Taylor, C John --- "An Old Tax is a Simple Tax: A Back to the Future Suggestion for the Simplification of Australian Corporate-Shareholder Taxation" [2006] JlATaxTA 4; (2006) 2(1) Journal of The Australasian Tax Teachers Association 30
AN OLD TAX IS A SIMPLE TAX: A BACK TO THE FUTURE SUGGESTION FOR THE SIMPLIFICATION OF AUSTRALIAN CORPORATE-SHAREHOLDER TAXATION
C JOHN TAYLOR[*]
The system of corporate-shareholder taxation adopted in the Income Tax Assessment Act 1915 (Cth) was both elegant and simple. Dividends paid out of current year income were deductible to the company and assessable to shareholders. Where dividends were paid out of taxed retained income, shareholders received a credit (but not a gross up) for corporate tax paid on the distributed income. The system was generally consistent with vertical equity, provided equivalent treatment for debt and equity and for different forms of business organization, did not discriminate against nonresident investors and produced capital import neutrality at both the company and shareholder level. The costs of the system were largely borne as administrative costs of government. This paper argues that simplification benefits could be produced if an optional system of corporate-shareholder taxation based on the 1915 model were introduced for private companies with only resident shareholders. The paper explains how such a system could be constructed to produce equivalent results for these companies to those produced under the current system and examines whether adoption of the proposed system would involve a breach of Australia’s double tax agreement obligations. It details the design features of the proposed optional system and discusses whether its adoption would facilitate further simplification for companies in areas such as the debt/equity and thin capitalisation rules. The paper reflects on the impact that Australia’s international obligations have had on the complexity of domestic tax design. The concluding section argues that a new international consensus that allows countries greater flexibility in design of corporate-shareholder tax systems is desirable given the challenges posed by financial engineering and globalisation.
I INTRODUCTION
Film historian Kevin Brownlow titled his 1968 tribute to the silent film, The Parade’s Gone By. The title came from an incident reported to Brownlow by Monte Brice a writer and director of silent comedies. While on the set of the 1957 biopic The Buster Keaton Story Brice at one point had the temerity to suggest that a particular scene wasn’t being filmed the way it would have been in the 1920s. He was told: ‘Look, why don’t you go away? Times have changed. You’re an old man. The parade’s gone by.’ Those of you who have had the dubious pleasure of viewing The Buster Keaton Story might agree with me that it could not have been any worse if its makers had paid more attention to Brice.
I relate the incident because it not only reflects my eight year old son’s comments whenever I tell him how things used to be when I was a child, ‘that was the past; this is the future’, but also because it reminds me of the reactions that I receive (at best polite silence, at worst incredulity) when I say I am researching the history of corporate-shareholder taxation. In this context, my defence for looking at history is the same as Brice’s would have been, the product we have now could not be made any worse by looking at how things used to be. An understanding of where we have come from and of how we got to where we are can assist in remembering otherwise long forgotten alternatives and in identifying the causes of the constraints on our current policy choices. Having said this, before looking at the past, I will try to explain why I think this exploration is particularly relevant to the current tax reform debate in Australia.
II THE CURRENT TAX REFORM DEBATE IN AUSTRALIA
In Australia we are currently engaged in yet another debate about tax reform. The Board of Taxation recently reported to the Treasurer identifying those provisions in the Income Tax Assessment Act 1936 (Cth) (‘ITAA36’) and the Income Tax Assessment Act 1997 (Cth) (‘ITAA97’) which are no longer operative and for that reason could be repealed.[1] As a result, the Tax Laws Amendment (Repeal of Inoperative Provisions) Act 2006 (Cth) has just been enacted. The Treasurer has established a Task Force on how the regulatory burden on business, including those burdens imposed by Taxation Provisions (hereafter called the ‘Red Tape Project’), could be reduced.[2] The Task Force’s report was released on 7 April 2006.[3] The Australian Tax Research Foundation has held a conference in 2006 which has examined how Australia expresses the tax base in legislation.[4] The Taxation Institute of Australia is researching ways in which simplification of tax policy can lead to reduced compliance costs with emphasis on the identification and repeal of redundant provisions.[5] The first in an anticipated series of reports for the Taxation Institute of Australia (‘TIA’) on this issue was released on 22 June 2006.[6] In contrast with previous attempts to simplify language without simplifying policy and flirtations, via Option Two and the Tax Value Method, with more radical policy changes, the focus of many of the current projects (at the technical level at least) is on simplification in the organisation, approach and structure of the legislation rather than its language or its fundamental policy underpinnings.
The literature on tax complexity distinguishes between legal and effective simplicity. The legal simplicity or complexity of a tax law is determined by the ease or difficulty with which it can be read and understood. Tran Nam suggests that legal simplicity/complexity depend on: (a) the comprehensibility of the language used to express the law; and (b) content of the law. Tran Nam regards the content of the law as encompassing such matters as: the tax base, discretions, uncertainties, exemptions, special concessions, allowable deductions, rebates and multiple tax rates.[7] Effective simplicity measures the value of resources expended by society in raising a given amount of revenue. Raising a given amount of revenue involves ascertaining how the tax laws apply to a given set of circumstances.[8] As this process potentially involves taxpayers, tax administrators, judicial officers, and legislators, to measure the effective simplicity or complexity of a tax system, regard must be had to the costs borne by each of these potential participants.[9] As Tran Nam has noted, effective simplicity encompasses legal simplicity but is additionally affected by:
1 the number of taxpayers and tax administrators; 2 the size distribution of taxpayers; 3 the business cycle; and 4 the general level of tax avoidance and tax evasion in the economy and the government’s commitment to combat these.[10]
Many of the current projects, particularly the proposal to repeal inoperative provisions, appear to aim to improve the legal simplicity of Australia’s tax legislation by reducing its sheer bulk. Others, such as parts of the TIA’s submission to the Red Tape Project, aim to reduce effective complexity through the use of de minimis rules.
This paper is intended to be a contribution to this debate and is written on the basis that in the current Australian context achievable tax reform will neither be directed at fundamental shifts in tax policy, nor will it be a redrafting exercise, rather it will be concerned with ways in which the organisation and structure of the legislation can be changed so as to reduce its legal and/or effective complexity.
The major thesis of the paper is that many of the complex features of Australia’s current system of corporate-shareholder taxation are the products of the choice to not fully extend imputation benefits to nonresident shareholders and the constraints of the international tax environment. The result is a system of corporate-shareholder taxation that is more complex than it needs to be for those Australian companies which do not have nonresident shareholders. If those companies could be given the option of using a simpler form of corporate-shareholder taxation, it is likely that there would be improvements in the legal simplicity of the tax system for those companies and improvements in the overall effective simplicity of the system.
What then would be the features of a simpler system of corporate-shareholder taxation designed for companies which have no nonresident shareholders? In answering that question it made sense to me to look at the Australian approach to corporate-shareholder taxation before this discriminatory policy was in place and before the international tax system was so constrained. Hence Section II of this paper examines the Australian corporate-shareholder tax system in the period from 1915 to 1923 after which it was changed significantly as a result of efforts to harmonise the collection of Federal and State income taxes. Section III of this paper then outlines the current Australian system of corporate-shareholder taxation. Section IV examines the effect of international and revenue considerations on the transformation of the simple system of 1915-23 to the complex system that prevails in 2005. Section V of this paper examines some data on private companies in Australia in an effort to estimate the number of private companies that do not have nonresident shareholders. Section VI then identifies the major domestic and international tax policy objectives that appear to underpin the current Australian system. Section VII agues that for private companies with no nonresident shareholders the domestic tax policy objectives of the current dividend imputation system identified in Section VI could be achieved more simply by allowing those companies the option of using a dividend deduction system. Section VIII then examines whether giving such companies this option would infringe Australia’s Double Taxation Agreements (‘DTAs’). The paper concludes in Section IX by reflecting on the impact that Australia’s international obligations, primarily through DTAs, has had on the complexity of domestic tax design. The concluding section argues that a new international consensus that allows countries greater flexibility in design of corporate-shareholder tax systems is in fact desirable given the challenges posed by financial engineering and globalisation.
III THE SIMPLE PAST: AUSTRALIAN FEDERAL CORPORATE-SHAREHOLDER TAXATION 1915 TO 1923
From its inception in 1915 until 1923 the federal income tax used a dividend deduction system of corporate-shareholder taxation. Amendments introduced in 1916 limited the deduction to distributions of ‘assessable income’ which did not include items such as foreign source income and capital gains.[11] Capital gains were not included in the corporate tax base[12] but dividends funded from domestic source capital profits were assessable to shareholders at marginal rates.[13] Hence, capital gains and other domestic preferences were washed out on distribution. In relation to distributions of current year income there was no need for inter-corporate dividend relief measures as the effect of the deduction for dividends was that distributions of current year income were not subject to corporate tax.
At the time, Australia adopted a wholly territorial approach to income taxation with foreign source income not being subject to Australian tax.[14] In contrast to the treatment of domestic preferences, the portion of a dividend that represented a redistribution of foreign source income was not taxable to a shareholder.[15] The result was that the system produced capital import neutrality at both the resident company level and at the underlying shareholder level. Equivalent treatment was also given to investment in resident companies with income that had been subject to foreign tax and investment in foreign companies. In addition, foreign source income that was not taxed to a resident company could be passed through to nonresident shareholders free of Australian tax.
Retained income was taxed at the company level. In 1918 the dividend deduction system was supplemented by a form of imputation system in relation to distributions of previously taxed income. The Commissioner could permit companies to charge any income tax that they had paid pro-rata against shareholders.[16] From 1918 a rebate was allowed to both resident and nonresident shareholders for dividends funded from previously taxed corporate income. The rebate was limited to the lesser of tax at the corporate rate and tax at the shareholder’s rate on income from property.[17] The rebate was calculated by the Australian Taxation Office on the basis of the paying company’s income tax return. This procedure meant that rebates were never given in respect of income which had not been subject to corporate tax thus avoiding the problem of super-integration. The same rebate provisions applied to both corporate and non-corporate shareholders. In determining the tax rate applicable to the shareholder the shareholder’s assessable income was not grossed up for corporate tax paid or by the amount of the rebate. The failure to gross up and the limitation of the rebate meant that the total corporate and shareholder tax on distributions of previously taxed income did not reflect the shareholder’s marginal rate. The limitation on the rebate meant that, for shareholders on marginal rates below the corporate rate, in effect, a dividend exemption system applied in relation to distributions of income that had been subject to Australian corporate tax.
A defect in the system throughout the period was the significant disparity between the corporate rate and the top marginal rate. For the year ending 30th June 1916 the corporate rate was 7.5 per cent while the top marginal rate on income from property was 25 per cent.[18] By the end of the period in which the dividend deduction system operated, the year ending 30 June 1923, after taking into account the effect of provisions imposing additional tax, the top marginal rate on income from property was 38.375 per cent.[19] The rate of tax imposed on undistributed profits of a company was 12.08 per cent.[20] Because of the disparity of rates between the top marginal rate and the corporate rate, up to 1922 what was, in effect, a system of shareholder allocation by the Commissioner operated when a company had not made a sufficient distribution of taxable income in the year.[21] In addition, the company remained liable for tax on its retained profits. The 1922 consolidation, Income Tax Assessment Act 1922 (Cth), imposed additional tax on the company, equal to the excess of the tax that would have been payable by shareholders had there been a sufficient distribution over the tax paid by the undistributed income.[22]
Between 1915 and 1922, partnerships[23] and trustees[24] were taxed at the entity level. Partners were also assessed on their individual interest in partnership income[25] and beneficial interests’ income derived under a trust formed part of a beneficiary’s income.[26] When income was distributed partnerships and trustees were allowed a proportionate rebate of the tax paid at the entity level.[27] The result was substantially equivalent to the treatment of companies with the exception that distributions of capital profits were not taxed at the partner or beneficiary level. The system for taxing partnerships and trusts was changed in the 1922 consolidation. Under the 1922 provisions, partnership income was only taxed at the partner level, sui juris beneficiaries were taxed on trust income to which they were presently entitled and the trustee was taxed on trust income to which no beneficiary was presently entitled or where a beneficiary was under a legal disability.[28] Until 1923 returns on debt interests in companies were given substantially equivalent treatment to that given to dividends.[29]
A company was required to pay income tax on distributions to nonresidents or, at its discretion, withhold tax from them. A shareholder included the dividend in assessable income but, in determining tax payable, deducted the tax paid by the company.[30] The end result of the system was that equivalent treatment was given to resident and nonresident shareholders in respect of distributions of both current year income and taxed income of prior years. Although the treatment of nonresident shareholders was completely non-discriminatory the actual collection of tax from nonresident shareholders was clearly problematic.[31]
The treatment of nonresident shareholders was unaffected by DTA obligations for the simple reason that Australia had not entered into any DTAs by 1923. From 1921 until 1947, under the Dominion Income Tax Relief scheme, Australia provided double tax relief to United Kingdom (‘UK’) shareholders receiving dividends from Australian companies. Relief was granted by Australia where the tax payable by the shareholder, after relief had been given by the UK government, was greater than the highest rate of Australian and UK tax applicable to the shareholder.[32] Because this system was not based on classifying income into different categories in limiting the source country’s right to tax, Australia could attempt to tax nonresident shareholders on a net basis at normal statutory rates. Indeed, an impressive feature of the UK’s Dominion Income Tax Relief system was the degree of diversity in systems of corporate-shareholder taxation that it permitted throughout the British Empire.
The majority report of the 1921-22 Warren Kerr Commission recommended the retention of the dividend deduction system at the Commonwealth level. After noting the inequities in the treatment of distributions of taxed income, the majority recommended that either the tax previously paid be refunded to the company or that a gross up and fully refundable credit mechanism operate at the shareholder level.[33] Refunding tax paid by the company would have excised the dividend imputation elements from the system and would have represented the treatment of distributions of taxed income which, as argued below, is to be preferred in dividend deduction systems.
In 1923, notwithstanding the recommendations of the Warren Kerr Commission, the Commonwealth, by repealing the dividend deduction provisions, converted its system to a dividend imputation system. The demise of the Commonwealth’s dividend deduction system was not due to perceived technical problems in its operation but rather was a consequence of a 1923 Commonwealth–State agreement under which the Commonwealth took responsibility for the collection of state taxes. The Australian States at the time operated either classical or dividend exemption systems. The conversion of the Commonwealth system to a dividend imputation system brought it closer to the States’ systems in that it now taxed both the distributed and undistributed profits of companies and hence facilitated the common collection of Commonwealth and State corporate taxes.
In summary, the system of corporate-shareholder taxation adopted in the ITAA15 was both elegant and simple. Dividends paid out of current year income were deductible to the company and fully assessable to shareholders. Where dividends were paid out of taxed retained income shareholders received a credit (but not a gross up) for corporate tax paid on the distributed income. The system, was generally consistent with vertical equity considerations; provided equivalent treatment for debt and equity and for different forms of business organisation; did not discriminate against nonresident investors; and (due to the territorial basis of the Australian system and specific pass through provisions) produced capital import neutrality at both the company and shareholder level. Moreover the costs of the complexity in the system were largely borne as administrative costs of government. The most complex features of the system concerned the treatment of distributions of income that had been subject to Australian corporate tax. If the recommendation of the Warren Kerr Commission, that corporate tax be refunded to companies making such distributions, had been enacted, the potentially complex dividend imputation features of the system could have been excised and greater vertical equity would have been achieved. The absence of an effective withholding tax system meant that, although the treatment of nonresident shareholders was completely non-discriminatory and completely unaffected by DTA obligations, the actual collection of appropriate levels of Australian tax on dividends paid to nonresidents proved to be problematic.
IV THE HARROWING COMPLEXITY OF THE PRESENT: AUSTRALIAN CORPORATE-SHAREHOLDER TAXATION IN 2005-06
If we now jump forward over 80 years to the present we find Australian corporate-shareholder system which is highly complex by any measure. Since 1987 Australia has had a shareholder credit or variable credit type of dividend imputation system. Since the so-called Simplified Imputation System was introduced in 2002, it has been largely modelled on the New Zealand system and, arguably, has in fact become more complex.
The main features of the Australian system of corporate-shareholder taxation are well known and are set out only briefly here. A resident company and certain New Zealand (‘NZ’) companies are obliged (or, in the case of NZ companies, permitted) to maintain franking accounts which track Australian corporate tax paid or payable (in the form of franking deficit tax) by the company or the franked amount of dividends received from another resident company. Payments of foreign corporate tax do not generate franking credits. A company may attach franking credits to dividends that it pays. The maximum franking credit that a company can attach to a dividend equals the amount of the dividend multiplied by c / 1 - c where c is the currently applicable corporate tax rate. The first dividend paid by a company in a franking period (one year for private companies and six months for public companies) establishes the company’s benchmark franking per centage for that period. In an effort to prevent dividend streaming within a franking period the system, subject to some exceptions for public companies, imposes a penalty tax (overfranking tax) if a company attaches more franking credits to a dividend than is consistent with its benchmark per centage. Where a company attaches more franking credits to its dividends than it has or generates in its franking account before the end of franking year, it may produce a franking deficit at the latter time. Franking deficit tax is offset against the company’s future mainstream company tax instalments and can be classified as a form of equalisation or compensatory tax. Where the franking deficit exceeds 10 per cent of the company’s total franking credits for the year the offset against mainstream company tax is reduced by 30 per cent. Attaching fewer franking credits to a dividend than is consistent with the company’s benchmark per centage produces a debit in the company’s franking account but does not pass a corresponding franking credit on to shareholders. Hence the effect is simply that underfranking produces a loss of franking credits for the company. Specific anti-streaming rules and a requirement that companies report excessive variations in their benchmark per centage are aimed at preventing dividend streaming between periods. The system also contains a general anti-avoidance rule directed at dividend streaming and rules aimed at counteracting trading in franking credits.[34]
Receipt of a franked dividend produces a credit in the recipient company’s franking account equal to the franking credits attached to the dividend. The recipient company also is allowed a gross up and credit for any franking credit attached to the dividend. While this means that a dividend that is franked to 100 per cent will be tax free to the recipient company it also means that a recipient company will have a tax liability on any dividend received that is franked to less than 100 per cent. Any tax paid by the recipient company on dividends franked to less than 100 per cent will in turn generate credits in the recipient company’s franking account. This is in contrast to the treatment of inter-corporate dividends prior to 1999 where the rebate under s 46 of the ITAA36 meant that dividends paid by a resident listed company were effectively tax free to a resident recipient company whether or not they were sourced in profits that had borne Australian corporate tax.[35] Although excess franking credits are generally refundable under the Simplified Imputation System this is not the case where the recipient of the dividend is a company. Excess credits are divided by the corporate rate and are carried forward by the company as tax losses.
Franking credits attached to dividends paid to resident shareholders are included in the assessable income of the shareholder. Tax is assessed on the shareholder’s grossed up income and the shareholder is allowed a tax offset equal to the amount of the attached franking credit. Excess credits are refundable to resident shareholders other than companies and some tax exempts. Excess credits on dividends received by resident companies are converted into tax losses in the manner described in the previous paragraph. The unfranked part of a dividend paid to a nonresident is subject to withholding tax. Nonresidents are not entitled to a franking gross up and credit but the franked part of a dividend paid by a resident company (and by certain NZ companies) to a nonresident is not subject to withholding tax. Certain redistributions of foreign source income to nonresident shareholders are also not subject to withholding tax. Dividends paid to nonresidents that are subject to withholding tax or would be but for certain specified exemptions (including the two previously mentioned) are not subject to tax on an assessment basis.
For resident shareholders the effect of the system is that the overall level of Australian tax on distributions of income that has borne Australian corporate tax represents taxation at the shareholder’s marginal rate. The treatment of unfranked dividends and the operation of franking deficit tax mean that corporate tax preferences are washed out on distribution. As taxed foreign source income will either be within the ‘exemptions’ under ss 23AH or 23AJ of the ITAA36 or subject to a foreign tax credit it will not, at least to the extent that it has borne foreign tax, be subject to Australian corporate tax. Hence, to that extent, it will not generate Australian franking credits and the exemption or credit treatment given to foreign source income at the Australian corporate level will be washed out on distribution. The payment of foreign tax will, in effect, be treated as a pre-tax expense.
The complexities of the dividend imputation system are exacerbated by the fact that it intertwines with a whole host of other complex rules that affect corporate-shareholder taxation. Those rules include: the debt and equity rules; the rules relating to tainting and untainting the share capital account; the thin capitalisation rules; the CGT system; the consolidation rules; the dividend stripping rules; the rules governing the cost of bonus issues; the buy back rules; various deemed dividend rules; the anti capital benefit streaming rules; the direct and indirect value shifting rules; the CFC rules; the FIF rules; the trans-Tasman triangular taxation rules; the exempting company rules; the ‘exemptions’ for foreign branch profits and non portfolio dividends; the foreign tax credit rules; and the personal services income alienation rules. Merely listing potentially relevant rules is itself a complex operation. Explaining their interactions and getting a real sense of the overall operation of the corporate-shareholder taxation system can be truly mind-boggling.
Survey evidence supports the assertion that Australian corporate taxation is complex. Company income tax has been found to have considerably higher compliance costs than personal income tax both in absolute terms and as a per centage of revenue collected.[36] For companies, internal costs represented 48 per cent of compliance costs with external costs for professional fees representing 52 per cent of compliance costs.[37] Computational costs represented 76.2 per cent of all corporate compliance costs with planning costs representing 23.8 per cent of corporate compliance costs. Planning costs represented a higher per centage of the total compliance costs for smaller companies while computational costs represented a higher per centage of the total compliance costs of the largest companies.[38] Corporate compliance costs were found to be heavily regressive.[39]
In summary, Australia currently uses a dividend imputation credit type system which since 2002 has been modelled on the NZ system. While the current system produces vertical equity, by permitting either full pass through or deferred recognition of tax preferences it favours partnerships and trusts over the corporate form. It also discriminates against nonresident investors by not fully extending imputation benefits to them. Moreover, while producing capital import neutrality at the corporate level (where either s 23AH or s 23AJ of the ITAA36 is applicable), it only produces national neutrality at the underlying resident shareholder level as payments of foreign tax are merely treated as pre-tax expenses for purposes of the dividend imputation system. The system is also much more complex than its 1915-23 progenitor and most of the cost of this complexity is now borne at the corporate level as compliance costs.
V THE JOURNEY FROM SIMPLICITY TO COMPLEXITY: THE EFFECTS OF INTERNATIONAL AND REVENUE CONSIDERATIONS ON THE DESIGN OF THE AUSTRALIAN DIVIDEND IMPUTATION SYSTEM
Many of the most complex features of the current Australian dividend imputation system can be regarded as being products of the national and international tax environment in which Australia reintroduced a dividend imputation system in 1987.
The dividend imputation system introduced in 1923 lasted until 1940 when it was converted to a classical system to assist in generating greater revenue to fund Australia’s participation in World War II.
By the time Australia re-introduced a dividend imputation system in 1987, the national and international tax world had radically changed. The 1987 dividend imputation system was replacing a classical system that had been in place for over forty years at a time when fiscal responsibility, revenue neutrality and efficiency considerations were paramount. Governments of countries recovering from the recessions of the late 1980s and early 1990s were preoccupied with deficit reduction. By this time the practice of existing imputation countries was to not unilaterally extend imputation credits to nonresidents; to not grant credit for payments of foreign tax; and to wash out corporate tax preferences (including tax preferred treatment of foreign source corporate income) on distribution to resident shareholders.[40]
Four key international developments in the intervening period also arguably had a significant effect on Australia’s choice of a dividend imputation system. These were: (a) the gradual development between the 1920s and the 1960s of an international framework for the interaction of tax rules between developed countries based on a network of bi-lateral DTAs; (b) the development of Australia’s DTA network; (c) the rejection of effective reciprocity of withholding taxes in the Commentary to the OECD Model Double Taxation Convention; and (d) the inclusion of the non-discrimination article in the OECD Model Double Taxation Convention.
A The Development of an International Framework for the Interaction of Tax Rules
The current international framework for the interaction of tax rules is a network of bilateral treaties which are either based on or heavily influenced by the OECD Model Double Taxation Convention. The OECD Model Double Taxation Convention as we now know it had its origins in work by committees of the International Chamber of Commerce and of the League of Nations in the 1920s and in the treaty practices of member states of the League of Nations. The process of development was lengthy, interrupted as it was by the involvement of the major capital exporting countries in World War II. From the beginning, the process reflected a tension between the interests of capital exporting and capital importing countries. By the mid 1940s a compromise was reached between developed countries. Source countries restricted their rights to tax investment income, by either not taxing it all or by imposing limitations on withholding taxes, thus assigning the major right to tax investment income to residence countries which were obliged to relieve international juridical double taxation. Conversely, source countries were given the right to tax business income attributable to permanent establishments at full statutory rates with the effect that the residence country only had a right to tax business income to the extent that its rate exceeded the source country rate. Fundamental to this consensus was a classification of taxes into origin (or source) taxes and residence taxes first developed in draft I-b of the 1928 draft of the Committee of Technical Experts of the Fiscal Committee of the League of Nations. Origin taxes were those imposed on non- residents on income from sources within the country of origin. Residence taxes were those imposed on residents on their worldwide income. Previous and other contemporary drafts by the Committee of Technical Experts had distinguished between personal and impersonal taxes and had consigned the right to levy personal taxes to the country of residence and the right to levy impersonal taxes to the country of origin.
In the context of this paper the significant point to stress is the influence of countries that used classical tax systems in the development of the eventual compromise. Germany switched to a classical system of corporate-shareholder taxation in 1925. German tax treaty practice, reflecting its classical and earlier schedular system was extremely influential in continental Europe in the 1920s and was reflected in draft 1c of the League of Nations 1928 Draft Model Taxation Convention.[41] The involvement of another classical corporate tax system country, the United States (‘US’), from 1927 onwards, was also extremely significant. A distinction between personal and impersonal taxes was problematic for countries like the UK and the UK (and for that matter Australia) which had global rather than schedular systems (the UK system not being a true schedular system in the sense that many continental European systems were) and hence arguably had no impersonal taxes. The development of the origin taxes – residence taxes distinction allowed countries using global systems to tax nonresidents.[42] The distinction, and the assignment of investment income to residence countries and business income to source countries, also dovetailed neatly with existing domestic rules in the US particularly with the foreign tax credit[43] and with the use of a classical system of corporate-shareholder taxation. The distinction also proved to be compatible with European schedular systems.
Taxation of foreign owned subsidiaries as residents, ceding the primary right to tax business profits to the source country where a permanent establishment is present, taxation of dividend income at source, reciprocal reductions in withholding taxes and no or limited relief to portfolio shareholders from economic double taxation all make sense in the context of schedular systems and classical corporate tax systems. Common to both systems is a view of the corporate income tax as a tax on a legal person and of the tax on dividends as a tax on distinct legal persons. In schedular systems the conceptual basis for having distinct and separate company and shareholder taxation is reinforced by the fact that one is seen as a personal tax while the other is seen as an impersonal tax. In these systems differential corporate and shareholder taxation is justified as the income subject to tax at each level is seen as arising from different sources. The framework as developed was to prove to be less appropriate for integrated systems of corporate-shareholder taxation.
B Development of Australia’s Bilateral Double Taxation Convention Network
Australia’s first bilateral DTA was concluded with the UK in 1946 and entered into force in 1947.[44] It was followed by DTAs with the US in 1953,[45] Canada in 1957 (entered into force 1958)[46] and New Zealand in 1960.[47] All of these conventions reflected the international consensus that had developed by the mid to late 1940s. By 1987 Australia had a well developed network of DTAs (excluding Airline Profits Agreements) with 23 countries.[48] In general these DTAs followed the OECD Model subject to Australia’s reservations in relation to some articles. In some instances, in treaties with then less developed countries, Australia’s DTAs were influenced by the UN Model (for example in relation to the definition of ‘permanent establishment’). By 1987, Australia’s DTAs almost universally imposed a reciprocal limitation of 15 per cent on the dividend withholding tax that the contracting states could levy. It is important to note that, as will be discussed in more detail below, by 1987 only Australia’s DTA with the US contained a non-discrimination article.[49] At this time Australia’s long-standing position had been not to agree to the non-discrimination article.[50] Nonetheless, subsequent protocols to several of Australia’s other DTAs obliged it to agree to a non-discrimination article with the state in question if Australia subsequently entered into a convention with a third state which contained a non-discrimination article.
C The Rejection of ‘Effective Reciprocity’
From the mid 1960s onwards a trend towards integrated systems of corporate-shareholder taxation developed within the OECD. Classical systems had become popular within the OECD during the Great Depression and World War II as part of an effort to increase revenues. For many countries the adoption of integrated systems represented a return to systems akin to those they had used earlier in their tax history. Nonetheless, the international framework for the interaction of tax rules eventually proved to be unwilling to adapt to the shift to integrated systems.
Some countries, most notably Germany but also others such as Norway, introduced split rate or dividend deduction systems that provided relief at the corporate level. Internationally the use of these systems was problematic due to the reciprocal reduction in withholding taxes which had become standard in double taxation conventions based on the OECD Model. Some earlier post World War II double taxation conventions had tried to take the overall level of corporate-shareholder taxation in the respective contracting states into account in determining the extent and nature of the treaty reductions in corporate-shareholder taxation. For example, under the 1945 United Kingdom – United States Double Taxation Convention,[51] the US reduced its withholding tax on dividends from 30 per cent to 15 per cent while the UK agreed not to levy surtax on dividends paid to nonresident shareholders.[52] As the standard tax rate payable by UK companies was 50 per cent plus a 5 per cent national defence contribution, while the US corporate rate was 40 per cent the treaty reductions meant that the total corporate and shareholder source country tax on dividend income was 55 per cent.[53] Consistently with this approach the DTA provided for an indirect foreign tax credit irrespective of the level of ownership of the shareholder.[54] In the 1960s, Germany tried to persuade its major treaty partners, especially the US, to permit it to levy a non-reciprocal and higher withholding tax on dividends. The German argument was that such treatment would represent ‘effective reciprocity’ as the split rate system meant that Germany was levying a lesser tax on distributed corporate profits. Germany’s lack of success in its negotiations with the US would not have provided Australia with any encouragement at all if it had been disposed to pursue either a split rate or a dividend deduction system.[55]
The attitude of the Commentary on the OECD Model to ‘effective reciprocity’ of withholding taxes reversed itself during this period. Under effective reciprocity a comparison is made, not just between the withholding rates of the parties to a bi-lateral convention, but between the combined levels of corporate and shareholder taxes on corporate distributions in each jurisdiction. The Commentary on the 1963 OECD Draft Convention appeared to have accepted effective reciprocity. The statements supporting effective reciprocity were removed from the 1977 edition of the Commentary. The attitude to effective reciprocity hardened in the 1992 Commentary which regarded ‘balancing’ withholding taxes that were not creditable for foreign tax credit purposes as discrimination by the source country against foreign investors and as discriminating against foreign investment as compared to domestic investment from the residence country’s viewpoint.[56]
D The Development of the ‘Non-Discrimination’ Article
Vogel notes that non-discrimination rules specifically in regard to tax law had been introduced as early as 1892 in bilateral treaties of friendship, commerce and navigation.[57] The 1921 and 1923 resolutions of the International Chamber of Commerce had affirmed non-discrimination between foreigners and residents.[58] Nonetheless non-discrimination articles only gradually found their way into model DTAs. A non-discrimination article was not included in the 1927 Draft Bilateral Convention for the Prevention of Double Taxation prepared by the League of Nations Committee of Technical Experts on Double Taxation and Tax Evasion.[59] The Mexico Draft Double Taxation Convention produced by the League of Nations Committee on Fiscal Affairs did contain a non-discrimination article requiring that taxpayers with fiscal domicile in one contracting state should not be subject in the other contracting state to higher or other taxes than a taxpayer having fiscal domicile or nationality in the other state.[60] Similarly, the early DTAs entered into by the US did not contain comprehensive non-discrimination articles. Neither the US – France Conventions of 1932[61] and 1939[62] nor the US – Canada Convention of 1942[63] contained non-discrimination articles. The US – Sweden Convention of 1939[64] did contain an article to the effect that citizens of each state residing within the other contracting state would not be subjected in the latter state to other or higher taxes than were imposed upon citizens of the latter state. The US – UK Convention of 1945[65] contained a similar article directed at discrimination against nationals. Contemporary commentators questioned whether corporations could be classified as nationals of either state under the article. No equivalent to the current art 24(5) of the OECD Model Double Taxation Convention was included in any of the early US Conventions.[66] None of Australia’s DTAs prior to the publication of the 1963 Draft OECD Model included a non-discrimination article. A non-discrimination article was included in the 1963 OECD Draft Model Double Taxation Convention. In the context of this paper the most important provision in the non-discrimination article was art 24(5) which prohibited other or more burdensome taxation of enterprises controlled by residents of the other contracting state as compared with similar enterprises of the first contracting state. Australia reserved its position on art 24. Nonetheless an overwhelming majority of OECD members did not reserve their position on the non-discrimination article and its inclusion in DTAs became the international practice.
E The Impact of these Developments on Australia’s Choice of an Imputation System
As Warren has noted, the 1920s division of the income tax base failed to evolve in response to corporate tax integration.[67] A relatively early consequence of that failure was the demise of systems of corporate-shareholder integration providing relief at the corporate level and the growing popularity by the mid 1980s of systems providing relief at the shareholder level. As will be discussed in more detail below, systems providing relief at the shareholder level were able to discriminate against nonresident shareholders and to wash out corporate tax preferences for foreign source income on distribution. These features meant that these types of integration systems operated like classical systems internationally.
In combination the non-discrimination article and the rejection of effective reciprocity meant that the use of a split rate or dividend deduction system came to be seen internationally as generating insufficient revenue from companies controlled by nonresidents when compared with an imputation system where unilateral credit was not required to be given to nonresidents. The rejection of effective reciprocity meant that a country using either a split rate or a dividend deduction system could not use higher withholding taxes on dividends to compensate for corporate tax revenue lost[68] when deductions were obtained for dividends distributed to nonresidents.
If a country using either a split rate or a dividend deduction system tried to protect its corporate tax base by denying a company a deduction for dividends paid to nonresidents then technically it was in breach of the non-discrimination article in the OECD Model. By contrast, the better view was that denial of an imputation gross up and credit to a nonresident shareholder technically did not breach the non-discrimination article.[69]
The influence of these international developments was evident in the proposal in the Australian Treasury Draft White Paper of 1985, Reform of the Australian Tax System, that a full or partial dividend imputation system for resident individual shareholders be introduced.[70] The ability to deny imputation credits to nonresident shareholders was explicitly stated to be an advantage of an imputation system[71] while difficulties associated with taxing profits of local subsidiaries of foreign parents was regarded as the decisive consideration against a split rate system.[72] Both considerations had previously been noted by the Asprey Committee in 1975.[73] Reflecting the mood of fiscal responsibility of the period these reforms were all required to be revenue neutral.[74]
F The Impact of Revenue Considerations on Australia’s Choice of an Imputation System
The imputation system eventually introduced in Australia in 1987 did depart in some respects from the then dominant model of imputation in the OECD. Initially Australia had intended to adopt an imputation system with a compensatory or equalisation tax modelled on the then current UK Advance Corporation Tax system. The variable credit type of imputation system ultimately introduced by Australia appeared to be heavily influenced by a model developed in the interim by Benge and Robinson.[75] Under the system as initially implemented a company, within certain limits (such as the rule requiring all dividends paid on shares in the same class to be franked to the same amount) was able to choose the extent to which it franked a dividend. Adverse consequences could follow, however, for a company that did not attach franking credits to a dividend to the extent that, having regard to matters such as committed future dividends, credits were available in its franking account. In effect, the required franking amount rules operated as ordering rules which regarded taxed profits as being distributed first. Under the rules it was possible, and even usual, for companies to distribute their taxed profits as fully franked dividends and to retain their untaxed profits. Where a company chose to retain its untaxed profits the system enabled credit at the shareholder level to be restricted to distributions of profits that had borne corporate tax thus avoiding the problem of super-integration without imposing a compensatory tax. When a company distributed untaxed profits without attaching franking credits to the distribution, no gross up and credit was received by the shareholder. Where a company chose to attach franking credits to distributions of untaxed profits it could render itself liable for a compensatory or equalisation tax in the form of franking deficit tax. The principal reason for the choice of this form of dividend imputation system appears to have been unease in corporate Australia at the prospect of payments of Advance Corporation Tax adversely affecting reported after tax profits of listed companies.[76] Similar concerns continued to be raised by corporate Australia when the Review of Business Taxation (‘RBT’) canvassed the possibility of using a deferred company tax as a solution to the problem of dividend streaming.[77]
At the same time the complex franking account mechanism introduced in 1987 can be seen as being driven by revenue considerations aimed as it was at preventing the granting of imputation credits on distributions of untaxed income. The wash out of corporate tax preferences on distribution was not an inevitable consequence of that system but was a feature which carried over from the classical system, was consistent with the practice of other imputation countries, and assisted the achievement of the goal of overall revenue neutrality in reform package. A concern to minimise revenue costs was also evident in the non-refundability of imputation credits in the 1987-1999 system.
G The Legacy of the Choices Made by Australia in Designing its Imputation System
In combination these features produced biases in the system which led to the development of a complex set of rules directed against dividend streaming, franking credit trading and capital benefit streaming. The problem of dividend streaming was the price that Australia paid for adopting a variable credit type of imputation system and for deciding not to extend franking credits to nonresident shareholders. The so called Simplified Imputation System introduced in 2002 following recommendations of the RBT changed the whole basis of franking under the system to benchmark franking in an effort to curtail dividend streaming.
Arguably the recommendations of the RBT did little to remove biases in the system or to simplify it. The RBT rejected a general pass through of corporate tax preferences in companies because of revenue considerations.[78] This coupled with the decision to continue to not grant franking credits to nonresidents meant that the changes implemented in response to RBT recommendations, with the exception of the introduction of refundable franking credits, while simplifying the operation of the system in some respects and while making imputation planning more difficult, did not really address the underlying biases that caused complexity in the system. In fact the changes to the inter-corporate dividend rebate exacerbated the biases against investment by residents in resident companies with taxed foreign income.
VI CURRENT KEY DESIGN FEATURES IN AUSTRALIA’S DIVIDEND IMPUTATION SYSTEM AND THEIR POLICY UNDERPINNINGS
The Australian dividend imputation system is one of several alternative approaches to integrating a country’s corporate and shareholder tax systems. The following may be regarded as the key design features of Australia’s current dividend imputation system:
1 The overall rate of tax on taxed income passing through a resident company to a resident individual shareholder represents taxation of that income at the shareholder’s marginal rate; 2 Where a distribution produces an excess credit for a resident shareholder the excess credit is fully refundable to shareholders other than corporates and some tax exempts; 3 Where a distribution produces an excess credit for a resident corporate shareholder the excess credit is converted into a tax loss for the shareholder; 4 Except in the case of dividends paid within a consolidated group (which are ignored for tax purposes) the same gross up and credit mechanism operates for dividends received by Australian resident companies as operates for dividend received by Australian resident individuals; 5 Tax preferred income (including tax preferred foreign source income) distributed by a resident company to a resident shareholder (individual or corporate) is taxed at the shareholder’s marginal rate; 6 Tax credits given to resident shareholders are limited to the franked portion of the distribution which represents tax paid or payable by the resident company; 7 Except where a resident company attaches franking credits to a distribution in circumstances that ultimately trigger a liability for franking deficit tax, a compensatory tax is not levied on distributions of income that has not borne Australian corporate tax; 8 Resident companies (because of the benchmark franking rule) can retain untaxed income and distribute taxed income with franking credits attached; 9 Taxed income retained by resident companies is only taxed at the corporate rate until it is distributed but personal services income alienation provisions may mean that the income is taxed at the marginal rate of the person providing the personal services; 10 The streaming of franking credits away from shareholders who have little or no use for them is circumvented by the benchmark franking rule and by several specific rules proscribing dividend streaming; 11 Trading in franking credits by shareholders is circumvented by specific measures against franking credit trading; and 12 To the extent that a distribution by a resident company to a nonresident shareholder (individual or corporate) represents taxed income it is exempt from Australian withholding tax but no gross up or tax offset is given to the nonresident shareholders.
These key design features may be regarded as being consequences of a more limited number of Australian domestic and international tax policies relevant to corporate-shareholder taxation. Domestically, the broad policy objectives of the dividend imputation system can be stated as: (a) that the overall level of taxation of profits that pass through a resident company to a resident shareholder should represent taxation at the relevant shareholder’s marginal rate; (b) as a consequence corporate tax preferences are washed out on distribution; and (c) that companies can attach credits to dividends in a manner that achieves these results without paying a compensatory tax.
Internationally the broad policy objectives of the imputation system can be stated as: (a) preventing credit from being given for payments of foreign tax; and (b) discriminatory treatment of nonresident shareholders in situations where reduced rates of withholding tax apply under a DTA. The former policy arguably produces a bias at the resident shareholder level against investment in resident companies with foreign source income that has been subject to foreign tax. This bias is mitigated by the ability of resident companies to fund franked dividends from taxed profits and to retain untaxed profits. In respect of the latter policy, as noted previously, the better view is that technically the discrimination does not amount to a breach of the non-discrimination article in Australia’s DTAs.
The argument of this paper is that it is the combination of the domestic and international policy objectives, and particularly the objective of discriminating against nonresidents, that produces the key design features and complexities of Australia’s dividend imputation system noted above. The consequence of this analysis that Australian companies which do not have nonresident shareholders are obliged to comply with requirements of the Australian dividend imputation system that only exist because Australia wishes to discriminate against nonresident shareholders. Assuming that Australia wishes to continue this discrimination, this observation begs the question of whether it is possible to design a simpler form of corporate-shareholder taxation that produces equivalent results for resident shareholders to the dividend imputation system for those Australian companies that do not have nonresident shareholders. This in turn begs the further question of whether such a simpler system can be designed in a way that does not infringe the non-discrimination article in Australia’s DTAs.
VII SOME CHARACTERISTICS OF PRIVATE COMPANIES IN AUSTRALIA
Before examining the design features of a simpler corporate-shareholder tax system for companies with no nonresident shareholders, it will be useful to obtain some estimate of the numbers of Australian companies in that category. Such an estimate should be of assistance in quantifying any reduction in the effective complexity of the tax system that might result from a simpler form of corporate-shareholder tax system. Public companies for tax purposes are not considered because of the existence of the the secondary market in shares in listed companies. The secondary market means that it would be impossible to enforce a rule restricting optional dividend deduction treatment to listed companies with no nonresident shareholders.
In the 2002-2003 income year, 248,880 or 89 per cent of tax paying companies were private companies. These companies paid 36.74 per cent of the net company tax paid in 2002-2003. There were also 349,733 private companies which were either loss making or which had zero taxable income. This represented 97 per cent of non tax paying companies. The total income of all private companies in 2002-2003 was $577,824,895,424 while the total taxable income of all private companies $47,417,578,189. Of all private companies in 2002-2003 only 1,192 or 0.2 per cent paid more than $1,000,000 in company tax. This represented 0.47 per cent of all tax paying private companies in 2002-2003.[79]
Clearly a significant proportion of private companies do not have nonresident shareholders. Published Taxation Statistics do not directly disclose the number of nonresident shareholders in Australian companies but published statistics are consistent with this conclusion. In the 1999-2000 income year, the last year in which statistics on dividend withholding tax payments were separately reported, only 406 companies (excluding nominee companies of securities dealers) paid a total of $60,849,000 in dividend withholding tax.[80] In 2002-2003 a total of 225 nontaxable and a total of 246 taxable private companies claimed a total of $19,064,582 in s 46FA deductions. A total of nine non-taxable and five taxable public companies claimed a total of $75,684,082 in s 46FA deductions in the same period.[81] Given that some companies may have paid franked dividends to nonresident shareholders, the total number of private companies with nonresident shareholders is likely to be higher than the number of private companies that paid unfranked dividends or claimed s 46FA deductions but is still likely to be a clear minority of private companies.
VIII A BACK TO THE FUTURE SOLUTION FOR CLOSELY HELD COMPANIES: OPTIONAL DIVIDEND DEDUCTIBILITY
Where a company has only resident individual shareholders substantially equivalent results to those produced under the Australian dividend imputation system can be achieved more simply by permitting the company to use a form of dividend deduction system.[82] To achieve taxation of distributed income (whether subject to corporate tax or not) at the resident individual shareholder’s marginal rate the company would need to be permitted to carry tax losses resulting from dividend deductions both forward and back. Companies would not be permitted to carry losses back earlier than the year when they joined the dividend deduction system. Adjustments would need to be made in the value of losses carried forward or back for changes in the corporate tax rate between periods and an ordering rule for the application of losses attributable to different periods would need to be developed.
Because of the existence of the secondary market for shares in listed companies, such a system would not be appropriate for listed companies, or subsidiaries of listed companies. To prevent dividends being diverted to nonresidents through the use of intermediate entities it would also be necessary to limit the entities who could own shares or other interests in an optional dividend deduction company to resident individuals or intermediate entities in which the only stakeholders were resident individuals who were not trustees.
A company electing for dividend deduction treatment would not need to maintain a franking account or apply the franking rules to distributions, and would not be subject to either the anti-dividend streaming or the anti-franking credit trading rules. Shareholders receiving dividends would simply be taxed on the amount of the dividend without the need to apply the gross up and credit mechanism.
Companies in the imputation system (and other interposed entities such as trusts and partnerships) receiving dividends from dividend deduction companies would simply treat the dividend as they would an unfranked dividend.
A rule would need to be developed for the situation where a dividend deduction company received a franked dividend. One possibility here would be to either require the company to track franking credits on dividends received and to attach them to dividends that it paid using an ordering rule like the benchmark franking rule. Such an approach would reintroduce much of the complexity that a dividend deduction option was aimed at avoiding. Another approach would be to treat the dividend deduction company as a conduit in much the same way as partnerships and trusts are treated in the dividend imputation system. A difficulty with this approach is that the current rules giving conduit treatment to partnerships and trusts for imputation purposes are premised on the allocation of net income or of partnership losses on basis of entitlement rather than actual distribution.[83] By contrast, under a dividend deduction system corporate income is only assessable to a shareholder on distribution. To make conduit treatment of franking credits received by a dividend deduction company dependant on actual distribution would again impose the complexities of tracking and franking requirements on dividend deduction companies.
A better alternative may be to pursue the trust analogy from a different perspective and to only allow a shareholder in a dividend deduction company a franking credit if and to the extent that the franked distribution received was redistributed as a dividend in the year of receipt. Such a rule might be thought to be appropriate as a concessional measure and would be likely to lead to immediate redistributions of franked dividends received. It would not be practical to implement a rule of this nature where a company had more than one class of shareholder. Hence it would appear to be necessary that a company electing to receive dividend deduction treatment only have one class of shareholder. If this limitation were in place an alternative approach would be to regard franked dividends received by a dividend company as being included in the assessable income of shareholders irrespective of their actual distribution. This approach would be unfair to shareholders who had no control over the distributions of the company although similar unfairness might be thought to be present in the existing rules governing the taxation of some trusts and partnerships. Unfairness in the application of this rule to companies opting for dividend deduction treatment could be mitigated by imposing limits on the number of shareholders that a company choosing the dividend deduction option was permitted to have.
A company under an optional dividend deduction system would be able to choose to retain its tax preferred income and to distribute its taxable income. To the extent that it did this it would achieve the same end effect as does a company under Australia’s dividend imputation system which chooses to retain tax preferred income and to distribute its taxed income as franked dividends. An optional dividend deduction system could produce a result that a company which distributed tax preferred income generated a larger loss carry forward which could then be offset against its future assessable income with the result that income of the later year could be retained tax free. As illustrated in Examples 1 and 2, the end result of such action would be equivalent to that produced in the Australian dividend imputation system if the company had only distributed its taxed income as franked dividends in both years. The combined corporate and shareholder tax paid under both systems would be identical if the companies adapted their dividend policies to the characteristics of the two corporate tax systems.
Example 1: Dividend Deduction System
Year One
Company
Profit $200
Income $100
Dividend $200
Deduction $200
Taxable Income $0
Loss carry fwd ($100)
Tax paid $0
Shareholder (48.5 per cent marginal rate plus Medicare levy assumed)
Dividend $200
Tax $ 97
After tax dividend $103
Total tax $ 97
Year Two
Company
Profit $100
Income $100
Loss carry fwd ($100)
Dividend $0
Taxable Income $0
Tax paid $0
Total tax $0
Total tax Year 1 and Year 2 = $97
Example 2: Variable Credit Imputation System
Assume the facts in Example 1 with the variation that an Australian/New Zealand style variable credit dividend imputation system is used.
Year One
Company
Profit $200
Income $100
Tax $ 30
Franking credits $30
Dividend $70
Franking credit attached $30
Shareholder
Dividend $70
Gross up for franking credit $30
Grossed up dividend $100
Gross tax $48.50
Tax offset $30
Net tax $18.50
After tax dividend $51.50
Total tax $48.50
Year Two
Company
Profit $100
Income $100
Tax $30
Franking credits $30
Dividend $70
Franking credits attached $30
Shareholder
Dividend $ 70
Gross up for franking credit $ 30
Grossed up dividend $100
Gross tax $48.50
Tax offse $30
Net tax $18.50
After tax dividend $51.50
Total tax $48.50
Total tax Year 1 and Year 2 = $97
Under an optional dividend deduction approach corporate income would only bear tax at the corporate rate so long as it remained in corporate solution. Hence introducing an optional dividend deduction approach would not of itself do anything to counteract any tax planning that diverted personal services income to companies.
Assuming that capital gains on shares would continue to be given preferential tax treatment, there would be a continuing need for functionally equivalent rules to those dealing with tainted share capital accounts and anti-capital benefit streaming. Obviously debits in franking accounts could not be used as a sanction in these situations. In fact the solutions to these problems are likely to be simpler in a dividend deduction system. For example, the problem of profits being capitalised and then distributed as returns of capital could be dealt with by defining the share capital account as excluding an account containing capitalised profits or polluted in other ways and by continuing the current rule that distributions from a polluted share capital account are regarded as dividends. Similarly, the rules concerning off-market buy backs would need to be adapted to the optional dividend deduction system but this would not give rise to any significant problems. Deemed dividend rules for loans and debt forgiveness would also need to be adapted but there would be no point in applying deemed dividend rules in relation to excessive salary payments.
The tax treatment of returns on domestic debt and equity interests would be virtually identical under an optional dividend deduction system. The only difference of substance between the two treatments would be that losses attributed to dividend deductions could be carried both backward and forward under a dividend deduction system. Hence if an optional dividend deduction system were to be introduced it may be desirable to permit at least all dividend deduction companies to carry all losses back to the year of entry into the dividend deduction system as well as forward. If this were done then, as returns on both debt and equity would be deductible to the paying company, there would be no need to classify securities as being either debt or equity for domestic tax purposes.
Furthermore, as no interests in an optional dividend deduction company were controlled, directly or indirectly, by nonresidents, there would be no reason for applying the inbound thin capitalisation rules to them. In the case of the outbound thin capitalisation rules the concern would be that the company would obtain inflated deductions in respect of non-assessable non-exempt foreign source income. Hence, optional dividend deduction companies which did not derive non-assessable non-exempt could be exempted from both the debt and equity rules and the outbound thin capitalisation rules on an annual basis. Arguably there would be little need to apply the outbound thin capitalisation rules to an optional dividend deduction company whose only foreign source income was from portfolio investments. In any event, given that returns on both debt and equity would be fully deductible to the company there would be little point in distinguishing between debt and equity in limiting deductions in situations where optional dividend deduction companies did have non-assessable non-exempt foreign source income. In these cases it may be better to have a simpler limit based on an arbitrary per centage of the non-assessable non-exempt foreign source income derived by the optional dividend deduction company.
Hence it is argued that the legal complexity and associated compliance costs of a dividend deduction system at both the company and the shareholder level would be likely to be substantially less than the legal complexity and associated compliance costs of the current Australian dividend imputation system. As fewer companies and shareholders would be subject to the greater legal complexity and higher compliance costs of the dividend imputation system the effective complexity of the income tax system would also be likely to be reduced. It is also likely that administrative costs associated with a dividend deduction system would be lower than those associated with the current Australian dividend imputation system (auditing of dividend deduction companies which did not maintain franking accounts nor frank dividends would appear to be simpler as would assessing tax returns by shareholders who generally would not be claiming a gross up and credit). This would also be likely to reduce the effective complexity of the income tax system.
For these benefits to arise the rules for electing for dividend deduction treatment would need to be simply and clearly set out. In order to be permitted to elect for dividend deduction treatment:
• The company would have to be a private company for tax purposes; • The company could have only one class of shareholders; • No shares or other interests in the company could be held by nonresidents; • No shares or other interests in the company could be held by tax exempt entities; • The only entities which would be permitted to own shares in the company would be resident individuals or other resident intermediate entities in which the only stakeholders were resident individuals who were not trustees; • The number of shareholders in the company could not exceed a specified number (eg the number equal to the number of partners possible in a general partnership).
The company would also be excluded from the operation of the debt and equity rules and the thin capitalisation rules where no interests in the company were held by nonresidents and where it did not have any foreign branch profits or any non portfolio foreign source dividend income.
IX DOES THE PROPOSAL INFRINGE THE NON-DISCRIMINATION ARTICLE IN AUSTRALIA’S DTAS?
Only two of Australia’s DTAs (those with the US and the UK) actually contain a non-discrimination article.[84] By a Protocol entered into in 1989[85] the following provision, art 27A, was inserted in the Australia – France DTA[86]:
If, in an agreement for the avoidance of double taxation that is made after 19 June 1989 between Australia and a third State, being a State that is a member of the Organisation for Economic Co-operation and Development,
(a) [Paragraph (a) has not been extracted in this paper]
(b) there is included a Non-discrimination Article,
the Government of Australia shall immediately inform the Government of the French Republic in writing through the diplomatic channel and shall enter into negotiations with the Government of the French Republic, in the case of paragraph (a), to review the provisions specified in that paragraph in order to provide the same treatment for France as that provided for the third State and, in the case of paragraph (b), in order to provide the same treatment for France as that provided for the third State.
Similar provisions are contained in protocols to Australia’s DTAs with Korea,[87] Finland,[88] Spain,[89] South Africa,[90] Romania,[91] and Mexico.[92] The DTAs with non-OECD countries merely make reference to subsequent treaties with a third country not to subsequent treaties with a third OECD country. All of the protocols containing provisions similar to article of the Australia – France DTA were entered into prior to Australia entering into its 2003 DTA with the UK.[93]
There are some differences between the non-discrimination articles in the Australia – US DTA and the Australia – UK DTA. The provisions, however, that are relevant to whether the proposal infringes the article are substantially the same. Paragraph 1(c) of art 23 of the Australia – US DTA reads as follows:
a corporation of a Contracting State, the capital of which is wholly or partly owned or controlled, directly or indirectly, by one or more residents of the other Contracting State, shall not be subjected in the first-mentioned State to any taxation or any requirement connected therewith which is more burdensome than the taxation or connected requirements to which other similar corporations of the first-mentioned State in the same circumstances are or may be subjected.
The corresponding provision, which more closely follows the OECD Model, paragraph 4 of art 25 of the Australia – UK DTA reads as follows:
Enterprises of a Contracting State, the capital of which is wholly or partly owned or controlled, directly or indirectly, by one or more residents of the other Contracting State, shall not be subjected in the first-mentioned State to any taxation or any requirement connected therewith which is other or more burdensome than the taxation and connected requirements to which other similar enterprises of the first-mentioned State in similar circumstances are or may be subjected.
For some years, the UK Inland Revenue’s view of non-discrimination articles was that they are limited to discrimination against enterprises controlled by residents of the treaty partner state as compared with enterprises controlled by residents of a third state.[94] The better view, however, is that in determining whether there is a breach of the non-discrimination article in the OECD Model Convention (art 24(5)) a comparison should be made with an enterprise the shareholders or partners of which are exclusively residents.[95] For example, Vogel notes that Swedish and Dutch courts have found that rules granting relief on corporate reorganisations which were limited to cases in which both participants were created under domestic law infringed the non-discrimination article.[96] Most recently in the UK Court of Appeal decision in NEC Semi Conductors Ltd v IRC [2006] STC 606, Lloyd LJ (with whom Mummery LJ and Sedley LJ agreed on this point) said of the comparison to be made under the non-discrimination article,
What are the ‘other similar enterprises’? It cannot mean other UK enterprises owned or controlled by residents of Japan; that would not prohibit any discrimination against Japanese owned UK subsidiaries generally. For the same reason it cannot mean UK companies owned or controlled by residents of countries other than the UK generally……..The comparison which is required by the non-discrimination article includes, among several possibilities, a comparison with a UK company wholly owned or controlled by a single shareholder. It is true that the non-discrimination article does not deal with whether the sole shareholder is an individual or a corporation, but it seems to me that it would be irrational to exclude that as a relevant element in the comparison, where the circumstances of the actual case are such that the foreign parent is a corporation. As Park J said ‘it is necessary to assume that, if the actual company is a subsidiary of another company, then so is the hypothetical company.’[97]
Hence under the better view, either denying deductions for dividends paid by optional dividend deduction companies to nonresidents or only permitting companies with no nonresident shareholders to become dividend deduction companies would infringe the non-discrimination article in both the US and UK DTAs.[98]
For many years the US has not permitted nonresidents to be shareholders in IRC Subchapter S corporations.[99] The Technical Explanation to the United States Model Double Taxation Convention[100] argues that this restriction does not infringe the non-discrimination article. The Technical Explanation argues that while resident shareholders in an S corporation are taxed by reference to the pro rata shares of the income, losses, deductions and credits of the corporation that are allocated to them, nonresident aliens are not taxed on a net basis and thus does not generally take into account corporate losses, deductions or credits. Hence the Technical Explanation characterises the exclusion of nonresident alien shareholders as arising not because the shareholders are foreign but because they are not taxed on a net basis.[101] The argument has not impressed the commentators who point out that the problem would be easily resolved by applying to S corporations the withholding tax rules that the US applies to partnerships with nonresident partners.[102]
Irrespective of the merits or otherwise of the arguments on this point in the US Technical Explanation, they would not be able to be used without modification to justify denial of dividend deduction company status to companies with nonresident shareholders. Analogous logic to that used in the Technical Explanation, however, may provide a justification for such treatment. This would be to ask why Australia would want to exclude nonresident shareholders from owning shares in optional dividend deduction companies. The answer is not so we can discriminate against companies with nonresident shareholders but rather a fear of erosion of our corporate shareholder tax base through a combination of excessive distributions and low treaty rates of withholding tax. Once the problem is characterised as one of preventing base erosion it is possible to identify other classes of taxpayer that Australia would want to exclude from being shareholders in optional dividend deduction companies. Base erosion would be a risk in all instances where tax exempt entities were shareholders in optional dividend deduction companies. Given that some resident tax exempt entities are currently entitled to fully refundable franking credits, to minimise the risk of infringing the non-discrimination article, it would be desirable to exclude those entities from restrictions on share ownership in optional dividend deduction companies. If both nonresident shareholders and certain resident tax exempt shareholders were excluded from owning shares in optional dividend deduction companies it would be arguable that the exclusion was not based on the residential status of the shareholder but on the risk of erosion of the corporate-shareholder tax base presented by ownership of shares in optional dividend deduction companies by tax exempt or low taxed shareholders. This approach would parallel that taken in the existing regime denying the ability to frank dividends to exempting entities.
Alternatively, one of the suggestions made by commentators on non-discrimination problems with the US Subchapter S corporation might be able to be adapted to resolve base erosion concerns if nonresidents were permitted to own shares in optional dividend deduction companies. This is the suggestion that nonresidents be permitted to hold shares in Subchapter S corporations if they elect to be taxed on a net basis.[103] Applying this suggestion in the context of an optional dividend deduction system, however, may be more complex as shareholders in a Subchapter S corporation are not taxed on dividends and hence the suggestion would not infringe limits on dividend withholding tax in US DTAs. In an optional dividend deduction system shareholders are taxed on dividends received and Australia’s DTAs impose limits on the tax the Australia can levy on dividends paid to nonresident shareholders. Allowing companies with nonresident shareholders to be optional dividend deduction companies if the nonresident shareholders elect to be taxed at standard nonresident rates on a net basis would not infringe the non-discrimination article but might infringe the dividend article in Australia’s DTAs.
Interestingly, there are Australian precedents for allowing a nonresident shareholder the option to be taxed on dividends on a net assessment basis rather than on a gross withholding basis. Australia first introduced a withholding tax on dividends in 1959. Although the withholding tax was at a flat rate and was levied on the gross dividend income, nonresidents were given the option under ITAA36 s 128D(2) to elect to be taxed on an assessment basis on the dividend income under ITAA36 s 44(1)(b). If the election was made, s 128D(3) required the nonresident to file a return disclosing the nonresident’s dividend income and the nonresident’s worldwide income. A refundable credit against the tax on an assessment basis for withholding tax paid on the dividend was given under ITAA36 s 128E(1)(a). Where tax on an assessment basis exceeded the withholding tax and the tax on an assessment basis that would have been payable if the ITAA36 s 128D(2) election had not been made, ITAA36 s 128E(1)(b) entitled the nonresident to a non refundable rebate of the excess.[104] The rebate under s 128E(1)(b) meant that it would have been difficult to argue that the option to be taxed on a net assessment basis infringed the limits on Australia’s taxation of dividends in its then current DTAs.
The purpose behind allowing nonresidents to be shareholders in optional dividend deduction companies if they elected to be taxed on a net assessment basis (with credits for any prior withholding tax) would be to prevent corporate-shareholder base erosion. The existence of a such a rebate like the former ITAA36 s 128E(1)(b) rebate would defeat that purpose. Hence, despite Australia’s previous experience with optional net basis taxation of dividends for nonresidents the question remains whether such treatment without a refundable rebate would infringe the dividend articles in Australia’s DTAs.
Consideration of that question involves the issue of whether a taxpayer who has elected to be taxed at higher rates on a net basis on dividends can thereafter claim that the lower treaty rates of tax on a gross basis should have applied. The Technical Explanation to the United States Model Income Taxation Convention may provide some basis for an argument that there is no infringement of dividend article in these circumstances. The Technical Explanation expresses the view that the dividend article is not infringed if tax is withheld at the time of payment at full statutory rates and the treaty benefit provided by means of a subsequent refund.[105] While such procedure may be justified as a means of verifying the nonresident status of the shareholder, when it depends on the refund process being initiated by the shareholder it might not be thought to be too far removed from a process in which the shareholder elects to be taxed at full statutory rates.
On balance the alternative of restricting shareholding in optional dividend deduction companies to nonresidents and those tax exempts which are not currently entitled to franking credit refunds would appear to be the option least likely to infringe the non-discrimination article in Australia’s DTAs and the one most closely analogous to other current Australian practices.
X INTERNATIONAL OBLIGATIONS AND THE COMPLEXITY OF DOMESTIC TAX DESIGN
I began this paper by recalling an incident reported in Kevin Brownlow’s The Parade’s Gone By. Symmetry requires therefore that I should finish it by recalling a quote reported in the same book. Brownlow chose to end his book with a quote from America’s sweetheart, the Canadian, Mary Pickford. The quote was ‘It would have been more logical if silent pictures had grown out of the talkie instead of the other way around.’ Film critic Walter Kerr in his The Silent Clowns after wondering what the woman could have meant by this quote, why is it logical to wind up with a black and white dance of silent ghosts, concludes that it is artistically apt. Kerr’s conclusion is that ‘all art begins with a taking away’; that audiences responses to less complete canvasses are more engaged, responsive and intelligent than they are to those that are over cluttered with detail.
An analogy can be drawn with the development of Australian corporate-shareholder taxation since 1915. We have moved from an elegant and simple system to a complex and cluttered one which arguably distorts rather than promotes constructive business activity in response to it. No doubt, some of our technical mechanisms (such as our withholding tax regimes, arbitration procedures and exchange of information arrangements) represent significant improvements on their 1915 equivalents. Paradoxically many of the complexities in our system are consequences of our bilateral DTA network locking us into the international taxation consensus and compromises that emerged following World War I. Perhaps the focus on structural issues in the current tax reform debate in Australia represents a belated recognition of the art of taking away, of stripping down the system to its essentials. Certainly the argument of this paper is that the corporate-shareholder tax system is more complex than it needs to be for the majority of companies and, but for international considerations, much of that complexity would never have existed. The submission of this paper is that for these companies that complexity can be stripped away without a significant risk of infringing Australia’s DTAs obligations.
It has to be acknowledged, however, that international treaty obligations may limit Australia’s freedom of action in simplifying its corporate-shareholder taxation regime. Clearly one of the consequences of globalisation increasingly is the limitations that it, and international agreements entered into as part of and in response to it, places on national sovereignty. The talkie to silent analogy when viewed from a different perspective can also be helpful in developing a response to this concern. Historically income tax systems developed first as domestic systems and co-ordinated rules for their international interaction only emerged later. In broad terms the bi-lateral treaty networks that have developed since World War II have been largely concerned with limiting the taxing rights that contracting states would otherwise enjoy under their domestic law. As Vann has argued the OECD Model based as it is on a schedular approach and on compromises between residence and source taxation represents the post World War I consensus not the post World War II consensus.[106] Over time the result was that this superimposition of international rules based on a schedular approach as a gloss on disparate domestic systems increased the complexity of the domestic systems. It may have been more logical to have developed rules for the international interaction of systems first and to then design the domestic systems with the international rules in mind. The mobility of capital that is the defining characteristic of globalisation and the financial engineering that has developed symbiotically with it are increasingly highlighting the shortcomings of the schedular and classical approach of the post World War I consensus as embodied in the OECD Model.
Perhaps the time has come to truly reconsider international rules for the interaction of tax systems to allow for the flexible development of domestic rules built on them. One back to the future suggestion is Harris’ suggestion of developing a composite income tax system based on a refined version of the UK system of Dominion Income Tax Relief which operated from the 1920s to the mid to late 1940s.[107] The fundamental feature of this approach is that source countries would tax nonresidents at their source rate reflecting the services provided to them while residence countries would give residents foreign tax credits based on the residence country’s source rate. As was the case with the UK system of Dominion Income Tax Relief, a composite tax system would be compatible with the international operation of a wide variety of corporate tax systems including an optional dividend deduction system.[108]
[*] BA, LLB, LLM (Hons) (Sydney) Cer H Ed (UNSW); Associate Professor, School of Business Law and Taxation, University of New South Wales.
[1] Board of Taxation, Commonwealth, Identification and Possible Repeal of the Inoperative Provisions of the 1936 and 1997 Income Tax Assessment Acts — A Report to the Treasurer (2005); Peter Costello, ‘Income Tax Act Reduced by around 30 per cent’ (Treasurer’s Press Release No 102, 24 November 2005).
[2] Commonwealth Treasury, ‘Taskforce on Reducing the Regulatory Burden on Business’ (Treasurer’s Press Release No 87, 12 October 2005).
[3] Taskforce on Reducing Regulatory Burdens on Business, Commonwealth, Rethinking Regulation (2006).
[4] Email from Graeme Cooper, Conference Organiser, to C John Taylor, 21 November 2005.
[5] Letter from Noel Rowland, Chief Executive Officer, TIA to C John Taylor, 18 August 2005.
[6] C John Taylor, Beyond 4100: A Report on Measures to Combat Rising Compliance Costs through Reducing Tax Law Complexity (2006).
[7] Bihn Tran-Nam, ‘Tax Reform and Tax Simplification: Some Conceptual Issues and a Preliminary Assessment’ [1999] SydLawRw 20; (1999) 21 Sydney Law Review 500, 506.
[8] Ibid 507.
[9] Ibid 507.
[10] Ibid 508.
[11] Income Tax Assessment Act 1915 (Cth) (‘ITAA15’) s 16(1). Income Tax Assessment Act No 2 1916 (Cth) substituted a new s 16(1) which limited the deduction to distributions of ‘assessable income’ as opposed to ‘income’ and inserted a definition of ‘assessable income’.
[12] Income was not defined but it was clear from the second reading speech of the Attorney-General, W M Hughes that the intention was that capital gains should not form part of the income tax base in order to encourage productive investment: Commonwealth, Parliamentary Debates, House of Representatives, 18 August 1915, 5845.
[13] ITAA15 s 14(1)(b) included in the income of any person: ‘dividends, interest, profits, or bonus credited or paid to any member, shareholder, or debenture-holder of a company which derives income from a source in Australia or of a company which is a shareholder in a company which derives income from a source in Australia’.
[14] ITAA15 s 10(1) levied tax, irrespective of residency considerations on taxable income from all sources in Australia.
[15] ITAA15 s 14(1)(b) second proviso, inserted by Income Tax Assessment Act (No 2) 1915 (Cth).
[16] ITAA15 s 16(1) repealed by Income Tax Assessment Act (No 2) 1916 (Cth). The same provision was re-enacted at s 16(1B) by Income Tax Assessment Act (No 2) 1916 (Cth).
[17] Section 16(2A) inserted by Income Tax Assessment Act 1918 (Cth).
[18] Income Tax Act (No 1) 1915 (Cth) sch 2 (rate on income from property), sch 3 (rate of tax on income of a company).
[19] Income Tax Act 1922 (Cth) s 5 imposed additional tax and sch 2 set the rate of tax on income from property.
[20] Income Tax Act 1922 (Cth) sch 4.
[21] ITAA15 s 16(2).
[22] Income Tax Assessment Act 1922 (Cth) s 21.
[23] ITAA15 s 25(1).
[24] ITAA15 s 26(1).
[25] ITAA15 s 25(2).
[26] ITAA15 s 14(c).
[27] ITAA15 s 27(1)-(2A).
[28] Income Tax Assessment Act 1922 (Cth) s 29 imposed tax only at the partner level, while s 30 taxed sui juris beneficiaries on income to which they were presently entitled, and taxed the trustee on income to which either no beneficiary was presently entitled or where the presently entitled beneficiary was under a legal disability. These key features of the Australian tax treatment of partnerships and trusts have changed little since 1922.
[29] Up to 1923 the provisions relating to dividend deductibility and assessability also applied to distributions to debenture holders. After 1923 the deductibility provisions no longer applied to dividends but continued to apply to interest payments. The provisions relating to payments to nonresidents continued to apply to interest payments after 1923. See the amendments to s 20 of the ITAA 1922 (Cth) by s 6 of the Income Tax Assessment Act 1923 (Cth).
[30] The obligation to pay tax on assessable income distributed to nonresident shareholders and the credit for the tax paid by the company were inserted by the Income Tax Assessment Act (No 2) 1916 (Cth). The option to withhold tax at source and to pay tax or withhold tax on interest paid to nonresidents and on distributions in relation to bearer debentures and share stock were inserted by the Income Tax Assessment Act 1918 (Cth).
[31] Given that the corporate rate, and from 1918 onwards the rate payable by companies on distributions to nonresidents, was considerably lower than the top marginal rate there was clearly a risk that nonresidents with Australian source income that placed them in the top marginal rate would not file returns. By 1945, when nonresidents were taxed on an assessment and prior year provisional tax basis, the Australian Commissioner of Taxation admitted in correspondence that it was not practical to enforce the Australian rules on taxation of nonresidents. Letter from I S Jackson (Australian Commissioner of Taxation) to Sir C Gregg (Chairman, UK Board of Inland Revenue) 25 September 1945, UK National Archives, Item Details DO 35/1157 File 0503/41.
[32] Pursuant to an agreement with Great Britain in 1921, the ITAA15 was amended to provide relief from international juridical double taxation. The relevant provision was s 12A inserted by the Income Tax Assessment Act 1921 (Cth). Section 12A interacted with s 27 of the Finance Act 1920 (UK) which introduced the Dominion Income Tax Relief system. The overall aim of the Dominion Income Tax Relief system was that the total tax on cross border income should not exceed tax at the higher of the two countries rates.
[33] Commonwealth, Royal Commission on Taxation, Report of the Royal Commission on Taxation: Second Report (1922), [295], Recommendation 1 and see discussion at [290] and [292].
[34] For a discussion of anti-dividend streaming and anti-franking credit trading rules see F Gilders, J Taylor, G Richardson, and M Walpole, Understanding Taxation Law: An Interactive Approach (2nd ed, 2004), 676-91.
[35] Limitations on the inter-corporate rebate were imposed in 1999 prior to the introduction of the Simplified Imputation System by restricting the rebate on dividends paid by listed companies to the franked portion of the dividend. This had been the position for some years in relation to dividends paid by unlisted companies. When the 1999 restrictions were introduced, resident companies receiving non-portfolio dividends that were not fully franked from other (non-group) resident companies received a deduction under s 46FA of the ITAA36 to the extent that they redistributed those dividends to nonresident companies.
[36] See J Pope, R Fayle and D L Chen, The Compliance Costs of Companies’ Income Taxation in Australia (1994), 69; J Pope, R Fayle, and D L Chen, The Compliance Costs of Public Companies’ Income Taxation in Australia 1986/87 (1991), 88; and J Pope, R Fayle, and M Duncanson, The Compliance Costs of Personal Income Taxation in Australia 1986/87 (1990), 48.
[37] Pope, Fayle and Chen, above n 36, 69.
[38] Ibid 67.
[39] Ibid 69.
[40] See K Messere, ‘Half a Century of Changes in Taxation’ (1999) 53 Bulletin for International Fiscal Documentation 340, 351 and OECD, Taxing Profits in a Global Economy: Domestic and International Issues (1991), 57, 164.
[41] See the discussion in P A Harris, Corporate/Shareholder Income Taxation And Allocating Taxing Rights Between Countries (1996), 303, 297, 291-3.
[42] The US only became involved in the Committee of Technical Experts of the Fiscal Committee of the League of Nations in 1927. For an account of the influence of the US representative, T C Adams, on the eventual consensus and on draft 1b of the 1928 drafts see M J Graetz and M M O’Hare, ‘The Original Intent Of US International Taxation’ (1997) 44 Duke Law Journal 1021, 1081-90.
[43] Significantly, none of the League of Nations draft model conventions nor any of the versions of the OECD model convention have, in the text of the model itself, said anything about the prevention of the economic double taxation of portfolio investors. The US Internal Revenue Code at the time only allowed an indirect foreign tax credit to parent companies owning at least 50 per cent of the relevant subsidiary company.
[44] Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income, 29 October 1946, Australia–United Kingdom, [1947] ATS 18 (entered into force 3 June 1947).
[45] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income, 14 May 1953, Australia–United States, [1953] ATS 4 (entered into force 14 December 1953).
[46] Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income, 1 October 1957, Australia–Canada, [1958] ATS 12 (entered into force 21 May 1958).
[47] Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income, 12 May 1960, Australia–New Zealand, [1960] ATS 6 (entered into force 23 June 1960).
[48] Commonwealth Treasury, Australian Tax Treaty Table (2006) <http://www.treasury.gov.au/docu-ments/625/XLS/Australian_Tax_treaty_tableSept2006.xls> at 28 September 2006.
[49] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income, 6 August 1982, Australia–United States, [1983] ATS 16, art 23 (entered into force 31 October 1983).
[50] See the summary of the Australian approach to DTA negotiations by L Bury, Treasurer, Cabinet Submission No 65, 5 January 1970, [8]. In this connection it is notable that the non-discrimination article under the Australia – United States Double Taxation Convention was not given the force of law. See International Tax Agreements Act 1953 (Cth) s 6.
[51] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 16 April 1945, United Kingdom–United States, 12 Bevans 671 (entered into force 25 July 1946).
[52] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 16 April 1945, United Kingdom–United States, 12 Bevans 671, arts VI(1)-(2) (entered into force 25 July 1946). The US withholding tax was reduced to 5 per cent in the case of dividends paid to parent companies owning at least 95 per cent of the subsidiary company.
[53] See the discussion in F E Koch, The Double Tax Conventions — Volume 1: Taxation of Income (1947), 73-4.
[54] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 16 April 1945, United Kingdom–United States, 12 Bevans 671 (entered into force 25 July 1946).
[55] See the discussion of the negotiations between Germany and the US in the 1960s in D R Tillinghast, ‘The Revision of the Income Tax Convention between the United States and the Federal Republic of Germany’ (1966) 21 Tax Law Review 399, 404-6. For a briefer summary which places these negotiations in a more general policy context see H J Ault, ‘Corporate Integration, Tax Treaties and the Division of the International Tax Base: Principles and Practices’ (1992) 47 Tax Law Review 565, 583-4. Ault notes that Germany had some success in achieving non reciprocal withholding tax rates in negotiations with treaty partners other than the US.
[56] For a discussion of the initial acceptance and ultimate rejection of ‘effective reciprocity’ in the commentary on the OECD Model see P A Harris, ‘Origins of the 1963 OECD Model Series: Working Party Twelve and Article 10, Part I’ [1999/2000] 1 Australian Tax Forum 3, 22-3.
[57] K Vogel, Klaus Vogel on Double Taxation Conventions (3rd ed, 1997), 1280.
[58] See the discussion in Graetz and O’Hare, above n 42, 1068-70.
[59] League of Nations, Double Taxation And Tax Evasion: Report Presented By The Committee of Technical Experts on Double Taxation and Tax Evasion (1927), League of Nations Doc C.216.M.85.1927.II.
[60] League of Nations, London and Mexico Double Tax Conventions: Commentary and Text (1946), League of Nations Doc II.A.7, Article XIV of the Mexico Draft, 4388.
[61] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 27 April 1932, France–United States, 7 Bevans 977 (entered into force 1 January 1936).
[62] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 25 July 1939, France–United States, 7 Bevans 1046 (entered into force 1 January 1945).
[63] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 4 March 1942, Canada–United States, 6 Bevans 244 (entered into force 15 June 1942).
[64] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 23 March 1939, Sweden–United States, 11 Bevans 809 (entered into force 14 November 1939).
[65] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 16 April 1945, United Kingdom–United States, 12 Bevans 671 (entered into force 25 July 1946).
[66] See M B Carroll, ‘Double Taxation Conventions Concluded by the United States Since 1939’ V (1947) Cahiers de Droit Fiscal International 25, 61-3.
[67] See A C Warren, ‘Income Tax Discrimination against International Commerce’ (2001) 54 Tax Law Review 131 at 147.
[68] See the discussion in J F Avery Jones et al, ‘The Non-discrimination Article in Tax Treaties, I and II’ [1991] British Tax Review 359, especially 443-4.
[69] Ibid.
[70] Commonwealth Government, Reform of the Australian Tax System: Draft White Paper (1985), ch 17. A summary of the proposed options is at [17.33].
[71] Ibid [17.23].
[72] Ibid [17.19].
[73] On the lack of obligation to extend imputation credits to nonresidents, see Commonwealth Government, Taxation Review Committee: Full Report (1975), [16.65]-[16.71]. On difficulties associated with a split rate system, see [16.35].
[74] Commonwealth Government, above n 70, 2.
[75] M Benge and T Robinson, How To Integrate Company And Shareholder Taxation: Why Full Imputation Is The Best Answer (1986).
[76] See the discussion in R J Vann, ‘Overview Of Company Tax Reform’ in R J Vann (ed), Company Tax Reform (1988) 3, 7.
[77] Commonwealth Treasury, RBT, A Tax System Redesigned (1999), 411.
[78] Ibid, 62.
[79] Australian Taxation Office (‘ATO’), Taxation Statistics 2002-03 (2005), 63, see especially Detailed Table 1: Company Tax: Selected Items, by Net Tax and Company Type, 2002-03 Income Year.
[80] ATO, Taxation Statistics 1999-2000 (2001), Table 12.2.
[81] ATO, Taxation Statistics 2002-03 (2005), 63, Detailed Table 2.
[82] If a policy decision were made to allow corporate tax preferences to pass through to closely held companies then a system similar to the US Subchapter S corporation would be another alternative. Arguably this option would be less problematic from an international point of view. It would, however, involve a significant shift of policy in relation to the tax treatment of distributions of corporate tax preferences. As this paper was written as a contribution within the parameters of a tax reform discussion in Australia that excludes fundamental changes in tax policy, this alternative will not be examined in detail in this paper. Incidentally, using a S corporation system to some extent would amount to a ‘back to the future’ solution given that between 1915 and 1922 the Commissioner allocated profits to shareholders where companies had not made a sufficient distribution. Allowing small companies the option to be taxed on a flow through basis was also proposed by the Asprey Committee. See Commonwealth, above n 73, [16.79]-[16.90].
[83] See C John Taylor, ‘Problems with Franking Credits Flowing through Partnerships and Trusts: The 2004 Amendments and a Simpler Alternative’ (2005) 34 Australian Tax Review 154, 188.
[84] As noted earlier the non-discrimination article in the Australia – United States Double Taxation Convention was not given the force of law by the International Tax Agreements Act 1953 (Cth) s 6.
[85] Protocol Amending the Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 19 June 1989, Australia–France, [1990] ATS 26 (entered into force 19 July 1990).
[86] Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 13 April 1976, Australia–France, [1977] ATS 21 (entered into force 21 September 1977).
[87] Protocol to the Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 12 July 1982, Australia–Korea, [1984] ATS 2, para 6 (entered into force 1 January 1984).
[88] Protocol to the Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 12 September 1984, Australia–Finland, [1986] ATS 6, para (b) (entered into force 28 March 1986).
[89] Protocol to the Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 24 March 1992, Australia–Spain, [1992] ATS 41 (entered into force 10 December 1992).
[90] Protocol to the Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 1 July 1999, Australia–South Africa, [1999] ATS 34, para 1 (entered into force 21 December 1999).
[91] Protocol to the Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 2 February 2000, Australia–Romania, [2001] ATS 4 (entered into force 11 April 2001).
[92] Protocol to the Agreement for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 9 September 2002, Australia–Mexico, [2004] ATS 4, para 11(b) (entered into force 31 December 2003).
[93] Convention for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with respect to Taxes on Income, 21 August 2003, Australia–United Kingdom, [2003] ATS 22 (entered into force 17 December 2003).
[94] Vogel, above n 57, 1331; J D B Oliver, ‘Current Notes: Similar Enterprises’ (1989) British Tax Review 141; J D B Oliver, ‘Current Notes: Differential Treatment or Discrimination’ (1993) British Tax Review 435.
[95] Vogel, above n 57, 1331; Oliver, ‘Similar Enterprises’, above n 94; Oliver, ‘Differential Treatment or Discrimination’, above n 94.
[96] Vogel, above n 57, 1331.
[97] NEC Semi Conductors Ltd v IRC [2006] STC 606, [34]-[41].
[98] Avery Jones et al, above n 68, 443. The comment is made in relation to split rate systems.
[99] 26 USC § 1361(b)(1)(C). For a discussion of the general requirements applying to US Subchapter S corporations see J K Mc Nulty, Federal Income Taxation of S Corporations (1992); Eustice and Kuntz, Federal Income Taxation of S Corporations (1985).
[100] US Treasury, Technical Explanation to the United States Model Income Taxation Convention <http://www.ustreas.gov/offices/tax-policy/library/techxpln.pdf> .
[101] Ibid [360].
[102] American Bar Association, Section of Taxation, Committee on S Corporations and Partnerships, ‘Report on the Comparision of S Corporations and Partnerships: Part I’ (1991) 44 The Tax Lawyer 483; American Bar Association, Section of Taxation, Committee on S Corporations and Partnerships, ‘Report on the Comparision of S Corporations and Partnerships: Part II’ (1991) 44 The Tax Lawyer 813, recommendation 4, [4.2]; S H Goldberg and P A Glicklich, ‘Treaty Based Non-Discrimination: Now You See It Now You Don’t’ (1992) 1 Florida Tax Review 51, 105.
[103] Goldberg and Glicklich, above n 102, 105.
[104] The election was introduced largely to accommodate NZ resident shareholders. In 1959 NZ continued to exempt dividends sourced in British Commonwealth countries from NZ tax. This meant that NZ shareholders who had interest expenses relevant to gaining dividends from Australian companies could not obtain a NZ tax deduction for those expenses. Hence, even though the rate of dividend withholding tax might be lower than the marginal rate which would otherwise apply to the shareholder the fact that it was levied on a gross rather than a net basis could mean that, where the NZ shareholder had expenses (typically interest) relevant to gaining her or his Australian dividend income the effective tax rate on a withholding basis could be higher than the effective rate on an assessment basis. The election enabled a NZ shareholder to claim interest and other expenses on her or his Australian source dividends as a deduction against tax levied by Australia on an assessment basis. The background to the election is explained in Commonwealth Committee on Taxation, Report (1961), 108-10.
[105] Ibid 45-6.
[106] R J Vann, ‘A Model Tax Treaty for the Asia-Pacific Region?’ (1991) 45 Bulletin for International Fiscal Documentation 99.
[107] See the discussion in Harris, above n 41, chs 7 and 8.
[108] Harris, above n 41, 518.