1. The Discussion Paper proposes amendments to the GAAR which are aimed at extending the powers of SARS to disregard genuine transactions honestly concluded and to determine their tax consequences in a manner which SARS regards as appropriate. It is fair to say that the Discussion Paper presents the SARS view on the alleged defects in section 103 of the Income Tax Act, 1962 ("the Act") and on the amendments required to remedy them.
    2. The SARS proposals give rise to other concerns. They seek to confer even greater powers on the collection agency and effectively nullify the safeguards against abuse as currently contained in the section. They significantly increase the level of uncertainty in relation to the tax consequences of common commercial transactions and raise important constitutional issues. They are also far reaching and will empower the agency to negate transactions which even Lord Templeman would have regarded as legitimate. SARS frowns upon sale and leaseback transactions; in England, the birthplace of the notion of impermissible tax avoidance, they have unanimously been pronounced legitimate by the House of Lords.
    3. Moreover, we do not accept that section 104 has been shown to be ineffective when applied to transactions such as those described in the Discussion Paper. It proved ineffective in cases such as Louw, Conhage and others (which we refer to below) only because, in our respectful submission, the allegations of abnormality made by SARS in those cases were unjustified. In relation to some of the transactions described in the GAAR was successfully applied by SARS; in relation to others SARS has not brought the cases before the Courts.
    4. SARS states sophisticated tax "products" are marketed by banks and other entities including "law firms". We are unaware of any law firm in South Africa which has in the past marketed, or which is currently marketing any tax "products". Typically the contingent nature of the fee arrangements referred to by SARS would be precluded by professional rules. Be that as it may, this memorandum has been prepared without the benefit of the views of law firms, if there are any, engaged in the marketing of such "products".

    1. The first point of principle to note is that a general anti-avoidance rule (a "GAAR") is not required in South Africa or, one would think, in any modern legal system, to combat criminal conduct on the part of taxpayers. All of the taxing statues and the South African common law deal adequately with such conduct. They also adequately cater for tax driven transactions which rely on non-detection by SARS for the achievement of the targeted tax benefits. Such conduct is illegal and universally castigated as tax evasion (SARS 2.2.1).
    2. The second principle worthy of note is that South African courts, applying ordinary principles of common law, deal adequately with attempts to manipulate the tax laws by means of artificial and contrived schemes that have little economic substance. (SARS 4.6). Our courts will not give effect to "a complex structure …. used to disguise the true nature" of a transaction (SARS 6.2.3) or to schemes which are arbitrary or contrived (SARS 4.2). In this context reference may be made to recent tax cases such as Erf 3183/1 Ladysmith (Proprietary) Limited v The Commissioner for Inland Revenue 1996 58 SATC 229 and Relier (Proprietary) Limited v The Commissioner for Inland Revenue 60 SATC 1 in which the Commissioner successfully challenged tax avoidance schemes manipulating the provisions of the tax statutes without even relying on the general anti-avoidance provision. The legal principle illustrated in these cases is well established in South Africa and is not confined to the field of tax law.
    3. The purpose of a GAAR is to strengthen the hands of the fiscus, i.e. the revenue collection officials, to alter the tax consequences of perfectly legitimate and genuine transactions in relation to which one or other of the Commissioner's officials might entertain a real or, even an imagined, grievance. It empowers the revenue official, to disregard the sanctity of commercial contracts lawfully and honestly concluded in determining the tax consequence to the parties thereto. As such, it has to be emphasised, a GAAR provides the tax collecting agency with a drastic and intrusive power. It is drastic because it empowers the tax collection agency to ignore the legal consequences of honest bargains lawfully concluded; it is at the same time intrusive in that it limits the right of the taxpayer freely to conduct its affairs in a manner which would attract the least amount of tax.
    4. A GAAR in a constitutional democracy such as South Africa, and indeed in any progressive economy, requires the balancing of the interests of business, on the one hand, against the strong arm of the fiscus. Almost 60 years ago Schreiner JA, in considering the GAAR then contained in the 1941 Act, expressed himself in the following terms (CIR v King 14 SATC 184 at 199) :
    5. "It seems to aim at a truer or fairer determination of the liability to the taxes imposed by the Act and at their due payment when so determined. It is intended, I think, to deal with cases in which the Commissioner, as representing the fiscus, is properly aggrieved by a transaction or operation designed to enable one of the parties thereto to escape tax. The Commissioner is not properly aggrieved merely because at a stage before income has accrued to a taxpayer it might have been predicted with confidence amounting even to certainty, that if the taxpayer took no steps in the matter such income would accrue to him and because he then takes the avoiding steps. But the Commissioner would be properly aggrieved if a transaction or operation were entered into which prevented income from accruing to the taxpayer, while leaving him in the position of one to whom the income would normally and naturally accrue. The section is not, in my opinion, designed to implement the expectations, however reasonable, of the Commissioner that there will be no change in the taxpayer’s affairs which will result in his getting less income; it is designed to meet the Commissioner’s objections to the creation of abnormal or unnatural situations, to the detriment of the fiscus." (Our emphasis)

      It is submitted that the same sentiment is expressed in the passage quoted at 2.4 of the Discussion Paper from the New Zealand case of CIR v BNZ Investments [2002] 1 NZLR 450.

    6. It is for this very reason that a GAAR normally does not confer an unfettered discretion on revenue officials; it usually contains appropriate safeguards against the abuse of the drastic powers conferred thereby. In South Africa these safeguards are embodied, for all practical purposes, in the so-called "abnormality" requirement and the special provisions contained in section 103 of the Income Tax Act, 1961 with regard to the onus of proof. The Discussion Paper proposes the negation of these very safeguards on grounds which in our submission are not justified.

    1. In the Discussion Paper SARS seeks to draw a distinction between impermissible tax avoidance (SARS 2.2.2) and "legitimate tax planning" (2.2.3). The notion of impermissible tax avoidance is not a concept recognised in South African law. It has its origin in the Courts of England. The parameters of impermissible tax avoidance and the features which would distinguish it from legitimate tax planning have been the subject of considerable judicial controversy in England. The most recent decisions of the House of Lords have avoided reference to the concept. See for instance, Barclays Mercantile Business Finance Ltd v Mawson (Inspector of Taxes) [2005] 1 ALL ER 97.
    2. For present purposes it is sufficient to note that the concept of a GAAR is foreign to the laws of England. The legislature in that country has contented itself with a legal system in which the courts, applying principles of English common law, are required to resolve disputes with regard to the scope of legitimate tax planning. This is the environment in which the English courts, with Lord Templeman at the forefront, grappled with the notion of impermissible tax avoidance without ever defining its scope.
    3. South Africa does have a GAAR and has had one for at least the last 65 years. It has a long history of case law in which legal principles on the subject of tax avoidance and the scope of our anti-avoidance provisions have been established and successfully applied. The reliance in the Discussion Paper on ill-defined foreign notions in the quest to erode the existing safeguards contained in section 103 is in our submission misplaced. A GAAR operates not only within the framework of the taxing statute but also within the legal system as a whole. Thus, whilst the Courts of England had to develop new theories to address complex tax avoidance schemes, there was no need for this in South Africa. For instance, in FA + AB Ltd v Lupton (Inspector of Taxes) [1971] 3 ALL ER 948, the House of Lords found a dividend stripping scheme to be impermissible; the same scheme failed in South Africa on the application of ordinary principles of law in CIR v Nemojim (Proprietary) Limited 45 SATC 241. The film scheme considered by the House of Lords in Ensign Tankers (Leasing) Ltd v Stokes (Inspector of Taxes) [1992] 2 ALL ER 275 failed in South Africa when SARS invoked section 103. It is submitted that many of the transactions considered by the Court’s of England as examples of impermissible tax avoidance, will fail in South Africa by reason of section 103.

    1. SARS contends (at 8.3.1) that section 103 has proven to be an inconsistent, and at times, ineffective deterrent "to abusive avoidance schemes and other impermissible tax avoidance". The first problem with this proposition is the content of what SARS regards as abusive avoidance schemes and impermissible tax avoidance. We confine ourselves to three examples taken from the reported cases.
      1. In CIR v Louw 1983 (3) SA 551; 45 SATC 113 a partnership of professional engineers "incorporated" their practice so that they could henceforth practise as a company as opposed to a partnership. The "scheme" entailed the incorporation of a new company with the partners as the directors and only shareholders of the company. The partners then sold their practice to the company on credit, without requiring the payment of interest and with the payment terms depending on the financial circumstances of the new company. The Appellate Division (now the Supreme Court of Appeal) unanimously held these arrangements to be "perfectly sound and businesslike" and rejected the Revenue contention that they were abnormal. Importantly (we revert to this below) the court accepted "the realities of the situation" which were that the company to which the practice was sold by the partners had as its shareholders and directors the self-same partners and was controlled by them.
      2. In CIR v Conhage (Proprietary) Limited 1994 (4) SA 1149; 61 SATC 391 the taxpayer entered into a transaction which, in the world of finance, is internationally described as a "sale and leaseback transaction". The facts may be summarised briefly. The taxpayer was the owner of a manufacturing plant. It required finance for expansion and, instead of simply borrowing the money, it accepted a proposal from its bankers to conclude a sale and leaseback transaction because such a transaction improves the after-tax cost of funds to the taxpayer. The taxpayer sold its manufacturing plant to a bank and immediately hired it back so as not to lose the use thereof. Predictably, the highest court in this country rejected the Revenue allegation that transactions of this nature are "abnormal". It is important to emphasise that even in England, the very source of the concept of impermissible tax avoidance, so passionately embraced by SARS in the Discussion Paper, sale and leaseback transactions have recently been accepted as perfectly legitimate transactions. See Barclays Mercantile Business Finance Ltd (supra).
      3. In ITC 1712 63 SATC 499 the taxpayer concluded a financial lease for a period of 61 months in respect of container tanks. The Bank as lessor, required the payment of an initial rental for the first year of the lease equal to 43% of the rentals payable over the full period of the lease. Relying on section 103(1) SARS disallowed the deduction of the full amount of the initial rental contending that irregular rental payments over the period of the lease are abnormal and agreed to solely or mainly with a tax avoidance purpose. Again, predictably, the Tax Court rejected the SARS contentions. The initial rental was required by the Bank concerned in order to reduce its credit risk.

    2. The above examples illustrate that perceptions of what constitutes impermissible tax avoidance are not absolute. Moreover, the reported cases show that the SARS perception of the content of impermissible tax avoidance is often commercially unrealistic and contrary to international trends. In our submission the above examples illustrate the need for proper safeguards against the abuse of a GAAR by a collection agency seeking to invoke it in respect of transactions which, after due process, have been held to be perfectly sound and businesslike.
    3. The second problem with the SARS proposition (at SARS 8.3.1) is that it is not supported by readily available empirical evidence with regard to the fate of the "schemes" referred to in the Discussion Paper. Without attempting an exhaustive analysis of the efficacy of the current GAAR the following observations illustrate in our view that the SARS proposition is at least open to doubt.
      1. In ITC 1606 58 SATC 328 the Cape Tax Court upheld the SARS attack on a transaction referred to in the Discussion Paper (at 6.2.3) as a "bare dominium scheme". In that case the SARS attack was based on section 103(1) and why SARS, armed with the precedent established by this case, did not attack all of the bare dominium schemes remains a mystery. Certainly, the case does not support the contention that the current provision was found ineffective in relation to these schemes.
      2. Next there are references in the Discussion Paper to the so-called "film schemes". They are described in some detail in Annexure A to the Discussion Paper and, perhaps less objectively, by Lord Templeman in his speech in the House of Lords in Ensign Tankers (Leasing) Limited (supra). The targeted tax benefits were disallowed by the House of Lords as an illustration of what the learned law Lords regarded as the application of the concept of impermissible tax avoidance. In an unreported case in the Tax Court of South Africa they were disallowed by the application of section 103, the proudly South African GAAR, without resort to an ill-defined and foreign doctrine. In the only reported case in South Africa on film schemes [Rane Investments Trust v C:SARS 65 SATC 333] SARS en route to the Supreme Court of Appeal inexplicably abandoned its reliance on section 103. It remains just another mystery to South African tax lawyers why SARS, armed with precedent in the Tax Court, did not attack all of the film schemes described in Annexure A to the Discussion Paper.
      3. Next, there are references in the Discussion Paper to convertible loans (SARS 10.2). Essentially they are loans requiring the capital to be settled in specie i.e. by the delivery of something other than cash. Typically and most commonly, but not invariably, a convertible loan is settled by the issue of shares issued by the borrower at a subscription price equal to the capital outstanding under the loan. It is but an example of the internationally recognised concept of convertible debentures. The tax consequences of such transactions differ from country to country and are largely dictated by the nuances of the legal and tax environment of the country concerned. Notably, in South Africa the generally accepted tax consequences of a convertible loan have not been challenged by SARS under section 103 and it is accordingly not open to SARS to contend that the GAAR is ineffective in addressing its grievances in relation to these transactions. Certainly, to the extent that it involves trickery and pretence, as SARS contends (at 10.2) that it does, it will be struck down by our courts either in terms of common law principles [our courts do not countenance the pretence] or in terms of section 103.
      4. In assessing the efficacy of our GAAR, it is as well to have regard to those cases in which SARS successfully invoked section 103 to negate the targeted tax benefits of specific transactions. An analysis of all of the reported cases is beyond the scope of these submissions. A reference randomly selected from volume 63 of the SATCs (the primary publication of judgments delivered in the Tax Courts) to ITC 1714 63 SATC 507 and ITC 1699 63 SATC 175 illustrates the effectiveness of the GAAR, read with established principles of common law.

    4. A review of the empirical evidence presented by the reported cases suggests that there is insufficient cause for further eroding the safeguards to the potential abuse of the powers conferred on the tax collector by section 103 of the Act. We favour the approach that tax disputes, alike with all other commercial disputes, should be resolved only by resort to courts of law. In that context, equity, common sense and the Constitution dictate that the procedural rules should be fair. The LSSA is opposed to rules which would render the contest uneven and the outcome inevitably predictable.

    1. The motivation in the Discussion Paper for the proposed amendments is based on an incorrect premise that the GAAR does not cater effectively for what SARS considers to be "impermissible tax avoidance". In our view the reported cases illustrate that it does. But there are other equally fundamental reasons for not supporting the proposed amendments. Under this heading we confine ourselves to the technical aspects of the SARS proposal.
    2. The most fundamental and salutary safeguard enshrined in the current section 103 is the requirement that in any proceedings involving its application the burden of proving the alleged abnormality of the transaction lies with SARS. Allegations of abnormality are frequently made injudiciously without regard to the peculiar circumstances of a particular case. Reference may be made in this regard to CIR v Louw (referred to above at 4.1.1). It is an appropriate safeguard against abuse and consistent with established procedural rules that the burden of proof in relation to an allegation of abnormality should be on SARS. It is this very safeguard which the SARS proposal seeks to nullify.
    3. That nullification is achieved by the proposed section 103(2) read with the proposed section 103(4)(b). It might as well be legislated that whenever the Commissioner invokes section 103 it is for the taxpayer to prove the normality of the transaction. It is difficult to imagine a transaction which even SARS would regard as "legitimate tax planning" in relation to which the onus of proof would not be shifted to the taxpayer in terms of the proposed amendment. It should be noted that in terms of section 82 the burden of proof in all tax disputes is placed on the taxpayer. The abnormality requirement in section 103 is an exception to this general approach precisely because it is intended to be a safeguard against the abuse of injudicious allegations of abnormality. As is noted in the Discussion Paper (2.4) the GAAR is not a revenue raising provision and, in our submission, there is no reason why the burden of proof should not remain with SARS.
    4. The most alarming feature of the proposed amendment is that contained in section 103(2)(h) which is rather obviously directed at the finding of Corbett JA (as he then was) in Louw’s case. The proposed amendment requires that a court of law should ignore what Corbet JA in that case described as "the realities of the situation" and castigate as abnormal what was found to be "perfectly sound and businesslike arrangements". The proposed amendment is unsound. It will render abnormal, unless the taxpayer proves otherwise, all transactions between related parties even if in the circumstances of a particular case they are perfectly sound and businesslike. (As they were found to be by the Supreme Court of Appeal in Louw’s case).
    5. An example of such a transaction is the interest free loan without security by a holding company to its wholly-owned subsidiary with no fixed terms for repayment. Such a transaction is perfectly sound and businesslike having regard to the fact that there is a special relationship between lender and borrower. The amendment proposed by SARS is that this special relationship should be ignored; the transaction must be castigated as abnormal, unless the contrary is proved, because a lender of money does not make unsecured interest free loans to an unrelated borrower. Ironically, on this aspect, it is SARS which proposes a legislative directive that our courts should ignore the substance and reality of a given transaction, take a blinkered view of the facts and decide the issue without having regard to the relationship between the parties. Imagine the absurdity of a holding company incurring the costs of a mortgage bond to secure a loan to its subsidiary simply to avoid the allegation of abnormality by some or other SARS official!

    1. As has been submitted, the erosion of the safeguard contained in the GAAR by shifting the burden of proof to the taxpayer cannot be supported. For as long as the safeguard is retained, it makes no sense to permit SARS to raise section 103 in the alternative. The reason for this submission lies in the rules governing proceedings in a court of law. To permit SARS to raise section 103 in the alternative, is to make mockery of the proceedings in the Tax Court.
    2. In terms of section 34 of the Constitution, 1996 a taxpayer has a right to a fair hearing before an independent and impartial tribunal. The content of section 34 was considered by the Constitutional Court in, inter alia, De Beer NO v North-Central Land Council and South-Central Local Council and Others (Umhlatuyuzana Civic Association Intervening) 2002 (1) SA 429 (CC). At paragraph [11] of the judgment the following is stated –
    3. "This s34 fair hearing right affirms the rule of law, which is a founding value of our Constitution. The right to a fair hearing before a court lies at the heart of the rule of law. A fair hearing before a court as a prerequisite to an order being made against anyone is fundamental to a just and credible legal order. Courts in our country are obliged to ensure that the proceedings before them are always fair. Since procedures that would render the hearing unfair are inconsistent with the Constitution courts must interpret legislation and Rules of Court, where it is reasonably possible to do so, in a way that would render the proceedings fair. It is a crucial aspect of the rule of law that court orders should not be made without affording the other side a reasonable opportunity to state their case."

    4. By the simple contrivance of invoking the provisions of section 103(1) in the alternative, the Commissioner effectively escapes all the implications to a litigant burdened with the onus of proof as is exemplified by the following example –
      1. Assume that on the facts of Nemojim (supra) SARS raises an assessment disallowing the losses claimed by the taxpayer in that case on the ground that those losses do not qualify for deduction in terms of section 11(a) of the Act. In terms of section 82, the burden of proof is on the taxpayer. In practice this means that the taxpayer must produce the witnesses and permit them to be cross-examined by the SARS representative.
      2. However, should the assessments be raised only in terms of section 103 it is SARS which bears the onus of proof. The SARS witnesses may then be cross-examined by the taxpayer's representatives.
      3. A difficulty arises where section 103 is raised in the alternative. The issues that are required to be proved are invariably similar. The result is in respect of the same issues the burden of proof is at the commencement of the trial on both the taxpayer (in relation to its right to the deductions claimed) and on SARS (in relation to the abnormality requirement).
      4. The result is unfair. It is an unfair procedure because, if it were permitted, it would deprive the taxpayer of the protection of the incidence of the evidential burden in relation to an important jurisdictional prerequisite to the application of section 103(1). In effect, by the simple device of an allegation made in the alternative, the Commissioner is rendered entitled to cross-examine witnesses on allegations in respect of which he bears the onus of proof.

    5. Of course, to the extent that burden of proof is in terms of the SARS proposals shifted to the taxpayer, the above criticism loses its force.

    1. SARS correctly points out (9 at p 44) that a GAAR often results in uncertainty with regard to the ultimate tax consequences of transactions lawfully concluded by taxpayers. SARS appears to favour the view (at p 46) that, although certainty is important, it is not an absolute value; a GAAR cannot be overly precise if it is to be effective.
    2. Depending on the view that one takes of tax avoidance as a "problem" legal certainty may well have to yield to harsh revenue collection tactics. However, this reasoning cannot apply to the imposition of penalties on "promoters" of tax avoidance schemes. We favour the age-old adage that there can be no penalty without a crime. The suggestion that "promoters" engaged in perfectly legal commercial activity, but frowned upon by SARS, should be exposed to some sort of penalty will undoubtedly raise issues of a constitutional nature. The suggestion cannot be supported.

    1. In Nemojim (above), the Supreme Court of Appeal said the following at 45 SATC 267.
    2. "It has been said that there is no equity about a tax. While this may in many instances be a relevant guiding principle in the interpretation of fiscal legislation, there is nevertheless a measure of satisfaction to be gained from a result which seems equitable, both from the point of view of the taxpayer and from the point of view of the fiscus. And it may fairly be inferred that such a result is in conformity with the intention of the legislature."

      Taxing provisions and practices which operate unfairly and oppressively on honest taxpayers engaged in lawful commercial transactions do nothing to promote respect for the law and a culture of compliance. In our view, the SARS proposals tip the scales too heavily in favour of the fiscus and, if there were indeed to be a need for amendments, more equitable alternatives should be explored.

    3. Moreover, we do not consider it appropriate that the "uneasy tension" between a GAAR and the basic notion of the rule of law (SARS at 45) should be understated, as it is in the Discussion Paper. The more Draconian the powers conferred by a GAAR, the greater this "uneasy tension" becomes. As has been held by the Constitutional Court in a number of cases the responsibility to promote and protect constitutional values is imposed by the Constitution, 1996 on all organs of State including the Legislature. This constitutional obligation on Parliament requires that proper and effective safeguards be incorporated in all provisions confering discretionary powers on SARS. In our submission, it is not appropriate that the existing safeguards in section 103 should be further diluted.






LSSA Tax Committee

22 March 2006