Chapter 4 - Does the Use of General Anti-Avoidance Rules to Combat Tax Avoidance Breach Principles of the Rule of Law?

A Comparative Study[1]

Rebecca Prebble* and John Prebble§

4.1 Introduction

4.1.1 The rule of law, property rights, taxation, and tax avoidance

“The rule of law” is a compendious term for a number of related values that people generally think good laws should adhere to.[2] Dicey’s familiar formulation held that the rule of law requires “the absolute supremacy or predominance of regular law as opposed to the influence of arbitrary power”.[3] It is theoretically possible to interpret this condition as requiring merely that there should be laws, as opposed to a series of isolated commands. Nevertheless, theorists writing since Dicey have supplemented Dicey’s basic formulation with a number of additional requirements that the statement logically must entail if it is to be of value. In the present context, the most important of these is that the law should be capable of guiding people. In order to guide people, laws must be relatively clear and their application relatively certain; otherwise no one will know what is permitted and what is forbidden.

That laws should be relatively certain seems at first to be a reasonable demand. Indeed, governments generally manage to ensure that their laws adequately satisfy this condition. However, the criterion has proven very difficult to satisfy in the field of taxation.[4]

Viewing this issue in terms of the New Zealand Law Foundation’s Regulatory Reform Project, we see that taxation, and especially income taxation, gives rise pervasively to two of the New Zealand questions that inform the project: issues of property rights, and issues of the rule of law.

As to property rights, taxation at its base involves the forced taking of property.[5] Modern tax systems almost invariably provide that the property to be taken is specified in terms of money, but the money in question is taken by state fiat,[6] a fiat that may ultimately be enforced by taking of property for sale to discharge tax debts.[7]

As to the rule of law, regulations that are uncertain attract considerable criticism because they fail to provide people with sufficient information about what is required of them. A certain lack of clarity is, however, unavoidable since even the most specific rules will always give rise to borderline cases.

The rule of law principle of certainty requires that the legislature (or other law-maker) should specify with precision what actions constitute crimes, so the fact that taxation entails the compulsory taking of property gives rise, to state the matter in generic terms, to the need to define precisely what property is to be taken. More particularly, in terms of taxation, the need is precisely to define the tax base. That is, in income tax terms, what gains fall into the net of taxable income?

In turn, this general question about income taxation gives rise to a number of more specific rule of law issues. A function of this chapter is to show generically how these issues arise. The chapter identifies and analyses relevant qualities of the principle of the rule of law and, by way of example, examines and evaluates general anti-avoidance rules, such as s BG 1 of the New Zealand Income Tax Act 2007, in the light of this discussion. The chapter concludes by identifying several other aspects of regulation as applied in the context of taxation that give rise to similar concerns, though, at first sight, less obviously than does anti-avoidance legislation.

4.1.2 The rule of law and tax avoidance

Tax avoidance is a problem for every country. Avoidance is not evasion. Evasion means lying about one’s income. For example, cash businesses may under-state their takings or fail to file any return of tax at all. Avoidance is not mitigation. “Mitigation” is not a term of art, but in this chapter, and generally in the present context, it means reducing one’s tax in ways that a governing statute clearly encourages or permits; for example, taking a deduction for a gift to charity.[8]

Avoidance is between the two. Avoidance means, approximately, contriving transactions, typically but not necessarily artificial in nature, to reduce tax that would otherwise be payable according to what appears to be the policy of the taxing provision in question. This is a description rather than a definition. Terminology in the area is controversial. Some people deny that we can draw a meaningful distinction between avoidance and mitigation. Some people deny that the word mitigation has any right to exist as a meaningful term in this context.[9]

As a general rule, the law does not require people to arrange their affairs so that they incur the greatest possible tax liability. When faced with two possible ways in which to organise their money, taxpayers are legitimately entitled to choose the option that requires them to pay the lesser amount of tax. There comes a point, however, where governments begin to think that taxpayers are going too far in their attempts to decrease their tax liability: at this point, taxpayers cease to engage in legitimate tax mitigation and embark on unacceptable tax avoidance.

Useful definitions of the point at which tax mitigation becomes tax avoidance are elusive. Lord Denning has said that for an arrangement to constitute tax avoidance, “you must be able to predicate … that [the arrangement] was implemented in that particular way so as to avoid tax”.[10] This definition brings us no closer to knowing what constitutes tax avoidance, because all it says is “tax avoidance arrangements are those arrangements that look like tax avoidance arrangements”. Nevertheless, the definition highlights the difficulty of exhaustively defining tax avoidance, or, indeed, the difficulty of defining tax avoidance in terms of legal rules at all.

Tax avoidance is perhaps best understood through examples, rather than by analysis. Examples of tax avoidance transactions from different jurisdictions abound. They tend to have a number of identifiable features, for example, artificiality,[11] lack of business or economic reality, [12] lack of true business risk,[13] and the exploitation of statutory loopholes.[14] Avoidance often involves taxpayers exploiting rules that were designed to reduce unfairness in the tax system[15] or using existing legal structures in enterprising ways that the legislature, had it thought about the matter, would not have approved.[16]

To help to recognise avoidance, take, for example, Inland Revenue Commissioners v Bowater Property Developments Ltd,[17] a United Kingdom case that the House of Lords decided in 1988. The case is a particularly good illustration because, although the avoidance scheme employed was conceptually simple, its various steps were complex enough to involve the artificiality and complexity that is typical of tax avoidance cases. Nevertheless, unlike many avoidance cases it can be described and analysed in a few paragraphs. Bowater involved development land tax, a kind of capital gains tax that applied to land sales if the development value component of the sale was valued at more than £50,000. In a transaction potentially caught by the tax, Bowater proposed to sell land for more than £250,000 to a company called Milton Pipes Limited.

Instead of selling the land as one parcel, Bowater segmented the land into five undivided shares. It sold one share to each of five sibling companies in the Bowater group for £36,000 per share. Land in undivided shares looks just like land: there was no sub-divisional survey. There were no separate titles. The five Bowater companies owned the land in one title, just as a married couple owns their home in one title. The Bowater companies were a sort of modern marriage with five spouses. These five sales had no effect on the beneficial ownership of the land (using “beneficial” in its substantive sense rather than with the meaning that obtains in trust law). Both before and after the sales the ultimate owners were the shareholders in the Bowater group.

The five companies then sold their undivided shares to Milton Pipes for £50,000 each. That is, each company bought for £36,000 and sold for £50,000, making a profit of £14,000, well under the threshold of £50,000.

Legally, there were five separate sales from Bowater and five more sales to Milton Pipes. Economically, there was just one sale from Bowater to Milton Pipes. Ignoring this economic reality, however, the House of Lords treated the transactions as genuine. Bowater accordingly escaped development land tax. Had the case arisen in New Zealand the result would probably have been different, as will become apparent from a study of the New Zealand avoidance cases that are discussed in this chapter.

4.1.3 General anti-avoidance rules

Typically, governments combat avoidance by adding specific and often very detailed rules to tax legislation, being rules that frustrate one kind of avoidance transaction or another. For instance, jurisdictions might allow taxpayer companies to carry losses forward and to set them off against the profits of future years. As an anti-avoidance measure, such jurisdictions tend to require certain minimum continuity of ownership between the loss year and the profit year.[18] Tax statutes are replete with such rules. However, specific anti-avoidance rules cannot combat the more creative forms of tax avoidance that employ transactions that governments cannot predict. Consequently, many tax systems feature general anti-avoidance rules in addition to specific ones.

There is considerable variation in the form that general anti-avoidance rules take in different countries. Nevertheless, the various forms have roughly the same effect, at least in theory. General anti-avoidance rules allow tax authorities to disregard schemes that would otherwise reduce tax liability. The transactions to which they apply are void for tax purposes. A transaction being void, the tax lies where it falls, although modern general anti-avoidance rules often allow tax authorities to reconstruct a transaction to reflect the economic reality of the circumstances and to tax the taxpayer on the basis of the reconstructed transaction.[19]

An example of a typical general anti-avoidance rule is s 99 of New Zealand’s Income Tax Act 1976 (New Zealand’s current rule is not so readily quotable because it is disaggregated into several elements,[20] but it has roughly the same meaning and effect). Section 99 relevantly reads:

Every arrangement made or entered into, whether before or after the commencement of this Act, shall be absolutely void as against the Commissioner for income tax purposes if and to the extent that, directly or indirectly,—

(a) Its purpose or effect is tax avoidance; or

(b) Where it has 2 or more purposes or effects, one of its purposes or effects (not being a merely incidental purpose or effect) is tax avoidance, whether or not any other or others of its purposes or effects relate to, or are referable to, ordinary business or family dealings,—

whether or not any person affected by that arrangement is a party thereto.

Despite the great difference between the legal systems and cultures of the two countries, the corresponding German rule is to very similar effect:[21]

(1) The Tax Act may not be circumvented by an abuse of possible legal arrangements. If there is such an abuse, the taxpayer shall be taxed as if he had chosen an adequate legal arrangement.

(2) Subsection 1 is applicable if its applicability is not excluded expressly by the law.

Countries that have anti-avoidance rules broadly similar in form to New Zealand’s and Germany’s include Canada,[22] South Africa,[23] Hong Kong,[24] and France.[25] The rule in Australia was formerly similar,[26] but since 1981 has been framed in much more detail.[27] The United Kingdom does not have a statutory general anti-avoidance rule, but it does have a judicially developed anti-avoidance rule that can sometimes have roughly the same effect. This United Kingdom common law anti-avoidance doctrine was first propounded by the House of Lords in W T Ramsay Ltd v IRC.[28] At the risk of gross over-simplification, one can say that the common law anti-avoidance doctrine essentially allows the court to look at a series of transactions and to determine whether the transactions have any economic purpose other than the avoidance of tax. There have been suggestions in the United Kingdom that its common law anti-avoidance doctrine is insufficient to combat tax avoidance and should be replaced by a statutory general anti-avoidance rule,[29] but so far these suggestions have not been taken up. As a result, United Kingdom courts continue from time to time to deliver judgments in tax cases that are hard to defend, even to a common lawyer steeped in the study of formalism. For instance, in Mayes v Revenue and Customs Commissioners Toulson LJ had to add this brief conclusion to his judgment:[30]

I add a brief summary to explain the reason for my reluctant concurrence in a result which instinctively seems wrong, because it bears no relation to commercial reality and results in a windfall which parliament cannot have foreseen or intended.

Until 2010, the United States was similar to the United Kingdom in that it resisted pressure to enact a statutory general anti-avoidance rule.[31] Instead of a statutory anti-avoidance rule, the United States had a judicially developed anti-avoidance rule, which was first established by the Supreme Court in Gregory v Helvering.[32] The rule is often referred to as the economic substance doctrine. It operated in a similar manner to the United Kingdom judge-made rule. In 2010, however, the United States codified its economic substance doctrine. It did so by means of a somewhat improbable vehicle: the Health Care and Education Reconciliation Act of 2010,[33] which was primarily concerned with sweeping changes to the United States’ healthcare system. At the time of writing, the United States’ new statutory general anti-avoidance rule has not yet been tested, but it is expected to operate in much the same way as statutory rules in other countries. Nevertheless, the fact that most major jurisdictions have now adopted general anti-avoidance rules suggests that New Zealand was on the right track when it enacted what appears to have been the world’s first general anti-avoidance rule, s 29 of the Property Assessment Act 1879, the ancestor of the current s BG 1 of the Income Tax Act 2007.

Some civil law countries rely on the “abuse of rights” concept, which forbids the use of rights for improper purposes.[34] Others have statutory general anti-avoidance rules with broadly the same effect as those found in common law countries.[35] The different forms that general anti-avoidance rules take do not affect associated rule of law issues; problems and justifications that concern general anti-avoidance rules are equally relevant to all of them.

4.1.4 How do general anti-avoidance rules breach the principles of the rule of law?

The exact content of the concept of the rule of law is the focus of a continuing debate between legal theorists.[36] Nevertheless, as far as certainty is concerned there is near unanimity: most, and probably all, legal philosophers consider that a law must be relatively certain in order to conform to the principles of the rule of law.[37] In terms of taxation, the principle of certainty requires that the law should precisely specify the income figure that is subject to tax. It is this requirement of certainty that general anti-avoidance rules offend. Although a number of countries have statutory general anti-avoidance rules, the legislation adds little to the common understanding of what constitutes tax avoidance. In most jurisdictions there is uncertainty as to which transactions fall inside the general anti-avoidance rule.

In this context, “certainty” and “uncertainty” refer to certainty as to the effect of transactions and structures for tax purposes. Start from the proposition that tax law follows the general law. That is, for instance, the effect of a contract for tax law is the same as its effect for the general law. General anti-avoidance rules say that, for transactions that they affect, the tax consequences of arrangements are different from their consequences at general law. Indeed, general anti-avoidance rules sometimes provide that taxpayers will be taxed not on the legal transactions that they have entered but on notional, different, transactions, constructed by the Commissioner.[38] A lack of certainty arises because taxpayers say that they are not certain sure whether the general anti-avoidance rule applies to their circumstances. How does this uncertainty arise?

The uncertainty surrounding tax avoidance stems from the fine line that separates unacceptable tax avoidance from acceptable tax mitigation. Lord Templeman in Challenge Corporation Ltd v CIR[39] considered the two concepts with reference to s 99 of the Income Tax Act 1976, as the New Zealand general anti-avoidance rule was then numbered. His Lordship took an example from United Kingdom practice, namely a covenant to assign income, which, if in due form and for a duration of at least six years, can shift liability for tax from the assignor to the assignee. He said:[40]

Income tax is mitigated by a taxpayer who reduces his income or incurs expenditure in circumstances which reduce his assessable income or entitle him to reduction in his tax liability. Section 99 does not apply to tax mitigation because the taxpayer’s tax advantage is not derived from an “arrangement” but from the reduction of income which he accepts or the expenditure which he incurs.

Thus when a taxpayer executes a covenant and makes a payment under the covenant he reduces his income. If the covenant exceeds six years and satisfies certain other conditions the reduction in income reduces the assessable income of the taxpayer. The tax advantage results from the payment under the covenant.


Section 99 does not apply to tax mitigation where the taxpayer obtains a tax advantage by reducing his income or by incurring expenditure in circumstances in which the taxing statute affords a reduction in tax liability.

Section 99 does apply to tax avoidance. Income tax is avoided and a tax advantage is derived from an arrangement when the taxpayer reduces his liability to tax without involving him in the loss or expenditure which entitles him to that reduction. The taxpayer engaged in tax avoidance does not reduce his income or suffer a loss or incur expenditure but nevertheless obtains a reduction in his liability to tax as if he had.

Although it is generally accepted that general anti-avoidance rules apply to tax avoidance and not to tax mitigation, drawing the line between the two is often problematic. A literal application of general anti-avoidance rules would catch many legitimate transactions.[41] General anti-avoidance rules therefore mean something more than their bare words.

4.1.5 Why are general anti-avoidance rules especially Bad?

The preceding sections of this chapter have demonstrated that general anti-avoidance rules are vague. However, all legislation is vague to some extent. The most specific of rules will always have borderline cases. Why, then, do some people single general anti-avoidance rules out as particularly egregious breaches of the rule of law?[42] Drafters of most laws cannot foresee all relevant fact situations. As Hart pointed out, all laws admit of “core” situations, where the law will definitely apply, and “penumbra”, where it is less certain whether the law will apply.[43] To criticise general anti-avoidance rules because it is not clear whether they apply in some situations appears to subject them to a higher standard than we demand of law in general.

The difference is that general anti-avoidance rules have far larger penumbras than most laws. Arguably, general anti-avoidance rules are nothing but penumbra. The reason why legislators decide that they need general anti-avoidance rules is that situations where the rules may be needed cannot be defined in advance. If legislators could foresee all varieties of tax avoidance, they would pass specifically targeted rules to frustrate those endeavours. No doubt, most tax policy makers could give examples of the sorts of arrangement that might be caught by general anti-avoidance rules, but these examples would be cases that have been found to constitute avoidance in the past. The fact that general anti-avoidance rules exist at all is evidence that policy-makers and legislators themselves cannot predict what structures taxpayers will eventually contrive. The following sections of this chapter examine the deeper values that the requirement of certainty seeks to preserve, and consider whether general anti-avoidance rules truly offend those values. If they do so, are there situations in which the rule of law must give way to countervailing considerations? And is tax avoidance one of those situations? An important factor is public tolerance of general anti-avoidance rules. It appears that the rule of law is seen as more important in some areas of law than in others. This chapter examines why this is so.

4.2 The underlying values of the rule of law

4.2.1 Guidance

The rule of law requires that the law must be certain so that it can provide guidance.[44] Generally, laws that are as vague as general anti-avoidance rules attract considerable criticism because they fail to provide people with sufficient information about what is and is not permitted to allow them to plan their lives. For example, on 29 August 1935 the Senate of the Free City of Danzig decreed an amendment to the Danzig Penal Code that criminalised acts “deserving of penalty according to the fundamental conceptions of a penal law and sound popular feeling”.[45] In an uncomfortable common law echo, the House of Lords in the English case of Shaw v Director of Public Prosecutions decided that it had jurisdiction to create new offences in order to punish acts that were contrary to public morals, but that had not previously been held to be illegal.[46] The Danzig legislation, which was enacted in order to align the city’s criminal law with that of Nazi Germany, is sometimes known as the “Danzig Decree”. There appears to be no connection between it and the Danzig Decree, but Article 386 of the Criminal Code of the Qing Dynasty, which ruled China from 1644 to 1912, furnishes an interesting comparison. The Qing Code contained a long list of specific offences, but taking a form very similar to the decree of the Senate of Danzig, Article 386 provided that “[Doing] that which ought not to be done” was an offence. It is hard to think of a norm that claims to be a rule of law that could authorise more arbitrary action on the part of the authorities. Even the Nazi rule incorporated the (admittedly spurious) criterion of “sound popular feeling”.

Both the Danzig Decree and Shaw v Director of Public Prosecutions have been heavily criticised. For example, the Permanent Court of International Justice delivered an opinion condemning the amendment to the Danzig Penal Code.[47] People criticise Shaw for similar reasons.[48] Should we be concerned that the reasons that make the Danzig Decree, the decision in Shaw, and Article 386 of the Qing Code objectionable appear to apply equally to general anti-avoidance rules?

It is difficult to know what effect general anti-avoidance rules have on people’s actions. It has been suggested that they act in terrorem, in that people are discouraged from constructing tax avoidance schemes because of the risk of being caught by the general anti-avoidance rule.[49] While this consequence may be what governments hope for when they resort to general anti-avoidance rules, such an effect is not what scholars mean when they argue that the law should be capable of guiding people. However, to demonstrate that general anti-avoidance rules offend the rule of law it is not sufficient simply to show that they do not guide people’s actions. To see what is so objectionable about general anti-avoidance rules it is necessary to examine the underlying values of the rule of law, and to reveal why it is important that people should be able to rely on its principles to guide them.

4.2.2 Liberty

The relationship between liberty on one hand and laws that can be relied upon on the other is a key part in many theorists’ conceptions of the rule of law. For Rawls, people must know exactly what legal rights they can claim because, “If the bases of these claims are unsure, so are the boundaries of men’s liberties.”[50] An essential part of being free, then, is knowing exactly how free one is. This argument has particular resonance when we look at general anti-avoidance rules. The argument is that general anti-avoidance rules’ truly objectionable aspect is that no one really knows how far their reach extends. People are prevented from taking action that might be allowed, the argument continues, because they do not want to take the risk of their action being disallowed.

Hayek also stresses the connection between the rule of law and liberty, but his conception of liberty is slightly different from that of Rawls. Where Rawls would describe knowledge of the degree of liberty that the law allows as an essential component of liberty itself,[51] Hayek simply sees liberty as the absence of coercion. If people know what the law is in advance, they can choose to put themselves in the position of being subject to it. Subjection to the law is therefore a wilful act.[52] This argument is particularly relevant to general anti-avoidance rules. Since no one knows exactly when general anti-avoidance rules will apply, people who are caught by them have not made a conscious decision to be subject to them, and are therefore coerced.

The argument in the preceding paragraphs appears to support the proposition that general anti-avoidance rules offend the rule of law as Rawls and Hayek explain that doctrine. But when tax professionals make this argument they are likely to put it in more specific terms, namely, that the existence of a general anti-avoidance rule has a chilling effect on legitimate tax planning, and that fear of general anti-avoidance rules prevents investors and businesses from utilising effective business structures that appear to be economically sensible.

There may be some truth in this claim, but it is not borne out by reported cases. All cases known to the present writers where the Commissioner has attacked an arrangement using a general anti-avoidance rule involve schemes that an informed but objective bystander would predicate entail tax avoidance. From another perspective, at meetings of tax professionals one of the writers has frequently asked for examples of transactions or structures that could reasonably be predicated to be legitimate, but that taxpayers have rejected because of fear of a general anti-avoidance rule. Examples have not been forthcoming.

4.2.3 Human dignity

For Raz, the criterion that the law should be capable of guiding action is closely linked to human dignity. The law must assume that people are capable of rational thought, and that they therefore want to plan their lives with the knowledge of what the law is.[53] Raz sees this factor as even more important than the rule of law’s connection with freedom.[54] Laws that do not conform to the rule of law are an affront to human dignity because the law “encourages autonomous action only to frustrate its purpose.”[55] Raz might well charge general anti-avoidance rules with such an offence. The detailed formality of tax law encourages people to find ways to circumvent it, but general anti-avoidance rules may frustrate their efforts.

4.2.4 Effective law and Fuller

It is unlikely that Lon Fuller would disagree with Rawls’s argument that the rule of law protects liberty or Raz’s proposition that it protects dignity. Fuller, however, focuses his argument on the theory that certain formal criteria of the rule of law must all be sufficiently satisfied in order for law properly so called to exist.[56] Laws must be public, prospective, understandable, non-contradictory, possible to conform to, relatively stable, there must be congruence between how the rules are written down and how they are enforced, and laws must be rules as opposed to ad hoc decisions.[57]

In order to demonstrate how continuous breaches of the rule of law reduce the effectiveness of legal systems, Fuller gives us the example of King Rex. King Rex is a ruler who tries but fails to make law on eight separate occasions. Each time that Rex attempts to make law, he manages to breach one of these eight criteria. For example, on one occasion Rex publishes a legal code that is so convoluted that no one can understand it and on another occasion he announces that all cases will be decided retrospectively.

Naturally, Rex’s subjects are dismayed at their king’s disregard for the rule of law, and are annoyed at the way the consequences of that disregard affect them.[58] For present purposes, however, the interesting point is the consequence for Rex. Rex is unable to rule effectively because his rules are incapable of being followed. There is really no point in Rex having laws at all, because his laws do not guide the behaviour of his subjects.[59] However much his subjects might want to obey Rex’s laws, they cannot. Fuller’s examples show that laws that do not conform to the rule of law can therefore be just as frustrating to law-makers as they are to law-followers.

4.2.5 Are general anti-avoidance rules effective?

General anti-avoidance rules tend to be counter-examples to Fuller’s general theory of effective law. They are frustrating to the citizen, but they are useful to governments. When general anti-avoidance rules work they are undeniably effective, because they allow governments to collect tax that they would otherwise lose. Nevertheless, the experience of some countries with general anti-avoidance rules reveals that they can sometimes be ineffective for reasons very similar to those that plagued King Rex.

For example, when Sir Garfield Barwick was Chief Justice of Australia the Commissioner was seldom successful in litigation where he deployed the general anti-avoidance rule.[60] Barwick CJ felt very strongly that “[i]t is for Parliament to specify, … with unambiguous clarity, the circumstances which will attract an obligation on the part of the citizen to pay tax.”[61] The Chief Justice had little time for the vagueness of the general anti-avoidance rule, and tended to find for the taxpayer even in cases of the most blatant tax avoidance.[62]

Barwick CJ’s pro-taxpayer stance reached its apogee in the cases of Slutzkin v FCT[63] and Cridland v FCT.[64] Slutzkin was a case of dividend-stripping. The taxpayer was a shareholder in FR Holdings Pty Ltd, a company that was pregnant with profits. Had the company distributed the profits as a dividend they would then have been taxable in the hands of Slutzkin and his fellow shareholders. The same result would have obtained had the shareholders put the company into liquidation and distributed the proceeds.

Instead, the shareholders cashed the company up by liquidating its assets. They then sold their shares to Cadiz Corporation, which was a trader in shares. Cadiz Corporation caused FR Holdings Pty Limited to distribute its retained profits as a dividend. Without its retained profits the company was now worth very little. Cadiz Corporation sold the shares in FR Holdings Pty for much less than the price that it had paid for the shares. Thus, the purchase and sale of the shares of FR Holdings Pty Limited resulted in a loss to Cadiz, but the loss was deductible as Cadiz was a trader in shares. Losses, preferably occasional rather than chronic, are incidental to the ordinary business of traders.

For Slutzkin, the fiscal effect of these transactions was that he sold his shares for a non-taxable capital receipt. Cadiz Corporation Limited, on the other hand, derived a taxable profit from the dividend, but sustained a deductible loss in selling the shares. The loss neatly cancelled the gain from the dividend and left Cadiz Corporation Limited with, in effect, a fee for its trouble. The fee was taxable, but was a very small fraction of the income that Slutzkin and his fellow shareholders had stood to derive from either a profit distribution or a liquidation.

Arguing that the only reason that Slutzkin and his fellows sold their shares was to avoid tax on profits that would otherwise have been distributed, the Commissioner submitted that the price of the shares was economically the same thing as a dividend and that the general anti-avoidance rule applied. Barwick CJ rejected this argument, holding that the sale of the shares was “no more than a realisation by them of the benefit of their shareholding in a way which would not attract tax”.

Cridland involved a scheme designed to take advantage of a rule that allowed primary producers to average their incomes over a number of years and to pay tax on that average. The rule was intended to make the tax system fairer for people like farmers, whose income often varies considerably from one year to the next. (Where there is a progressive scale, people with variable incomes can find themselves in one year or another propelled unfairly into very high bands of tax, bands that do not reflect their average income calculated over several years.) The scheme relied on rules that made anyone with even a small amount of farming income a “primary producer”, and that allowed for averaging of all income derived by primary producers, including non-farming income. Subscribing to the scheme, Cridland, a university student, bought a share in a unit trust. The trust was a primary producer. Cridland’s interest as a beneficiary of the trust was only one dollar a year. The years in which he was a beneficiary straddled his time as a student and also time as a salaried graduate, when his income was much higher. Cridland claimed to be a primary producer and therefore to average his income, spreading much of it back into his impecunious years as a student. Despite the general anti-avoidance rule, the Barwick court upheld the claim, with Mason J delivering the leading judgment. Both Slutzkin and Cridland were almost certainly situations where Australia’s general anti-avoidance rule should have applied, but Barwick CJ’s High Court found in both cases that the taxpayers had not avoided tax.

In response to this judicial attitude, which rendered Australia’s general anti-avoidance rule almost useless, the Australian Parliament in 1981 enacted a new type of general anti-avoidance rule that attempts to attain more precision of detail. It is certainly more prolix.[65] In hindsight, Parliament’s action was possibly not necessary: following Sir Garfield Barwick’s retirement, the High Court was able to re-inject some force into s 260, Australia’s then general anti-avoidance rule.[66] The history of how s 260 fared during Sir Garfield Barwick’s term as Chief Justice is an interesting example of how the rule of law defects of general anti-avoidance rules can make them ineffective.[67]

It is interesting to note that when general anti-avoidance rules are ineffective, this ineffectiveness is not due primarily to taxpayers being inadequately guided. Rather, when general anti-avoidance rules are ineffective it is because the judiciary do not know what to make of them. To return to general anti-avoidance rules’ sinister counterpart, the amendment to the Danzig Penal Code, it seems that the Nazis had a similar experience to that of the Australians with Sir Garfield Barwick. The same rule applied in Germany, as well as in Danzig, but it ultimately led to very few prosecutions in either jurisdiction, because its terms were too vague for even the compliant judges of the Nazi era to make much sense of them.[68]

There appears to be a parallel with Article 386 of the Qing Dynasty Criminal Code. A penalty that was rather limited for the times mitigated the wide embrace of the language of the rule. The punishment for breach of Article 386 was caning with the banzi, 40 strokes of the light bamboo, or 80 strokes of the heavy bamboo for more serious offences:[69] harsh enough by our lights, though then thought to be on the lenient side.[70] For this reason, it was generally understood that the catch-all Article 386 was intended to apply only to relatively minor misdemeanours.[71] Knowledge of the operation of the criminal law under the Qings is limited, but it may not be drawing too long a bow to suggest that Article 386 is another demonstration of Fuller’s thesis. Uncertainty as to its coverage may have stunted the operation of what, on its face, was a rule that offered unlimited scope for oppression.

General anti-avoidance rules in the tax area furnish a marked contrast to rules like the Danzig Decree and the Qing Article 386: situations where statutory general anti-avoidance rules are ineffective are relative rarities. The majority of jurisdictions that have general anti-avoidance rules find them to be a reasonably effective though not foolproof tool for frustrating tax avoidance.[72]

It is difficult to know what conclusion to draw from the fact that general anti-avoidance rules tend to be relatively effective. Fuller’s argument that laws are more effective when people know what they require certainly seems uncontroversial and likely to be true in most situations in respect both of law in general and of law employed in economic regulation, which is the concern of the Law Foundation Regulatory Reform Project. While Fuller does not demand that legal systems must satisfy each of his criteria perfectly in order to conform to the rule of law;[73] it is unlikely that Fuller would approve of the protracted and unapologetic breaches that accompany general anti-avoidance rules.

This point is even clearer if we use Fuller’s framework to assess individual laws, as opposed to entire legal systems. A state with some laws that offend Fuller’s criteria may still be able to be governed effectively, but, according to Fuller’s thesis, an individual rule that continuously breaches many of his criteria ought not to be effective. It is an interesting feature of general anti-avoidance rules that their criteria for effectiveness are almost the exact opposite of the effectiveness criteria of other laws.

4.3 Are general anti-avoidance rules justified despite breaching the rule of law?

4.3.1 Problems of income taxation

The intuitive alternative to a general anti-avoidance rule is a system of very many specific rules that detail exactly what is and is not subject to income tax. Of course, all tax systems already have such specific rules in at least some areas of economic activity, whether or not they also have general anti-avoidance rules. Unfortunately however, the more specific and detailed a system’s rules become, the more ways people will find to circumvent those rules.[74] Tax law is unusual in two key respects. First, there are very few other areas of law that people so aggressively try to avoid.[75] Secondly, the nature of tax law means that tax legislation contains a large number of potential loopholes.[76] The result is that in the absence of a general anti-avoidance rule, there is apt to be a great deal of tax avoidance that the government is powerless to stop.

It is tempting to suggest that if legislators cannot frame a tax avoidance rule that conforms to the rule of law they should not have an anti-avoidance rule at all. Governments should just put up with the adverse consequences. This suggestion, however, overlooks the fact that tax avoidance is not a problem for governments alone; it is a problem for society generally. Avoidance undermines two key purposes of a tax system. First, the principle of horizontal equity states that people in the same economic position should be taxed at the same rate.[77] Tax avoidance makes horizontal equity difficult to achieve, because successful tax avoidance results in some people being taxed less than others who are in the same economic position. In other words, people who avoid tax are not paying their fair share as measured by their wealth.

Secondly, tax avoidance makes it more difficult for tax systems to be economically neutral. Economic neutrality demands that tax systems should distort the normal workings of the market as little as possible; that is, that people should not make decisions for purely (or even partially) tax reasons. The existence of opportunities for tax avoidance frustrates this goal. To illustrate, consider the case of Peterson v CIR,[78] which the Privy Council decided in 2005. Peterson was a case in which films were funded principally by non-recourse loans. Pursuant to a scheme, Mr Peterson and others invested in films and deducted their investment from their other income. The deductions took the form of allowances for depreciation, which permitted investors in films to amortise the cost over two years when calculating assessable income.

The promoters of the film told the investors that the cost of the film was (say) $2,000, while in fact it was only (say) $1,000. To fund their investment in the films, Mr Peterson and his co-investors borrowed. The borrowing was in the form of non-recourse loans, that is, loans that were repayable only if the films were successful. Interest was not charged. Loans on such favourable terms naturally attract questions, and indeed it was found as a fact that the money was never borrowed at all.[79] The fact that the extra money from investors was not available did not bother the film’s promoters, because they had overstated the cost of the film anyway.

The reason for the overstatement of the cost of the films was the tax saving that it led to. Instead of being able to write off $1,000 over two years, investors were able to write off $2,000, even though they had never actually spent the second $1,000 (and, except on paper, had not even borrowed it). Whether or not the films were successful, the investors would gain a tax advantage. This tax advantage meant that a scheme that would not ordinarily be attractive to investors became worthwhile.

This situation is a clear example of the tax system creating market distortions: the transactions in Peterson were not attractive for their intrinsic merits; they were attractive because of tax advantages. Jurisdictions that have general anti-avoidance rules are able to counteract the effect of this distortion to some extent: to the extent that investors see the tax advantages of a particular scheme as unlikely to stand up to close scrutiny and therefore refrain from investing in it and the market will not be distorted.

The aims of the tax system are related to the wider purposes of taxation in general. Governments do not tax people only to amass wealth. Rather, tax is necessary to keep states functioning. Governments must provide public services such as defence and education. Furthermore, most societies use tax to redistribute wealth to some extent. Tax avoidance reduces the effectiveness of welfare systems,[80] a matter that is particularly important in the light of the public perception (that is probably accurate) that most tax avoidance is perpetrated by the rich or by people who are relatively well off. Though few people have reasoned the issue through to a sufficient depth to put it this way, the wide spread of general anti-avoidance rules, either statutory or judge-made, indicates that countries may consider that the negative results from not having a general anti-avoidance rule outweigh the breaches of the rule of law that general anti-avoidance rules entail.

This balancing exercise reveals much about the nature of the rule of law and its values. Adherence to the rule of law can often interfere with a society’s other goals. Some philosophers insist that the rule of law must be preserved without compromise.[81] Other writers, such as Raz, stress than the rule of law is only one yardstick against which a legal system may be measured.[82] Just as a society’s conformity to the rule of law does not ensure that the society is good, a breach of the rule of law does not make that society bad.[83] Rawls expands on this point, saying that a breach of the rule of law may be “the lesser of two evils”.[84] Tax avoidance is a very real evil for society: a breach of the rule of law seems to be a necessary remedy.

In modern days, at least in democracies that follow a Western model, there is seldom anything sinister about legislators breaching the rule of law. As Fuller observes, laws tend to be most effective when they conform to the rule of law;[85] so governments have a vested interest in making sure their laws conform to its values. In situations where laws offend the rule of law, it will often be the case that the alternative is even less desirable. Tax law is by no means the only situation where the rule of law must be sacrificed to the common good. It is easy to imagine situations where the preservation of human rights or the fulfilment of justice requires a breach of the rule of law. Take, for example, a law that at first sight may not appear to breach the principles of the rule of law at all: a retrospective law may be necessary to compensate fully for a human rights breach. The breach here is not a breach as far as the compensated victim is concerned. She is better off as a result of the retrospective law change. The breach is a breach of the interests of all other taxpayers, who did not expect their taxes to be charged with this particular impost. Nevertheless, in such situations most people would think that a breach of the rule of law by retrospective legislation is justified. Others, of a less analytical frame of mind, might not recognise a breach of the principle of non-retrospectively at all. Erroneously thinking of the fiscal position of the state as somehow independent of the fiscal position of its taxpayers and citizens, they might simply experience an imprecise feeling of warm satisfaction that the state had done something to remedy the injustice.

4.3.2 The importance of certainty

Certainty is clearly an important rule of law value. Usually certainty is important for both the law-follower and the law-maker. Most laws are more effective when people can be certain what they are meant to do or not do. That is, in most cases the rule of law helps to promote effective law. General anti-avoidance rules are therefore an aberration: it is their very vagueness that makes them effective. If they were not vague, they would not be effective.[86] This characteristic, together with the fundamental problems of tax law together with what many see as the dubious moral standing of tax avoiders, prompts some commentators to argue that certainty is simply an inappropriate value for general anti-avoidance rules to strive for.[87]

Challenge Corporation Ltd v Commissioner of Inland Revenue[88] is an example of the negative effect that certainty can have on the utility of an anti-avoidance rule. Challenge Corporation, the taxpayer company, acquired a subsidiary that had suffered heavy losses. Challenge Corporation then purported to set the subsidiary’s losses off against its own profits.

At the time, the provisions that allowed intra-group loss consolidation required only that loss-consolidating companies had to meet a minimum threshold of common ownership as at the end of the tax year. Through what seems to have been a drafting shortcoming, there was no requirement for a loss company to have suffered its losses at a time when it was owned by any of the same interests as owned the profitable company. Challenge Corporation had therefore complied with the letter of the law. There was a specific anti-avoidance rule, but it did not apply where one company simply acquired and retained ownership of the shares of another, as happened in the Challenge case. Without a general anti-avoidance rule, companies in the situation of Challenge Corporation would be able to take deductions despite having suffered no economic loss.

Where the principles of the rule of law negatively influence a law’s effectiveness, it is necessary to weigh the consequences of not having the law in question against the possibility that some people will be surprised by the manner in which the law operates. Certainty and related rule of law values are therefore extremely important where criminal sanctions are imposed, but are less important where the issue is tax avoidance.[89]

4.3.3 The morality of tax avoidance

In the face of such an obvious breach of the rule of law, the fact that so many countries have general anti-avoidance rules seems difficult to account for. The idiosyncrasies of tax law no doubt make general anti-avoidance rules necessary, but it is unlikely that the public tolerance of general anti-avoidance rules is caused by knowledge of these idiosyncrasies. Tax law is extraordinarily complicated, but it is unrealistic to suppose that most people see tax law as different in kind from other branches of the law. How, then, can we account for the lack of public condemnation of general anti-avoidance rules? The explanation may be a perception of tax avoidance as being questionable from a moral perspective.

The moral status of tax avoidance is contentious. There have been a number of cases that hold that since people have the right to arrange their money in such a way as to pay as little tax as possible, even holding that there is nothing immoral about tax avoidance.[90] Relying on such decisions, lawyers tend to assume that as a matter of law tax avoidance is morally unimpeachable. To say that because tax avoidance is not immoral as a matter of law it is not immoral in any sense entails two errors of logic. First, whether a certain act is moral must be determined according to principles of ethics, not by reference to statements in judgments. It is possible that judges who say that there is nothing immoral about tax avoidance are correct, but if that is so it must be because tax avoidance is moral according to ethical principles. As a matter of logic, a judge saying that a particular act is moral as a matter of law cannot determine whether the act is in fact moral.[91] Secondly, assume, contrary to the argument just adumbrated, that as a logical proposition a law-maker or judge can pronounce that a particular act is moral or immoral as a matter of law. Such a pronouncement still leaves open the issue of whether the act in question is moral as a matter of ethics, which brings us back to the starting point, namely, the question of whether tax avoidance is moral cannot be determined for purposes of morality by a legal rule or decision.

What is the moral status of tax avoidance according to basic principles of ethics? As a matter of morality untainted by law, people know that they have a duty to pay tax; so seeking to pay less tax than they otherwise might can appear to be shirking that duty.[92] Furthermore, despite the complexity of tax laws, most people have a reasonably clear idea of what the policy of the law would require them to pay. General anti-avoidance rules do not set out to catch individual taxpayers trying earnestly to comply with complex tax laws. Rather, they tend to catch instances of tax planning that is at least relatively aggressive. People who are ultimately caught by general anti-avoidance rules almost always know that they have engaged in something that they would at least concede to be “tax planning” – usually aggressive tax planning - even if they do not expect to be called to account. Taxpayers who engage in tax avoidance schemes are consciously putting other taxpayers at a relative disadvantage and may be criticised on moral grounds.[93]

If the arguably dubious moral status of tax avoidance partially explains the conspicuous lack of public outcry over general anti-avoidance rules, what can we deduce about the relationship between the rule of law and morality? It cannot be correct that people lose their right to rely on the law when they act immorally.[94] No one would suggest that the rule of law is unnecessary in the field of criminal law, which typically involves far more obvious immorality than tax avoidance. Possibly the real explanation is that the rule of law itself, as a strict formalist doctrine, inevitably allows people to some extent to circumvent the laws that conform to it. As far as criminal law is concerned, this shortcoming of the rule of law is far outweighed by the benefits that the rule of law offers. In contrast, when it comes to tax avoidance, the benefits to society of legal certainty are outweighed by its detriments.

The argument that the detriments of the rule of law in a particular area outweigh its benefits is nevertheless unsatisfactory. At least, it would not satisfy Hayek, although it might satisfy Rawls or Raz. Hayek would argue that the merits of the rule of law should not be evaluated on a case-by-case basis, leaving us free to disregard its principles where those principles are inconvenient. Rather, one of the reasons why societies value the rule of law is that it applies despite its resulting in a net societal detriment from time to time. Societies commit to adherence to the rule of law for the very reason that there will be instances where it is tempting to tolerate breaches.

This argument echoes David Cole’s criticism of Richard Posner’s Not a Suicide Pact: The Constitution in a Time of National Emergency.[95] In his book, Posner argues that the protections offered by the United States Constitution should be interpreted flexibly in order to allow the government to address the threat of terrorism. Posner argues, for example, that the United States Administration’s wiretapping of international telephone calls should be considered a “reasonable” search in the context of the threat of terrorism. Cole, however, points out that allowing the provisions of the Constitution to be interpreted more strictly or less strictly according to administrative convenience misses the point of having a constitution in the first place.[96] A constitution like that of the United States, and the rule of law, should be adhered to notwithstanding that doing so is not beneficial to society in every case. Any kind of cost-benefit analysis is simply inappropriate where the Constitution is concerned. The same considerations apply in respect of the rule of law.

It would follow from the principles advanced by Hayek and by Cole that the disbenefit to society at large that can accompany adherence to the rule of law when it is a matter of tax avoidance does not seem to explain the apparent acceptance of general anti-avoidance rules even among well-informed sectors of the public. Nor would that disbenefit justify the breach of the principle of the rule of law entailed in the uncertainty of general anti-avoidance rules. What, then, may be the explanation and the justification?

In respect of the public acceptance of general anti-avoidance rules, tax avoiders appear to be, and are seen as, fundamentally different from criminals. Generally speaking, when criminals break the law, they simply break it; they do not try to find ways to circumvent the law in order to avoid technical breaches. In contrast, there is an entire industry devoted to manipulating fiscal laws with a view to obtaining tax advantages without incurring a corresponding economic cost. In the light of this difference, the fact that the informed public appears to accept general anti-avoidance rules[97] despite their shortcomings as far as the rule of law is concerned is not surprising.

In respect of the justification for the breach of the rule of law, unlike criminal behaviour, tax avoidance takes advantage of the very nature of law itself. In particular, it takes advantage of law’s adherence to formality. The formality of law in general and of tax law in particular is an essential pre-requisite for contriving artificial transactions that enable the authors of the transactions or their clients to avoid tax. These are transactions that shift income from higher taxed people to lower taxed people, that enable revenue-to-capital conversions, that achieve the deferral of receipts or the acceleration of expenditure, that, through international arbitrage, permit the recharacterisation of receipts or expenditure, and so on.

The quality of relying on the formality of the law while circumventing the law’s policy distinguishes tax avoidance from criminal behaviour, being the area where rule of law questions tend to be most prominent. While it is true that there are difficult cases at the edges of criminal law (assisted suicide of very sick people and certain practices in cultures other than our own being prime examples) most criminal activity is clearly wrong by the lights of most people, whether or not there is law to forbid it. In contrast, tax avoidance exploits the formality of the law and, in doing so, exploits the values of the rule of law itself. It attacks those values while pretending to honour them. Enacting a general anti-avoidance rule to frustrate that exploitation presents as a justifiable counter-measure.

4.4 Conclusion as to general anti-avoidance rules

General anti-avoidance rules demonstrate that the rule of law is not an unqualified good. As with all principles, the rule of law can be outweighed by competing considerations. General anti-avoidance rules give an example of what those competing considerations might be. Furthermore, while general anti-avoidance rules themselves are justified, they are useful in showing exactly why we value the rule of law. Most societies with developed legal systems tend not to breach the rule of law very often. As a rare example of a breach, general anti-avoidance rules are a useful reminder of why values such as certainty are important.

4.5 Taxation-related issues relevant to the New Zealand questions in the Law Foundation Regulatory Reform Project

4.5.1 Introduction

As explained in the introduction to this chapter, there are a number of issues within the general area of the administration and regulation of taxation that give rise to rule of law questions. This final section of the taxation issues chapter lists and describes these issues. The rule of law questions discussed in the foregoing analysis of the chapter’s major example, namely general anti-avoidance rules, known as “gaars“, are relevant mutatis mutandis to the issues that are to be described below. The first five issues relate to attempts to create more certainty in the operation of gaars.

4.5.2 Interpretation and application of the general anti-avoidance rule

Courts always have difficulty with general anti-avoidance rules. One result is that judgments are often inconsistent with precedents, although purporting to follow those precedents. Another is that judgments are often inconsistent within themselves. If these kinds of issues could be resolved there would be at least a better chance of effective and uniform application of the gaar. A full study of the jurisprudence of the New Zealand gaar is beyond the scope of the Regulatory Reform Project, but a study of five or six issues and lines of reasoning that are currently to be causing difficulty would be of considerable utility.

4.5.3 An enumerated general anti-avoidance rule

For decades leading to the 1970s, the Australian statutory general anti-avoidance rule (Income Assessment Act 1933, s 260) was couched in very similar terms to New Zealand’s current rule. As this chapter has explained, while Sir Garfield Barwick was Chief Justice of Australia the High Court in effect eviscerated s 260. In response, the Federal Parliament passed Part IVA of the Act. The effect of Part IVA is much the same as the effect of the New Zealand s BG 1. The main difference is of appearance, in that Part IVA is much longer. Part IVA essentially codifies lists of factors that courts have in the past taken into account to determine whether s BG 1 or its predecessors apply to an impugned arrangement. It also adds provisions that policymakers thought desirable.

The main reason for the enactment of Part IVA was to provide a more fine-grained analytical framework against which courts could evaluate arrangements that the Commissioner impugns, in order to return some force to the section. A by-product is that, at least in appearance, the greater detail of the section affords greater certainty to taxpayers, though it is debatable whether there is greater certainty in fact or, if there is, whether that development is desirable. An issue for consideration is whether New Zealand should move to an enumerated rule. Within that issue there are a number of sub-issues. Probably the most significant such question is whether an enumerated rule should be inclusive or exhaustive.

4.5.4 GAARS and double tax agreements

Double tax agreements operate by overriding parts of the domestic tax laws of the treaty partners. Internationally, it is a live issue whether a double tax agreement overrides a domestic gaar. The underlying intention of state parties is almost certainly that treaties should not override gaars. One of the intentions of most treaties is to frustrate avoidance. If a treaty overrides a gaar it could well do the opposite. Some countries have addressed this problem by providing in their income tax legislation that treaties do not override gaars. Should New Zealand follow suit?

4.5.5 Interpretation statement from Commissioner

Since 1974, when the general anti-avoidance rule took what is substantially its current form, the Commissioner of Inland Revenue has on several occasions published statements that contain an exegesis of his view of the effect of the section. These statements do not purport to be law, but they have more or less the effect of law, since it is likely that the Commissioner will follow his interpretation statements in applying the law.

The reason for interpretation statements is similar to the reason for an enumerated rule. That is, it is thought that greater detail will offer greater certainty to taxpayers.

The history of interpretation statements is decidedly mixed. Some examples given in past statements have been wrong, or at least arguably so. Other examples or passages in interpretation statements have led taxpayers to persuade themselves of points of law later found to be untenable or given taxpayers a foundation to argue for decidedly doubtful interpretations.

Another question is whether the Commissioner should issue interpretation statements at all. Parliament has declined numerous opportunities to flesh out the general anti-avoidance rule. Is it appropriate for the Commissioner in effect to legislate where Parliament has decided not to tread? An issue, therefore, is whether the Commissioner should issue interpretation statements in respect of s BG 1. If one of the strengths of the statutory rule is its indeterminacy, it would be strange for the Commissioner to dilute that strength by attempting to make the rule more detailed.

If, contrary to the burden of the preceding paragraphs, the government should issue guidance on the operation of the gaar, further questions arise: Should the guidance be by regulation, ruling, or interpretation statement? And whatever the form, should the guidance include examples?

4.5.6 Private binding rulings on avoidance issues

By statutory authority, the Commissioner of Inland Revenue provides rulings as to the tax effect of proposed transactions. These rulings are binding on the Commissioner even if they are later found to be mistaken. Tax authorities in a number of countries issue such rulings, the conditions of issue varying somewhat from one country to another. In the present context, a particular variation is significant: some jurisdictions will not issue rulings on questions of avoidance.

In New Zealand, most, if not all, rulings on avoidance have taken the form of private rulings. That is, they are not published. There is, however, enough information in the public domain to suggest that some private rulings have been questionable. An issue is whether the New Zealand rulings regime should be modified to provide that rulings will not be available on avoidance questions.

4.5.7 Publicity of rulings

The original proposal for binding rulings was that rulings should all be published, with names and identifying details of taxpayers deleted. Provision for publication was omitted from the regime that was enacted. The result is that there is now a large body of quasi-law that is secret. Though this quasi-law is secret as far as the general body of taxpayers is concerned, parts of it become known to experts in the area, leading to a two-tier profession of tax advisers: an in-group and an out-group. An issue is whether publication of rulings should be revisited.

4.5.8 Fees for rulings

The rulings process appears to be subject to considerable delay. Should fees for rulings rise, in order to provide more resources?

4.5.9 Publicity about details of avoidance

Despite the best efforts of revenue authorities worldwide, major avoidance by large companies remains pervasive and may even be increasing. Some egregious avoidance is evasion under another name. Take, as an example, the “HOMER” prosecutions in the United States of 2010 and 2011.[98] In the past, New Zealand published names of delinquent taxpayers in The Gazette and in the press. Modern cases are formally different from the cases of such taxpayers, but share a good deal of substance.

Occasionally, litigation brings such avoidance to public view. It may be, however, that when avoidance is detected the more usual result is an eventual settlement that is not publicised. There is nothing new about secrecy in tax cases in general or in respect of settlements in particular, but developments in the United Kingdom over the Vodafone and Goldman Sachs settlements of 2010–2011 call such secret settlements into question. Studies of Vodafone’s published accounts have suggested that the Vodafone settlement was more generous than informed opinion could have expected, to the extent of raising the question of whether there was an element of agency capture. There may be an issue as to whether there should be more publicity about detected avoidance that does not reach the courts.

4.5.10 Transfer pricing

For tax administrations, transfer pricing is probably today’s major issue in the context of international trade and investment. Because of its heavy reliance on trade and on foreign capital, transfer pricing is very significant for New Zealand. It has become increasingly apparent that transfer pricing is an area of law that contains almost no norms that are recognisable as law. With the publication of the OECD’s Transfer Pricing Guidelines (2010) the position is transparently clear: here is a law book with virtually no law, reflecting a major area that claims to be law but that, likewise, has almost no law within it. This situation gives rise to pervasive rule of law issues. How are tax administrations to regulate and police intra-group transfer prices when there is no underlying law? Resort to the arm’s length principle, particularly in respect of goods that are never traded at arm’s length, takes us only so far. An issue for the Regulatory Reform Project is whether it is possible to reduce the elusive principles of transfer pricing to laws and, if so, whether that enterprise would be practical, bearing in mind that other nations are involved in any transfer pricing problem.

4.5.11 Responsive regulation

One model of the practice of responsive regulation that is beginning to be employed in the tax area is for the tax administration to employ a much lighter hand in respect of taxpayers that have demonstrated trustworthiness. Broadly speaking, in this context “trustworthiness” implies compliance with the law. A problem arises in that, in a strict sense, an avoidance transaction does comply with the law; that is the point of avoidance. On the other hand, it would defeat the purpose of responsive regulation in the tax area if a taxpayer could be deemed “trustworthy” by virtue of never evading tax and always complying with formal precision, if the same time the taxpayer engaged in avoidance transactions. At the time of writing, mid-2011, these kinds of questions are live issues in the United Kingdom. An issue for the Project is the extent to which responsive regulation is appropriate in an area like taxation, where the law is inherently uncertain.

4.5.12 The tax system and social programmes

Increasingly, New Zealand has used the Inland Revenue Department to administer and to deliver social programmes. Examples include the liable parents and student loans schemes. In some cases, New Zealand has gone further, and built social programmes into the tax system. Family support is an example. Undoubtedly, the Department has been able to administer these schemes more efficiently than other departments. But the system gives rise to a number of questions. Among them are, does the tax administration become less efficient as a result of giving the tax administration these extra roles? And, particularly relevant for the Regulatory Reform Project, what increased level of data sharing is appropriate to promote efficiency?

The questions just posed are major issues and likely to be beyond the scope of the Law Foundation Regulatory Reform Project. They are mentioned here to indicate that they have not been overlooked.

4.5.13 Excise and similar imposts on activities thought to be socially questionable

New Zealand imposes “sin” taxes on tobacco and alcohol.[99] It is sometimes suggested that we should impose a carbon tax. We tax liquid fuel apparently because it is an easy target. All these sorts of taxes give rise to issues under the New Zealand questions of the Regulatory Reform Project, especially in relation to the market economy. The broadest of these questions asks whether it is appropriate and/or efficient to use tax as market-based regulatory intervention.

4.5.14 The tax system and economic goals

One potentially significant issue that, at this stage, it is not proposed to address is the question of the merits and demerits of using preferences within the tax system to promote economic activity that the government favours. This issue falls squarely within the first two New Zealand questions of the Law Foundation Regulatory Reform Project, namely the question of small size and issues of the market economy and dependence on international trade. Tax preferences were a significant feature of the New Zealand tax system until the major reforms of the 1980s. Among the most notorious were measures that, along with subsidies, encouraged farmers to break uneconomic land in for pasture. So long as their fellow taxpayers shouldered the cost, farmers could profit from such land. But take all costs into account and the overall result was a continuing loss that was a major incubus on the shoulder of the country’s economy.

The reforms of the 1980s did away with virtually all preferences, leaving New Zealand with one of the world’s least adulterated tax systems. It is hard to find an economist or tax specialist who disagrees with the fiscal cleansing that occurred. During the nine years from 2001 a number of preferences re-entered the system, largely, apparently, for political reasons. Now that these preferences are present, the current government is finding it hard to bring itself to expel them. Nevertheless, the reasons apparently continue to be political rather than a matter sound and thought-out economic management. The reasons against tax preferences are well known and command general agreement among experts. There seems little point in going over them again in the context of the Regulatory Reform Project.

4.5.15 Conclusion

The 13 issues or sets of issues identified in this section of the chapter constitute a major work plan, far beyond the resources of the Law Foundation Regulatory Reform Project. Nevertheless, it is valuable to identify these issues (which, even now, are probably not an exhaustive list) in order to present a picture of the issues that potentially arise in an income tax context that fall within the New Zealand questions that the Project addresses. It is proposed that the project will not explore all, or even most, of these issues.

As for the Regulatory Reform Project itself, in addition to the bulk of the work in the present chapter (that is, at [4.1] to [4.4]) aspects of the issues described at [4.5.4], [4.5.5] and [4.5.6] will be addressed in the remaining stages of the project.


[1] The authors gratefully acknowledge the support of the New Zealand Law Foundation and the Henry Lang Fellowship, Institute of Policy Studies, Wellington, towards the writing of this paper.

* BA (Hons), LLB (Hons) Victoria University of Wellington, LLM Columbia; Barrister and Solicitor (NZ); Analyst the New Zealand Treasury.

§ BA, LLB (Hons) Auckland, BCL Oxon, JSD Cornell; Inner Temple, Barrister; Professor and former Dean of Law, Victoria University of Wellington; Senior Fellow, Taxation Law and Policy Research Institute, Monash University, Melbourne; former Henry Lang Fellow, Institute of Policy Studies, Wellington.

[2] The Regulatory Standards Bill 2011 (277-1), cl 17(1)(a) provides as a principle of responsible regulation that laws should be clear and accessible.

[3] Albert V Dicey Introduction to the Study of the Law of the Constitution (10th ed, Macmillan, London, 1960) at 202.

[4] Similar problems arise in the area of competition law see in this volume Paul G Scott “Competition Law and Policy”.(ch 3)

[5] See in this volume Richard Boast and Neil Quigley “Regulatory Reform and Property Rights in New Zealand” (ch 5) and Russell Brown “Possibilities and Pitfalls of Comparative Analysis of Property Rights Protections, and the Canadian Regime of Legal Protection against Takings” (ch 6).

[6] See cl 7(1)(d) of the Regulatory Standards Bill 2011 (277-1), which restates the rule found in s 22 of the Constitution Act 1986 that requires taxes to be imposed by statute.

[7] See Richard Boast and Neil Quigley “Regulatory Reform and Property Rights in New Zealand” in this volume (ch 5).

[8] Judith Freedman “Defining Taxpayer Responsibility: In Support of a General Anti-Avoidance Principle [2004] BTR 332 at 350.

[9] See, for example, Miller v CIR [2001] 3 NZLR 316 (PC) at 326 per Lord Hoffman.

[10] Newton v FCT (1958) 98 CLR 1 at 8, [1958] 2 All ER 759 (PC) at 764.

[11] FCT v Gulland (1985) 160 CLR 55 at 109 per Dawson J; FCT v Spotless Services Ltd (1996) 186 CLR 404 at 425 per McHugh J.

[12] Mangin v CIR [1971] NZLR 591 (PC) at 596–598 per Lord Donovan, quoting Turner J in the court below.

[13] Challenge Corporation Ltd v CIR [1986] 2 NZLR 513 (PC) at 561 per Lord Templeman.

[14] See, generally, Nabil Orow General Anti-Avoidance Rules: A Comparative International Analysis (Jordan Publishing Limited, Bristol, 2000) at 18.

[15] For example Challenge Corporation Ltd v CIR [1986] 2 NZLR 513 (PC) involved a corporate group taking advantage of rules that allowed it to consolidate the affairs of its members and to pay tax only on the resulting net profit. The group tried to minimise tax by buying an outside company that had suffered a loss and subtracting that loss from the profits of the original group.

[16] For example Mangin v CIR [1971] NZLR 591 (PC) involved an arrangement whereby the taxpayer each year leased the profitable part of his farm, which was a different section each year, to a family trust. The trust would then pay out the income from the section of the land to its beneficiaries, who were the taxpayer’s wife and children. The artificial element in this arrangement was that the part of the farm leased to the trust changed year by year, with the trust always receiving almost all of the farm’s income for that year. The result of the arrangement was that each beneficiary received a fraction of the farm’s income. The income was therefore taxed at a lower rate than it would have been had it been entirely derived by the taxpayer.

[17] Inland Revenue Commissioners v Bowater Property Developments Ltd [1989] AC 398 (HL).

[18] New Zealand, for example, requires companies to have a minimum continuity of ownership of 49 per cent between loss year and profit year, Income Tax Act 2007, s IA 5(2).

[19] See, for example, Income Tax Act 2007, s GB 1.

[20] Income Tax Act 2007, s BG 1, incorporating s GB 1 and certain definitions in s YA 1.

[21] Abgabenordnung (German Federal Code of Tax Procedure), s 42.

[22] Income Tax Act 1988 (Can), s 245.

[23] Income Tax Act 1962 (SA), s 103.

[24] Inland Revenue Ordinance 1997 (HK), s 61.

[25] Livre de Procédure Fiscale (Tax Procedures Code), art L 64.

[26] Income Tax Assessment Act 1936 (Cth), s 260.

[27] Income Tax Assessment Act 1936 (Cth), Part IV A and ss 177A–177G. See further below, text accompanying n 66.

[28] W T Ramsay Ltd v IRC [1982] AC 300 (HL).

[29] See, for example, Judith Freedman “Defining Taxpayer Responsibility: In Support of a General Anti-Avoidance Principle” [2004] BTR 332.

[30] Mayes v Revenue and Customs Commissioners [2011] EWCA Civ 407 at [100].

[31] Proposals to introduce a statutory general anti-avoidance rule to the United States have come before the United States House of Representatives on a number of occasions. See, for example, Rep Lloyd Doggett “Abusive Tax Shelter Shutdown and Taxpayer Accountability Act of 2003” HR 1555, 108th Cong, 1st sess 2 April 2003).

[32] Gregory v Helvering 293 US 465 (1935).

[33] The Health Care and Education Reconciliation Act of 2010 added new Code § 7701(o), 11 Pub L No 111-152, § 1409, 124 Stat 1029 (2010).

[34] Nabil Orow General Anti-Avoidance Rules: A Comparative International Analysis (Jordan Publishing Ltd, Bristol, 2000) at 373.

[35] John Prebble and Zoë Prebble “Comparing the General Anti-Avoidance Rule of Income Tax Law with the Civil Law Doctrine of Abuse of Law” (2008) 4 Bulletin for International Taxation at 151.

[36] See further, Jeremy Waldron “Is the Rule of Law an Essentially Contested Concept (in Florida)?” (2002) 21 Law and Philosophy 137.

[37] There is abundant support for this proposition. In particular, see Friedrich A Hayek The Constitution of Liberty (Routledge, London, 1960) at 144; John Rawls A Theory of Justice (The Belknap Press of Harvard University Press, Cambridge (Mass), 1971) at 235.

[38] For example, Income Tax Act 2007, s GA 1.

[39] Challenge Corporation Ltd v CIR [1986] 2 NZLR 513 (PC) at 562.

[40] Challenge Corporation Ltd v CIR [1986] 2 NZLR 513 (PC) at 562.

[41] See, for example, Richardson J in Challenge Corporation Ltd v CIR [1986] 2 NZLR 513 (CA) at 546 alludes to the somewhat paradoxical consequence situation of a literal interpretation of a general anti-avoidance rule being quite obviously not what Parliament intended.

[42] For example, the conference The Rule of Law and Anti-Avoidance Rules: Tax Administration in a Constitutional Democracy was convened in Sydney in 1995 to explore the apparent tension between anti-avoidance rules and the rule of law. Papers from the conference are collected in Graeme Cooper (ed) Tax Avoidance and the Rule of Law (IBFD Publications BV, Amsterdam, 1997).

[43] Herbert LA Hart “Positivism and the Separation of Law and Morals” in Herbert LA Hart Essays in Jurisprudence and Philosophy (Clarendon Press, Oxford, 1983) 49 at 63.

[44] See, for example, Joseph Raz “The Rule of Law and its Virtue” in Joseph Raz The Authority of Law: Essays on Law and Morality (Oxford University Press, Oxford, 1979) 210 at 213.

[45] Decree of the Senate of the Free City of Danzig (1935) art 2.

[46] Shaw v Director of Public Prosecutions [1962] AC 220 (HL).

[47] Consistency of Certian Danzig Legislative Decrees with the Constitution of the Free City (Advisory Opinion) [1935] PCIJ (series A / B) No 65.

[48] For example CC Turpin “Criminal Law – Conspiracy to Corrupt Public Morals” [1961] 19 CLJ 144 at 144–146.

[49] See, for example, Michael O’Grady “Acceptable Limits of Tax Planning: A Revenue Perspective” (paper presented to KPMG Tax Conference, Ireland, November 2003) at 6.

[50] John Rawls A Theory of Justice (The Belknap Press of Harvard University Press, Cambridge (Mass), 1971) at 235.

[51] John Rawls A Theory of Justice (The Belknap Press of Harvard University Press, Cambridge (Mass), 1971) at 235.

[52] Friedrich A Hayek The Constitution of Liberty (Routledge, London, 1960) at 144–145.

[53] Joseph Raz “The Rule of Law and its Virtue” in Joseph Raz The Authority of Law: Essays on Law and Morality (Oxford University Press, Oxford, 1979) 210 at 221.

[54] Joseph Raz “The Rule of Law and its Virtue” in Joseph Raz The Authority of Law: Essays on Law and Morality (Oxford University Press, Oxford, 1979) 210 at 221.

[55] Joseph Raz “The Rule of Law and its Virtue” in Joseph Raz The Authority of Law: Essays on Law and Morality (Oxford University Press, Oxford, 1979) 210 at 222.

[56] Lon Fuller The Morality of Law (2nd ed, Yale University Press, New Haven (CT), 1964) at 168.

[57] Lon Fuller The Morality of Law (2nd ed, Yale University Press, New Haven (CT), 1964) at 168.

[58] Lon Fuller The Morality of Law (2nd ed, Yale University Press, New Haven (CT), 1964) at 167.

[59] Lon Fuller The Morality of Law (2nd ed, Yale University Press, New Haven (CT), 1964) at 168.

[60] At the time, Australia’s general anti-avoidance rule was contained in s 260 of the Income Tax Assessment Act (Cth) 1936.

[61] FCT v Westraders Pty Ltd (1980) 54 ALJR 406 (HCA) at 461 per Barwick CJ.

[62] See further Geoffrey Lehman “The income tax judgements of Sir Garfield Barwick: A study in the failure of the new legalism” (1983) 9 Monash U LR 115 at 135. Lehman argues Sir Garfield Barwick did not deprive Australia’s general anti-avoidance rule of all effect, because to do so would invite speedy law reform. Barwick CJ allowed the former s 260 to continue to operate where an “antecedent transaction” was involved: see Mullins v FCT (1976) 135 CLR 290 at 302.

[63] Slutzkin v FCT (1977) 140 CLR 314.

[64] Cridland v FCT (1977) 140 CLR 330.

[65] Income Tax Assessment Act 1936 (Cth), ss 177A–177G.

[66] See FCT v Gulland (1985) 160 CLR 55.

[67] The United Kingdom’s experience with a judicially developed anti-avoidance doctrine might be used to illustrate the same point. The doctrine, as developed from its original formulation by Lord Wilberforce in W T Ramsay Ltd v IRC [1982] AC 300 (HL) at 323–326, is so vague that no one seems to be certain whether it even exists. Its application can therefore appear somewhat haphazard; see further Lord Walker of Gestingthorpe “Ramsay 25 Years On” (2004) 120 LQR 412.

[68] Martin Broszat The Hitler State (Longman, London, 1981) at 339.

[69] Charles Aylmer “Review of William C Jones, The Great Qing Code; a New Translation” (1994) 7 Bulletin of the European Association of Sinological Librarians, available online (last accessed 2 September 2011).

[70] See, for example, Jérome Bourgon “Abolishing ‘Cruel Punishments’: A Reappraisal of the Chinese Roots and Long-term Efficiency of the Xinzheng Legal Reforms” (2003) 37 Modern Asian Studies at 851–862.

[71] Several scholars of Chinese law have confirmed this point to the authors.

[72] Examples of cases where a general anti-avoidance rule has misfired include Slutzkin v FCT (1977) 140 CLR 314; Cridland v FCT (1977) 140 CLR 330, discussed earlier in this paper; and Peterson v CIR [2005] UKPC 5, to be discussed below.

[73] Lon Fuller The Morality of Law (2nd ed, Yale University Press, New Haven (CT), 1964) at 170.

[74] Judith Freedman “Defining Taxpayer Responsibility: In Support of a General Anti-Avoidance Principle [2004] BTR 332 at 346.

[75] But see, for some examples of such areas of law, Victor Fleischer “Regulatory Arbitrage” (2010) U of Colorado Law Legal Studies Research No 10–11.

[76] For an explanation of why tax law is more susceptible to loopholes than other areas of law, see Ross Parsons “Income Taxation – An Institution in Decay” (1986) 3 Australian Tax Forum 233; as developed in John Prebble “Income Taxation: a Structure Built on Sand” (2002) 24 Sydney Law Review 303.

[77] See, for example, Richard E Krever “Structure and Policy of Australian Income Taxation” in Richard E Krever (ed) Australian Taxation: Principles and Practice (Longman Cheshire Pty Ltd, Melbourne, 1987) 1 at 11.

[78] Peterson v CIR [2005] UKPC 5. The Privy Council in fact found that New Zealand’s general anti-avoidance rule did not apply to this scheme.

[79] Case U32 (2000) 19 NZTC 9,302 (Taxation Review Authority) at [80].

[80] Robert A McLeod “Tax Avoidance Revisited” (2000) 6 NZJTLP 103 at 103.

[81] See, for example, Friedrich A Hayek The Constitution of Liberty (Routledge, London, 1960).

[82] Joseph Raz “The Rule of Law and its Virtue” in Joseph Raz The Authority of Law: Essays on Law and Morality (Oxford University Press, Oxford, 1979) 210 at 211.

[83] Even Lon Fuller, who is strongly committed to the rule of law, accepts that isolated breaches do not automatically condemn a legal system. See Lon Fuller The Morality of Law (2nd ed, Yale University Press, New Haven (CT), 1964) at 170.

[84] John Rawls A Theory of Justice (The Belknap Press of Harvard University Press, Cambridge (Mass) 1971) at 242.

[85] Lon Fuller The Morality of Law (2nd ed, Yale University Press, New Haven (CT), 1964) at 168.

[86] As mentioned at [4.1.1] above, there are some cases that are definitely tax avoidance (although these cases are mainly ones that have been judicially decided to be tax avoidance); so general anti-avoidance rules’ sphere of application is not entirely unknown. Nevertheless, it is true to say that general anti-avoidance rules depend on their vagueness for their effectiveness.

[87] For example Judith Freedman “Defining Taxpayer Responsibility: In Support of a General Anti-Avoidance Principle” [2004] BTR 32 at 346.

[88] Challenge Corporation Ltd v Commissioner of Inland Revenue [1986] 2 NZLR 513 (PC).

[89] While taxpayers are usually extremely annoyed if their tax avoidance schemes are disallowed because of the operation of general anti-avoidance rules, general anti-avoidance rules do not impose criminal penalties, although some penalties are involved. It is arguable that it is more important for laws that impose criminal penalties to conform to the rule of law, see John Rawls A Theory of Justice (The Belknap Press of Harvard University Press, Cambridge (Mass), 1971) at 241.

[90] Probably the most famous statement on the morality of tax avoidance comes from Lord Tomlin in IRC v Duke of Westminster [1936] AC 1 (HL) at 19–20, where his Lordship stated that “every man is entitled if he can to order his affairs so that the tax attaching under the appropriate Act is less than it otherwise would be.”

[91] See further Zoë Prebble and John Prebble “The Morality of Tax Avoidance” (2010) 20 Creighton Law Review 693.

[92] Nevertheless, the exact amount of tax the each individual should pay is open to debate. It is questionable whether taxpayers who have paid the amount of tax specified by black-letter law can really be shirking a duty. See further Judith Freedman “Defining Taxpayer Responsibility: In Support of a General Anti-Avoidance Principle” [2004] BTR 332 at 337.

[93] See further Lord Templeman “Tax and the Taxpayer” (2001) 117 LQR 575 at 575.

[94] But see Stephen R Munzer “A Theory of Retroactive Legislation” (1982) 61 Tex L Rev 424. Munzer argues that people have no right to rely on their immoral acts not being retrospectively criminalised.

[95] Richard Posner Not a Suicide Pact: The Constitution in a Time of National Emergency (Oxford University Press, Oxford, 2006).

[96] David Cole “How to Skip the Constitution – review of Richard Posner Not a Suicide Pact: The Constitution in a Time of National Emergency” (2006) 53 NY Rev Books. The article is also available online at (last accessed 2 September 2011).

[97] People who move in the same circles as tax advisors may dispute this statement. But who has heard of a mainstream political party campaigning for support to repeal a general anti-avoidance rule?

[98] “Hedged Options Monetization of Economic Remainder” (HOMER) is a fraudulent tax evasion scheme. See for instance United States v. Ohle 2010 U.S. Dist. LEXIS 2150 (SDNY No. S2 08 Cr. 1109 (LBS)).

[99] See also discussion in Kate Tokeley “Consumer Law and Paternalism: A Framework for Policy Decision-Making” in this volume (ch 10).