative Views 107 Subsection 3 ( 2 ) of the Agreements Act 107 Profits of an enterprise include capital gains 111 Were the terms ' income' , ' profits' or ' gains' used interchangeably ? 113 Australia / Netherlands : Protocol Item 6 ( a ) 117 Alienation of Property Article 121 ATO View 121 ATO view - ' Income from the alienation of property' 122 ATO view - Why address income from alienation of property separately from the Business Profits Article ? 124 ATO view - ' gains' and ' profits' from the alienation of property 128 Alternative view - income ( or gains ) from the alienation of capital assets 133 Residual Article 136 ATO view 138 Alternative view 139 United States treaty 142 Pre-OECD treaties: United Kingdom, Japan and Germany 146 Background 146 ' Industrial and commercial profits' : United Kingdom and Japan 151 United Kingdom 151 Japan 155 Business profits : Germany 156 United Kingdom : Does the credit article deal with capital gains ? 157 Dual Resident Article confirms credit rules deal with residual income 162 ' in accordance with this Agreement' 163 Nevertheless , context shows capital gains not generally covered 164 Japan : credit article 168 German and subsequent treaties 169 Title: United Kingdom and Ireland 170 Detailed contents list 175 Your comments 176

If you wish to comment on this draft Ruling, please send your comments promptly by 29 September 2000 to : Contact officer details have been removed following publication of the final ruling.