Sydney Law Review
H DAVID ROSENBLOOM[†]
This paper examines the United States tax treatment of so-called ‘related party debt,’ especially in transactions involving entities under common control. As a matter of economics and common sense, it is hard to see how such transactions can meet the normal tests used by United States law to distinguish debt from equity. The hallmark of debt is an unconditional promise to pay a sum certain on demand or at a fixed maturity in the reasonably foreseeable future; a transaction between legal entities under common control each of which is ultimately just a piece of paper is hard to reconcile with this standard. The paper suggests that the widespread recognition of related party debt as debt may stem from unquestioning acceptance of legal fictions like the corporate form, tolerance of elective elements in the tax system, and the common use of hypothetical tests to determine tax consequences.
My objective in this paper is to bring a little sunshine to the subject of related party debt. Or, to be more specific, I propose to explore the treatment under the income tax of a relationship that purports to be debt entered into between legal entities subject to common control. The income tax treatment — the wide acceptance — of this relationship is puzzling. The untutored eye does not perceive a relationship between separate (if related) persons in any real sense, nor do established authorities and ‘in substance’ analysis support the existence of debt. Yet acceptance of related party debt as debt is generally unquestioned. My thesis is that this ready acceptance flows from certain features of the income tax — legal fictions, elections, hypothetical determinations – which themselves are seldom placed in question, and which combine to make the remarkable phenomenon of related party debt appear perfectly normal and to place it virtually beyond inquiry.
The income tax system of the United States, like most other income tax systems, generally reflects the Haig-Simons concept of income. This concept holds that ‘income is the money value of the net accretion to one’s economic power between two points of time.’ The definition is important in two ways: it relates income to the accretion, or increase, of economic capacity within a prescribed time period and it points to the subject of taxation — the ‘one’ — whose increase in economic capacity forms the base of the tax.
As an instrument of social and economic policy, the Haig-Simons concept, and thus the income tax, rests on the notion that it is appropriate, reasonable and fair to link the burden that falls on each taxpayer with that taxpayer’s ability to pay. Such ability, in turn, is appropriately, reasonably and fairly determined according to periodic accretions to the taxpayer’s wealth, in the form of income.
Since it is generally necessary to expend resources in the pursuit of wealth, the concept of accretion of economic power is logically net of expenses incurred by the taxpayer in order to derive the income subject to tax. Such expenses will vary depending on the type of income-producing activity, and thus the costs reasonably needed to achieve wealth accretion will differ substantially from taxpayer to taxpayer. For example, the costs incurred by a taxpayer to earn income from the manufacture and sale of a product may be greater or less than the costs incurred by another taxpayer in manufacturing and selling a similar product, and may differ even more from the costs incurred by another taxpayer in earning income from the rendering of services. An income tax takes such variances into account by permitting subtractions from the tax base, in the form of deductions, for profit-seeking expenditures, subject to certain limitations. As proper costs of earning income, these deductions are entirely consistent with the Haig-Simons concept.
There are, of course, many differences between the US income tax system (or any other income tax system) and the pure Haig-Simons view of income. Much has been written about the problematic concept of realisation, generally used to mark the point at which accretion is sufficiently objectified for the resulting income to enter into the tax base. In addition, various types of accretion have been declared to fall outside that base, for various reasons. Certain costs of earning income have been inflated, accelerated, or denied deductibility.
In addition to these obvious, and much discussed, divergences from the Haig-Simons concept, there are various features of the US income tax that operate in more subtle ways to subvert the ‘ability to pay’ concept. Among these are three features on which little comment is ever offered but which profoundly affect the economic and legal landscape of the income tax. These are (1) legal fictions, (2) elections, and, (3) hypothetical ‘as if’ determinations — all firmly embedded in the tax system and each worthy of independent analysis. When these features converge, as I believe they do in the relationship known as related party debt, the result is a leading tax reduction tool. Such debt is generally accepted as debt because a tax system that readily accepts legal fictions, countenances numerous elections and relies upon hypothetical tests to determine the tax base does not find it at all strange that related party debt should be treated, in general, just like debt between parties that are not related.
If taxable income is the periodic accretion of economic power, taxation must depend on the accurate measurement of that accretion. Clearly, a distortion in any element of the formula (taxable income equals gross income less related deductible costs), whether in the cognisable cost of an asset, recognised gain, or related costs, will result in a corresponding distortion of the base subject to tax. As distortions multiply, the gap between accretion of economic power and taxable net income widens. Distortions are, of course, exacerbated by unsurprising taxpayer efforts to bend elements of the measurement process to the goal of minimising liability. Intentional tax reduction involves, in essence, only a few basic techniques — reduction of gross income, inflation of deductible expenses, deflection of the tax base to others standing in a favoured tax position. All these techniques are greatly enabled by features of an income tax system that lend themselves to manipulation at the taxpayer’s will.
Most nations that follow a rule of law accept as given certain important legal fictions. Indeed, a basic aspect of legal training, absorbed by law students throughout the world, is that the ‘law’ operates with unshakable acceptance of such fictions. In these circumstances, it would be practically impossible for any jurisdiction to disregard fictions altogether in shaping its rules of taxation, if only because those fictions have important non-tax consequences.
One ubiquitous legal fiction is the corporate form — as fictitious a construct as could be imagined. ‘Legal persons’ owing their existence to rules of law, corporations generally can act in their own capacity, sue and be sued, take and hold property. They were adopted in 17th Century England as a mechanism to spur trade by shielding merchants from unlimited liability, and have always been an economic extension of their owners enjoying a special privilege granted by the sovereign. Developments in legal and economic systems in most industrialised countries and the advent of public shareholding (whereby ownership by large numbers of people became common) gave the corporation an entirely new role. In addition to being a business accommodation, a security mechanism, the corporation became a virtually autonomous economic unit conducting business without control, or even oversight, by its owners. The corporation thus came to form associations on its own, and to function itself as a shareholder and partner in business associations. In this manner, the corporation has become so accepted, so anthropomorphised, that it is perceived not only by those trained in the law, but in the popular press and by the public generally as having a real existence apart from both its owners and its managers.
The corporate ‘person’ is also malleable. Unlike the human being with which the tax law equates it, the corporation is perpetual; but it may be quietly terminated at the will of managers and shareholders. The corporation can also be brought quickly into life and then transformed into a different corporation, loaded with features to support ‘economic substance,’ or merged into other corporations by filing appropriate documents with appropriate authorities. It cannot be deprived of life and freedom under the US Constitution. In the modern world, it can migrate to countries other than the one in which it was formed, facing no immigration issues, with fiscal authorities in the new motherland welcoming the new corporate citizen without extensive inquiries into its lineage, substance, or purpose. In brief, the corporation has much greater possibilities than the natural person — requiring the tax laws to adopt commensurately multi-faceted rules and broadly (if quietly) affecting tax policy at both the stage of its formulation and that of its implementation.
Corporations, however, are only pieces of paper. There is nothing ‘real’ about a corporation, nothing tangible, nothing capable of acting except through human agents, nothing with an independent will, intent, or moral capacity of its own. It is hard to understand what prosecutors mean or intend when they occasionally speak of criminal penalties for corporate malfeasors. Penalties that do not affect individuals have little deterrent power, and the individuals harmed by corporate prosecutions may easily be persons (owners) other than the malfeasors (managers).
Non-legal policymakers in the field of taxation, especially economists, are perhaps less likely to think about entities, including corporations, as independent persons, as opposed to extensions of the persons who own or control them. And it may be that a truer concept of the modern corporation is an economic business unit, and that this concept should replace, at least for fiscal purposes, acceptance of the corporate form as a legal person. Certainly US law relating to international taxation, with its rich blossoming of complex rules, sub-rules and interpretations of rules, could be vastly simplified by looking through the form of the controlled foreign corporation (and perhaps other corporations) to the substance of its controlling shareholders and developing rules accordingly. Nor are the complications limited to the international tax field. The economic justification for regarding a group of ten corporations as ten separate persons is hardly compelling. There may well exist non-tax justifications for creation of the group, but it is not foreordained that the tax laws should respect the results, particularly when the multiplicity of choice may be damaging to the goals of those laws. By assigning tax consequences to corporations (and other entities), a tax system effectively permits choices of tax results unattended by economic consequences. The legal fiction accords taxpayers a substantial measure of control over the computation, and therefore ultimately the amount, of tax they are called upon to pay.
Since a corporation is a convenient mechanism for collection of tax, the US system considers it an income taxpayer and imposes tax on it. The corporation is thus a non-transparent person for US tax purposes, with its own tax base. Whether a corporate tax could be made consistent with otherwise ignoring the corporation as a separate person is an interesting and intricate question. At a high level of generality, there does not appear to be a fundamental inconsistency between collecting tax from the corporation and according it something less than full, individual-equivalent, ‘person-hood.’
In any event, other entities, constructed pursuant to other legal fictions, are generally regarded as transparent by the tax laws, so that income and tax consequences pass through them to their owners. Once the corporation is accepted for tax purposes as an independent person, it must be differentiated from these other persons that give rise to dramatically different tax consequences. This is not easy, and countries have taken varying approaches to the task.
Elections, like legal fictions, are probably inevitable in an income tax system. In numerous instances, a choice between alternatives would be unjust, or would lead to otherwise undesirable results. In addition, elections are politically attractive: they satisfy everyone. Forced to choose between solution A and solution B, the legislator passes the choice to the taxpayer, who can be counted upon to make the choice in his or her own favor. Everyone, ostensibly, wins.
The problem is that, like legal fictions, elections have costs. They add complexity to a tax system, because an explicit election requires rules to govern each of the elective possibilities. In addition, elections affect the fairness, or perceived fairness, of the system when their availability is limited to certain classes of taxpayers, or when they can be revoked from time to time, as a taxpayer’s underlying situation, and therefore his or her tax objective, changes. The overall impression tends to be a tax system subject to the taxpayer’s control, a concept antithetical not only to an income tax but to any tax. Taxation is compulsory, not elective, and involves no obligation on the part of government other than the provision of public goods and services. Taxation is thus different from raising government funds through charitable contributions, issuing debt or establishing slot machines. An election for a taxpayer — or certain taxpayers — to pay tax at either a 15 per cent rate or a 20 per cent rate would be absurd.
Some elections in the income tax are explicit: the taxpayer is allowed to choose whether to allocate expenses according to the cost basis of assets or their fair market value; interest expense of a foreign person is attributed to a host country business using either of several alternative methods. In these instances, the policymaker might ask whether it is really necessary to provide such alternatives. Why would a blended approach not work for everyone? Of course, political forces push in the direction of electivity (they always do — elections are never to the taxpayer’s detriment), but such forces also push in the direction of abandoning taxation altogether. The policymaker will strive for compromise and, in many instances, will find it impossible to withstand the temptation to make everyone content. But the use of elective alternatives undeniably contributes to the vulnerability of the tax system on a variety of grounds.
Nor are elections necessarily explicit — and therefore readily identifiable to persons attempting to evaluate them. When taxpayers have a choice of transactions that are economic alternatives but that produce different tax consequences, the choice raises as many issues as, and perhaps even more than, the explicit election. Here the problem is not the multiplicity of rules. Different rules are pre-existing, because the system has identified a difference in transactions: payment of a dividend is, arguably, different from payment on a derivative contract. When, however, a taxpayer may use either transaction to obtain identical, or very similar, economic results, most of the issues relevant to explicit elections are present. The economic difference between a dividend and payment on an equity swap is arguably not compelling. There will be differences in the creditworthiness of the payors, but those differences may not justify different tax consequences.
The two types of elections — explicit and transactional — and the relationship between them are nicely illustrated by the US experience with rules of entity classification and their incarnation in the famous ‘check the box’ regulations of recent vintage. Under the pre-existing Morrissey factors, which stemmed from a Supreme Court decision in 1935, entities were classified as corporations, partnerships, or trusts according to a mechanical application of factors: an objective of carrying on business and dividing the profits, associates, limited liability, centralisation of management, free transferability of interests, continuity of life. The original regulations adopting these factors after Morrissey contained a bias against the association (taxable as a corporation). There were an even number of factors to weigh and corporate status was assigned only when a majority of the factors pointed to association status. Tax administrators were concerned that professionals such as doctors and dentists would incorporate their practices to avoid application of what were then extraordinarily high rates of individual taxation. But as often happens in a complex world, the regulations remained on the books after the issues putting pressure on the question of entity classification changed. In the late 1960s and 1970s, the focus became tax shelters that depended on an entity affording limited liability to investors but that was transparent, allowing a flow-through of losses and credits. The bias against the non-transparent corporate form came to operate to taxpayers’ benefit, as they took advantage of the regulations to engage in a great deal of shelter activity.
That activity put increased pressure on the precise meaning of the various Morrissey factors, each of which developed a high gloss and a nuanced meaning. In addition, with expanding internationalisation in the 1980s, the entity classification rules came to apply with frequency to exotic foreign entities. It was far from clear how to apply the factors, with their refined interpretations, to these novel constructs. Generally, taxpayers could achieve the results they wished for a foreign entity, transparent corporation or non-transparent partnership, with little sacrifice of economic substance, by tweaking the entity to control the result under the Morrissey tests. The Internal Revenue Service was finding it necessary, in order merely to keep up, to track the latest developments in regard to Peruvian SAs and entity offerings in Niger. And the stakes rose substantially when states of the United States adopted, one after another, the now familiar limited liability companies which could, in general, be used instead of the corporate form to limit liability without meeting the Morrissey tests for an association. This permitted foreign persons investing in the United States to engage in many of the same elective transactions as US persons using foreign entities (or US limited liability companies) to invest abroad.
Thus it was that the Service raised the white flag of ‘check the box,’ allowing taxpayers an explicit election in place of the transactional election they had previously enjoyed. It was believed that costs in terms of tax administration would be reduced, the situation would be more open and accessible to the uninformed, and the law would be, in some sense, more ‘honest.’ This was a forthright and proper action by the Service, assuming there had to be any election at all. Years earlier, the Treasury had proposed that there ought not to be an election here — that if an entity afforded limited liability to its owners, that fact in and of itself should result in corporate classification. Protests from taxpayers (citing Morrissey) led the Service to withdraw the proposal and continue with the transactional election based on the Morrissey factors. It was only when the government’s patience with the factors ran out, and frustration with the implications for tax administration began to grow, that the transactional election gave way to the explicit election of ‘check the box’.
Many lessons for policymakers (and students of taxation) lurk in this story. If the legal fiction of the corporation as a separate non-transparent person is respected for tax purposes, there must be at least two sets of rules, one for the corporation and one for other entities, such as partnerships, that are transparent. But there is no reason why the dichotomy cannot turn on a difference of substance, as was suggested in the 1970s. The gift of an election permits taxpayers to have it as they wish, transparent or non-transparent, at will. Moreover, since the formation and dissolution of entities is simple, migration from one status to the other is not a problem. Allowing taxpayers to choose whether the legal fiction through which they operate should be transparent or non-transparent seems both unnecessary as an intellectual matter and questionable as a matter of tax policy.
The US income tax allows many explicit and transactional elections — situations where taxpayers are permitted to choose one tax treatment or another without substantive cost. It would be foolhardy to venture the judgment that such elections should be abolished entirely, but policymakers would do well to recognise that elections burden the system, in terms of fairness, appearance (of fairness) and complexity. The Internal Revenue Code did not come to look the way it does by simply characterising for tax purposes the persons and transactions that could be foreseen to engage in economic activity; the heft of the document is due in large part to an understandable taxpayer desire for choice, coupled with an equally understandable politician desire for accommodation. At the end of the day, however, a choice for everyone results in a separate tax law for everyone, and that would not be a desirable policy option for a variety of reasons.
To make the point another way, if ‘check the box’ represents a sound response to the problem of entity classification, why not employ the same technique for the equally difficult issue of instrument classification? Taxpayers and tax administrators commonly encounter difficulties in distinguishing debt from equity. Could the answer be to permit the taxpayer to ‘check the box’? For that matter, if ‘check the box’ is suitable for entity classification in the United States, presumably it would be equally suitable for France or Japan. ‘Check the box’ has produced some serious problems for the US income tax, but either of the foregoing perfectly logical suggestions would render matters worse.
Aggressive use of the entity classification rules and the corporate form allows sophisticated taxpayers to obtain advantageous tax treatment. This has both policy and practical implications. A tax system will not be considered fair if some taxpayers may restructure their forms of doing business, without economic change, to achieve results that are unavailable to other taxpayers. Furthermore, there are obvious costs to such restructuring in the form of reduced tax revenues and the burden placed on tax administration.
Since an election, by definition, results in a reduction of income tax liability (without a reduction in economic accretion), it can be viewed as a tax expenditure. Although the government periodically estimates revenue losses attributable to tax expenditures, no study estimating the fiscal effect of transactional elections seems ever to have been undertaken. Nor, as a matter of methodology, is it clear how measurement of the effect of taxpayer changes of forms and characteristics to attain tax reductions could be accomplished.
From multiple points of view, elective provisions of the Code tend to operate at cross-purposes with the goals of the income tax. Elections, express and transactional, add complexity and revenue cost, and burden the effective administration of the tax laws. Their use almost certainly cannot be eliminated, but a degree of sensitivity to the harm they cause would be highly desirable.
Like legal fictions and elections, hypothetical tests are needed in the income tax: transaction A to be analysed as if it were transaction B and the result applied to transaction A. And this is doubtless the case without regard to legal fictions. It is not hard to imagine situations in which transactions having no fictional aspects really must be subjected to an ‘as if’ analysis: a father sells widgets to his son and the tax law asks what the price of the widgets would have been if the transaction had been between parties that were not related. It is clear, however, that legal fictions multiply both the number and variety of situations in which transactions cannot be accepted at face value, and that such fictions have thus contributed greatly to the proliferation of hypothetical inquiries that must be undertaken if the tax system is to remain compulsory.
The poster child for hypothetical determinations is, of course, the ‘arm’s-length’ method, a technique employed by the income tax to determine and test pricing for transactions between related persons. There are several well-documented problems with this method. The standard employed is imprecise, because normally there is no such thing as a single ‘arm’s-length’ price for transactions between unrelated parties, since the price actually used will depend on a nearly infinite number of variables relating to the knowledge, understanding, needs and skills of the parties. In addition, the data required to apply the standard are hard to find in any form, and second- and third-best substitutes for ideal data are easily manipulated by persons with an understanding of the rules of the game. The promised impartiality of the method thus dissipates into a familiar trading of horses in light of the relative leverage of the parties — not the leverage of the (related) parties to the transaction being tested, but the leverage between taxpayer and tax administrator when a controversy surfaces. In the United States, at least, the taxpayer can threaten to extend the examination, using a variety of procedures and means at his or her disposal; the administrator can threaten penalties and rely on the expense of litigation. Some things never change.
The difficulty with the ‘arm’s-length’ method that is most troubling, however, is neither its falsely promised precision nor the difficulties of obtaining the information needed to employ it, but its very nature: a hypothetical test. A transaction between related parties (and there are a potentially infinite number of related parties, once the corporation is accepted as an independent person) is analysed as if the parties were not related. Or, as I suggested at a conference in the 1980s, the question asked by the ‘arm’s-length’ method is whether, if you had a brother, he would like cheese.
There is, of course, a reason why related parties that engage in a transaction are related, and that reason, however described, is almost certainly lacking in the case of parties that are not related. Moreover, the reason almost certainly has an important influence on any transaction in which the parties engage. Thus, the attempt to ask what the parties would have done, how they would have dealt with one another, if they had not been related (as the ‘arm’s-length’ method effectively asks) is, in the most fundamental sense, nonsense; it is doomed to send the asker wandering into speculative considerations. Furthermore, the question is inherently ambiguous. What aspects of the tested (related party) transaction are to be accepted before the ‘as if’ test is applied — everything save the actual price? Suppose it can reasonably be concluded that, if the parties had not been related, they would not have engaged in a transaction at all, or would have engaged in a different kind of transaction, or would have satisfied their business needs with other persons in other ways? What does ‘arm’s-length’ — what does ‘as if’ — imply? The level at which the inquiry is pegged is not specified, leaving the analyst to determine that crucial point for himself. As a result, tax consequences flow from a tug of speculations between taxpayer and tax administration, inevitably giving rise to spirited controversy (and a dispersion of results across the taxing jurisdiction).
The point here is not to weigh the ‘arm’s-length’ method against potential alternatives, but to note the problems that the method, a hypothetical test, inevitably spawns. Such ‘as if’ testing and determining of tax results are not uncommon in the US income tax and, by reason of the prominent leadership role that the United States plays in the field of taxation, in other income tax systems as well. As most lawyers would acknowledge, nothing is ever just like anything else. ‘As if’ tests invariably produce unsatisfying results because they veer into the subjective and therefore the manipulable. Once again, a tax world in which such tests are abandoned altogether is probably not possible, but there is no good reason to employ them except when they are truly necessary.
One cost that most income tax systems allow as recoverable, or deductible, in most circumstances is business-related interest. This is simply the cost of money borrowed, the functional equivalent of rent paid for a machine. The item obtained for this cost — temporary use of money or machine — is used to earn income that will go first to recover the cost of the temporary use and then, it is hoped by the borrower or renter, return a profit. This profit, of course, becomes the base of the tax. It is neither extraordinary nor objectionable, as a general matter, that the cost of funds should be so viewed in the context of an income tax. Debt and interest are normal features of engaging in an income-producing business.
Given the purposes of the income tax, and the rationale for the interest deduction, related party debt is an oddity. As its name indicates, this is indebtedness between parties having a relationship to one another. There are, obviously, many ways of defining that relationship — of establishing the line that separates related parties from parties that are not related — but these definitional matters are beside the present point. Regardless of where the line is drawn, in most jurisdictions the line is of only limited, and specific, significance. As a general matter, related party debt is seen as debt, and in that capacity viewed as giving rise to income from money lent in the hands of the creditor and deductible expense for the debtor.
Related party debt is a principal tool of the tax planner. Since related parties may differ between themselves in any number of tax characteristics — residence, method of accounting, functional currency, exempt versus non-exempt status — the interest deduction and commensurate income from related party debt may produce overall beneficial tax results within a single economic unit. This may be of little consequence when individuals (father and son, for example) are the parties that are related. But related party debt is not limited to, or even most common in, such circumstances. Since the tax laws accept legal fictions and view each legal entity as a separate person, related party debt is found in the greatest volume between commonly controlled entities. In these circumstances, the differences in taxation between debtor and creditor can be used to obtain a tax-free return from a source country, drive down effective rates of tax, achieve a deduction without a commensurate income inclusion, and in many, many other situations.
Debt can also easily be hybridised, or treated as equity in one tax jurisdiction while retaining its status as debt in another, and related party debt is a critical feature of such instrument arbitrage. And entity arbitrage, of the sort commonly referred to as the ‘hybrid branch,’ would have far less utility without the use of related party debt. In such cases, one tax jurisdiction perceives a non-transparent entity (outside the jurisdiction) while the other sees a transparent, and disregarded, entity. Related party debt owed to the hybrid supports interest that will drive down the tax base of the jurisdiction from which the payment is made, while not being recognised in the jurisdiction that deems the entity, and thus the interest payment, not to exist at all.
In fact, when an observer casts around for related party debt, it is everywhere — transferring the incidence of taxation away from one taxpayer and into the hands of another, lower-taxed or non-taxed, often foreign, related taxpayer. This versatile technique is accepted generally as the equivalent of debt between unrelated parties — subject, of course, to policing for an ‘arm’s-length’ rate of interest and perhaps extra scrutiny in suspected thin capitalisation situations, but never otherwise in substantial doubt.
There seems to be only one serious problem with related party debt: by most standards of economics, ‘substance,’ or common sense, it is not debt. That is, related party debt is generally not compensation for money lent by one person to another. Rather, it is a transfer of funds from one incorporated pocket to another, usually for tax-reduction purposes. Only a tax professional, considering indebtedness between commonly controlled entities, would perceive a similarity to raising funds from unrelated parties. As noted, interest is generally deductible in the unrelated party context because it is a legitimate cost of the funds needed to operate a business. The application of this rationale in a related party context, where no funds are being raised, is one of the tax miracles of our time.
The US Tax Court came close to this point in an entirely different context in Albertson’s, Inc v Commissioner, which involved the deductibility of ‘interest’ accrued on certain unfunded deferred compensation arrangements. The court noted that ‘one cannot lend that which one never had a right to possess, and petitioner [seeking to deduct interest expense] cannot borrow money that it always had the right to possess.’ [Emphasis in original.] By the court’s simple reasoning, one cannot borrow from oneself. When an entity ‘borrows’ from a related entity, the resulting use of funds is not much different, in substance or reality, from the use asserted in Albertson’s. But for the legal fiction that recognises each entity as independent, borrower and lender are one and the same.
Related party debt appears to reflect an admixture of techniques, assumptions and methods of approach to tax policy that have become ingrained in the income tax laws of most nations, and certainly in those of the United States. It builds on the legal fiction of the corporation, relies on the common use of hypothetical tests to determine tax consequences and survives in a policy environment that readily countenances elective elements of taxation. As noted previously, each of these features of the tax system has side effects that, independently, produce questionable, sometimes ridiculous, results. Together, they combine to support a proliferation of related party debt.
Under the US Internal Revenue Code, the deductibility of interest on debt represents a sharp distinction from the treatment of equity, dividend payments on which are generally non-deductible. The distinctive tax treatment of each of two common ways of financing a business reflects, in essence, a perceived difference in the nature of the relationship between the party providing funds and the party raising funds for operation of a business. Though both interest and dividends may be viewed, in non-tax parlance, as costs of raising capital, there is a marked difference, without regard to tax laws, between equity and debt. A holder of equity is an owner of the business and dividends may fairly be viewed as a return for the risk of investment. A debt-holder, on the other hand, is not an owner of the business but, rather, a renter-out of funds, entitled to demand ‘a fixed sum at some period set in advance.’ The risk assumed by the debt-holder is limited, and the cost of raising funds through an issue of debt is much closer in nature to other costs incurred in a business venture.
Dividends are paid to the equity-holder out of the profits of the business. A lender, in contrast, receives income from a loan even before, and regardless of whether, the borrower is profitable. For that reason, debt and equity are not merely different in kind but, to some extent, operate at cross purposes. There will not be any profit for the equity holder to enjoy if costs, including interest costs, are too large.
Policymakers and courts in the United States have struggled for years with the problem of distinguishing debt from equity in particular cases. Often the inquiry has focused on whether an identified payment represents deductible interest or a non-deductible dividend. The US Congress, in enacting section 385 of the Code in 1969, enjoined the Treasury to look into this subject and develop suitable rules to distinguish debt from equity. Treasury attempted to do this over a period of years in the late 1970s and early 1980s before giving up, withdrawing its ‘final’ regulations on the subject and sending it back for resolution on a case-by-case basis in the courts. The judiciary had already developed 13 or more ‘factors’ that would permit the testing of an instrument to determine how to categorise it. At first glance, such a multifarious means of reaching a determination appears to reside in the eye of the beholder, and an objective commentator could be forgiven for concluding that, in a given case, the factors can usually be massaged to steer the decision-maker in the desired direction.
In 1994, the Internal Revenue Service issued a Notice providing (in an all too common but, alas, none too helpful, formula) that the characterisation of an instrument as debt or equity would depend on the instrument’s terms and all surrounding facts and circumstances. Following the approach of section 385, the Notice went on to enumerate eight factors bearing on the determination, with an admonition that the weight to be accorded any of them would depend on all the facts and circumstances.
Interestingly, the first three of the eight factors enumerated in the Notice are: (1) whether the instrument contains an unconditional promise by the issuer to pay a sum certain on demand or at a fixed maturity in the reasonably foreseeable future; (2) whether holders of the instrument possess the right to enforce payment of principal and interest, and (3) whether the rights of holders are subordinate to the rights of general creditors. Although the Notice prophylactically states that no particular factor is conclusive in making a determination, it is hard not to ascribe some significance to the fact that the ‘unconditional promise’ factor appears in the list well in advance of the ‘label’ factor. The Notice, in that regard, is in step with much of the sprawling case law, which is apt to conclude that the real issue — what a Technical Advice Memorandum issued by the Revenue Service in 1998 called ‘the important issue’ — is whether there was a ‘genuine intention to create a debt, with a reasonable expectation of repayment, and [whether] that intention comport[ed] with the economic reality of creating a debtor-creditor relationship?’ This accords with the observation, for example, that debt provides for ‘repayment absolutely and in all events, or that principal be secured in some way as distinguished from being put in hazard.’
The debt/equity issue goes off in many, sometimes subtle, directions, and there can be circumstances in which an apparent entitlement to repayment may be accompanied by terms that place the holder at sufficient risk to generate a strong flavor of equity. It seems clear, however, that a fixed right to repayment, on the one hand, and participation in the risk of success of the business, on the other, are very basically at odds. In the absence of an explicit or clear participation in risk, the ability of a holder to demand repayment from a debtor at a reasonably near-term date certain is a definite hallmark of debt.
Many difficulties encountered in distinguishing debt from equity affect instruments issued by one related party to another. US courts purport to take this circumstance into account in their deliberations, if only obliquely, but there is little evidence that they have ever struggled with the implications of the relationship. Certainly, there is no recognition of the possibility that there might be a difference between related parties that are human beings and related parties that are entities under common control.
Related parties of the latter type warrant close scrutiny in light of the principles that have been developed in the debt/equity wars. As noted earlier, such related parties are potentially infinite in number. Entities constituting independent parties can be formed and unformed at will in the number that the person or persons ultimately in control desire. Whatever else may be said about acceptance of each entity as separate from the others, in the context of an ostensible ‘debt’ owed by one such party to another, it strains credulity to say that ‘repayment absolutely and in all events’ is envisioned or provided for. Two pieces of paper owned ultimately by a single economic interest cannot do business together, no matter what legal sophists say. There is lacking in such ‘transactions’ the frictional element that permits an outsider, such as tax authorities, to give credence and respect to the results.
It follows that the tax law must resort to its customary fallback, the hypothetical test, to determine whether, if the entities had not been related, they would have transacted as they ostensibly did. That type of inquiry, in the case of related party debt, focuses on the interest rate charged — whether it meets the ‘arm’s-length’ test for rate, terms and so on — but perhaps also on the fundamental question of whether the transaction would have occurred at all if friction had been present. That, of course, is an unanswerable question in a related party debt/equity context: not difficult, not complex, unanswerable.
Nevertheless, US law has proceeded to answer it, using the 13 factors (and others) heretofore described, or whatever substitute passes for a debt/equity test in the jurisdiction with decision-making authority. If the transaction passes muster under the test for thin capitalisation — that is, if the planners of the transaction have successfully drafted the instrument to accord with the expected jurisprudence — a finding of debt is entirely possible.
This approach makes related party debt a transactional election of the type described earlier, which the tax planner may choose, or not choose, to put in place. It would seem that this fact, by itself, would be inconsistent with leading indicators of debt status under the tests normally applied. For whatever an instrument says on its face, whatever artfully drafted surrounding documents may say, an obligation owed by one piece of paper to another (both controlled by the same interests) is hard to classify as an ‘unconditional promise’ to pay a ‘sum certain’ at a ‘fixed maturity date’ (or on demand) within the ‘reasonably foreseeable future.’ It may have the appearance of touching each of these bases, but the reality of the situation belies satisfaction of any of these points. There is no ‘unconditional promise’ when the parties can, with a wave of the pen, waive their prior handiwork. There is no ‘sum certain’ when the sum identified can be decreased or increased through purely formal (and easily accomplished) means. There is no ‘fixed maturity date,’ certainly not in the ‘reasonably foreseeable future.’
Related party debt is a common, much used, element in tax shelters, cross-border tax planning (both ways), earnings stripping, reduction of the income of controlled foreign corporations through hybrid branches, and many, many other troublesome aspects of US income tax law. If they were analysed dispassionately in light of the principles that have developed for distinguishing debt from equity, ostensible loans from one entity to another would probably be viewed as equity transfers. If related party debt were not accepted as debt, or at least were not capable of giving rise to deductible interest for the purported debtor, a critical elective element in the law would be eliminated. An important set of hypothetical ‘as if’ determinations would no longer be necessary. Many sections of the Internal Revenue Code would diminish in importance, perhaps to the point of irrelevance. A number of difficult ‘look through’ rules could be simplified. Remaining tax policy choices would be more transparent to legislators, other policymakers, tax professionals, perhaps even taxpayers. The result would be a clearer reflection of income as accretion of economic power, and a more equitable income tax system.
[*] This paper was prepared for delivery as the 2003 Ross Parsons Lecture, at the Supreme Court of New South Wales, Sydney, on 28 August 2003. In a slightly different form it was published in Tax Notes International Vol 32, 15 December 2003 at 989. The author would like to thank his colleague Vitaly Timokhov, for his very substantial contribution.
[†] Director, International Tax Program, New York University School of Law.
 Tax professionals necessarily remain tethered, to a lesser or greater extent, to the tax systems that are most familiar to them. The discussion that follows is grounded in the income tax system of the United States.
 See Robert Haig, ‘The Concept of Income — Economic and Legal Aspects’ in Robert Haig (ed), The Federal Income Tax (1921) at 1, 7; Henry C Simons, Personal Income Taxation: The Definition of Income as a Problem of Fiscal Policy (1938) at 50.
 There are, of course, differing views on the income tax, and specifically whether it is in fact an appropriate, reasonable and fair type of levy. One view is that a taxpayer’s fiscal burden, his or her contribution to government resources, should be measured not by what flows to the taxpayer but, instead, by what the taxpayer removes from society, or consumes. This approach supports a consumption tax, whether sales, value added, or other form of levy that meters the taxpayer’s burden in accordance with outflows rather than inflows. The approach is bolstered by the notion that accretions to wealth in private hands are requisite to economic growth, and that it is counterproductive to overall welfare to diminish that wealth by taxes on income not spent for consumption.
 An associated, but independent, rationale for the income tax is that persons with greater ability to pay have evidenced greater utilisation of government services and other benefits that taxes fund. If this were in fact the case, a requirement that such persons pay a greater amount of tax would be tied to the costs those taxes defray.
 This is independent of the point that periodic accretion includes, of necessity, periodic consumption.
 As a general matter, only profit-seeking expenditures should reduce income. The purpose of the limitation is to distinguish between legitimate costs of producing income and expenses for personal consumption. Another major limitation on deductibility involves timing and the factual connection between income and the costs of its production. Thus, a current expense is immediately deductible, whereas expenses conferring a lasting benefit are deducted over a period of time.
 See, for example, David Shakow, ‘Taxation Without Realization: A Proposal for Accrual Taxation’ (1986) 134 U Penn LR 1111.
 US income tax law nevertheless recognises the economic identity between corporation and shareholder in various places where the recognition is beneficial to taxpayers, such as S corporations, the deemed paid foreign tax credit and the treatment of limited liability companies. See IRC (Internal Revenue Code) ss1341, 902, 960; 26 CFR (Code of Federal Regulations) ss301.7701–2, 3.
 Thus, many corporations operate, in reality, wholly independently of their shareholders. It is clear that these owners, with some exceptions, have no say in regard to day-to-day operations of these large constructs. Managers (who need not be owners) are basically free of owner-determined constraints, except in extreme cases.
 The first indication of an abusive use of the corporate form in US tax matters came in the 1930s when, as a result of differences in tax rates, corporations were used to park income away from the higher rates imposed on individuals. In 1934, Congress introduced the personal holding company rules, imposing a penalty tax on these ‘incorporated pocketbooks’ by taxing earnings accumulated in corporate solution. The sanctity of the corporation was not questioned – although a simpler and economically more acceptable result would have been achieved by ‘piercing the corporate veil’ and attributing income of the incorporated pocketbook to its owners.
 With few exceptions, corporations have all the constitutional and legal rights of a natural person. Although corporations cannot vote, they can lobby elected members of the government and cannot be deprived of life, liberty, or property without due process of law. First National Bank of Boston v Bellotti  USSC 138; 435 US 765 (1978).
 See, for example, IRS Field Service Advice Memorandum 200117019 (27 April 2001), involving a corporate ‘continuation’ from the United States into Canada.
 See, for example, Larry Thompson, Deputy Attorney General of the United States, ‘ “Zero Tolerance” for Corporate Fraud’ Wall Street Journal (21 July 2003) at A10: ‘It is a bedrock principle of American law that business organizations, including corporations, may be held to account by the criminal law for the wrongdoing of employees or agents.’
 See H David Rosenbloom, ‘From the Bottom Up: Taxing the Income of Foreign Controlled Corporations’ (2001) 36 Brooklyn JIL 1525.
 The US tax world is made more complicated by the long-standing existence of transparent corporations and non-transparent trusts and estates. See IRC ss1301 and following. (S corporations); ss641 and following (trusts and estates).
 The differences among taxing jurisdictions on the definitional question – what, exactly, is a corporation – produce opportunities for taxpayers to ‘arbitrage’ those differences with a view to tax minimisation. See H David Rosenbloom, ‘International Tax Arbitrage and the “International Tax System” ’ (2000) 53 Tax LR 137.
 26 CFR s1.861–9.
 26 CFR s1.882–5.
 See generally Alvin Warren, ‘Financial Contract Innovation and Income Tax Policy’ (1993) 107 Harv LR 460.
 26 CFR s301.7701(a)–2, 3. The regulations were adopted by TD (Treasury Decision) 8697 (18 December 1996), 1997–1 Cumulative Bulletin 215, and went into effect generally as of 1 January 1997.
 Morrissey v Commissioner  USSC 212; 296 US 344 (1935).
 See, for example, Larson v Commissioner 66 TC 159 (1976); Foster v Commissioner 80 TC 34 (1983); Zuckman v United States 524 F 2d 729 (1975); MCA v Commissioner  USCA9 1502; 685 F 2d 1099 (1982); IRS Revenue Procedures 92–35, 1992–1 Cumulative Bulletin 790; IRS Revenue Procedures 89–12, 1989–1 Cumulative Bulletin 798.
 See, for example, IRS Revenue Ruling 77–214, 1977–1 Cumulative Bulletin 408; IRS Revenue Ruling 88–8, 1988–1 Cumulative Bulletin 403; IRS Revenue Ruling 93–4, 1993–1 Cumulative Bulletin 225.
 See, for example, IRS Revenue Ruling 93–6, 1993–1 Cumulative Bulletin 229 (Colorado); IRS Revenue Ruling 93–38, 1993–1 Cumulative Bulletin 233 (Delaware); IRS Revenue Ruling 94–51, 1994–2 Cumulative Bulletin 407 (New Jersey).
 For the reasons that led Treasury to abandon its initial approach to entity classification, see Notice of Proposed Rulemaking, Simplification of Entity Classification Rules, 61 FR 21989 (13 May 1996).
 An independent question addressed by the ‘check the box’ rules — whether a transparent entity owned by a single person should be disregarded — represents a distinct, and arguably more problematic, tax policy judgment. The rule for so-called ‘single-member entities’ was the product of rigorous logic (how could a transparent entity owned by a single person have an independent existence for tax purposes?) but, it would appear, little independent analysis.
 Notice of Proposed Rulemaking, Classification of Limited Liability Companies (17 November 1980), Proposed Regulation s301.7701–2(a)(2) (1980): ‘... an organization will be classified as an association [taxable as a non-transparent corporation] if under local law no member of the organization is personally liable for debts of the organization.’
 The withdrawal was accomplished by a Withdrawal of Notice of Proposed Rulemaking, Classification of Limited Liability Companies, 48 FR 14389 (4 April 1983). The decision in Morrissey was, of course, not constitutionally mandated.
 Trusts tend to be a stepchild in this analysis, though the many advisors who make a living by sweeping the Internal Revenue Code for gold are well aware that the trust form exists, and can be exploited.
 Tax expenditures are defined under section 3(3) of the Congressional Budget and Impoundment Control Act of 1974, Public Law Number 93-344 88 US Statutes-at-Large 297, as ‘revenue losses attributable to provisions of the Federal tax laws which allow a special exclusion, exemption, or deduction from gross income or which provide a special credit, a preferential rate of tax, or a deferral of tax liability.’
 See IRC s482 and the regulations thereunder. The purpose of section 482 and the regulations is to place ‘a controlled taxpayer on a tax parity with an uncontrolled taxpayer by determining the true taxable income of the controlled taxpayer.’ 26 CFR s1.482–1(a)(1). Or, in the somewhat starker terms used by the Tax Court in Your Host v Commissioner, section 482 ‘is designed to remedy only one abuse: the shifting of income from one commonly controlled entity to another.’ 58 TC 10 (1972), 24, affirmed USCA2 738; , 489 F 2d 957 (1973). In determining the true taxable income of a controlled taxpayer, ‘the standard to be applied in every case is that of a taxpayer dealing at arm’s length with an uncontrolled taxpayer.’ 26 CFR s1.482–1(b)(1).
 See, for example, Stanley Langbein, ‘The Unitary Method and the Myth of Arm’s Length’ (1986) 30 Tax Notes 625.
 It is no accident that the name of the taxpayer in the ‘cheese examples,’ Examples (2), (3), and (4) of 26 CFR s1.482–4(f)(3)(iv), is Fromage Frere. Alas, newly proposed regulations would dispense with these examples. See Notice of Proposed Rulemaking, Treatment of Services under Section 482, Proposed Treasury Regulation s1.482–4(f)(3) 68 FR 53447.
 See Staff of Joint Committee on Taxation, General Explanation of the Tax Reform Act of 1986 (Washington: JCP, 1987) at 1014: ‘Observers have noted that multinational companies operate as an economic unit, and not ‘as if’ they were unrelated to their foreign subsidiaries.’
 The observation is not intended to conflict with the classroom observation of the late Paul Freund, Professor of Constitutional Law at Harvard Law School, that, on the other hand, everything is just like everything else.
 See Deputy v du Pont  USSC 7; 308 US 488 (1940), 498, where the Supreme Court authoritatively defined interest as ‘compensation for the use or forbearance of money.’ In a later decision, Dickman v Commissioner  USSC 73; 465 US 330 (1984), 339, the court adopted a more ‘economic’ approach, defining interest as the equivalent of ‘rent’ for the use of funds.
 See, in the United States, ss163(e)(3), 163(j) and 267 of the Internal Revenue Code for examples of rules that limit the interest deduction on related party debt, and s904(d)(3)(C) for a different type of special treatment.
 See IRC ss61(a), 163.
 A modicum of care is, however, needed. In Laidlaw Transportation v Commissioner 75 TCM 2598 (1998), failure to observe the original terms of a purported debt instrument (treated as equity by a Canadian parent corporation) caused the Tax Court to recharacterise the instrument as equity for US tax purposes and disallow interest deductions to US subsidiaries.
 IRS Notice 98–11, 1998–1 Cumulative Bulletin 433, an early warning that such transactions might be questionable on policy grounds, has had no effect.
 Related party debt is not unknown in a purely domestic context, but the rules relating to consolidated returns for affiliated groups deprive it of much of its utility there.
 An exception would be the Tax Court Memorandum decision in Laidlaw, above n39, where the court purported to look to ‘substance’ in classifying ostensible related party hybrid debt as equity for US purposes and denying interest deductions claimed by the taxpayer. In Laidlaw, however, original maturity dates had been extended and interest payments had been re-lent to the putative borrower, and the court, at 2624 emphasised these facts: ‘The substance of the transactions is revealed in the lack of ‘arm’s-length’ dealing between [a foreign corporation] and petitioners [US subsidiaries], the circular flow of funds, and the conduct of the parties by changing the terms of the agreements when needed to avoid deadlines. The Laidlaw entities’ core management group designed and implemented this elaborate system to create the appearance that petitioners were paying interest, while in substance they were not.’ Thus, the court was able to sidestep the broader implications of its earlier observation that possibilities, not actual events, should drive substantive analysis, at 2616: ‘if a transaction is controlled by related entities, the form and labels used may not signify much because the parties can mold the transaction to their will’, citing Anchor National Life Insurance v Commissioner 93 TC 382 (1989), 407. Laidlaw is a Memorandum decision, with no precedential value.
 95 TC 415 (1990).
 Id at 423.
 It is true that, since the corporate fiction is accepted for non-tax as well as tax purposes, there can be real non-tax consequences, for example in a case of insolvency, to related party debt. But there is no reason why the mere possibility of non-tax consequences should make a general case for equating related party debt with unrelated party debt.
 Jewel Tea v United States 90 F 2d 451 (1937), 453. See generally William Plumb, ‘The Federal Income Tax Significance of Corporate Debt: A Critical Analysis and a Proposal’ (1971) 26 Tax LR 369.
 See, for example, In re Lane  USCA11 1404; 742 F 2d 1311 (1984). The discussion in the text is obviously not intended to deal comprehensively with the difference between debt and equity as a theoretical matter. It could, for example, be maintained that there should be no deduction for interest on debt. But the dichotomy described in the text is found in most tax systems and presumably reflects a common judgment that distinctions of the sort outlined in the text are justifiable.
 Public Law Number 91–172, s415(a). Notably, the section applies only to corporations, leaving case law as the only authority for transactions involving other entities.
 TD 7747, (31 December 1980) 1981–1 Cumulative Bulletin 141, withdrawn by TD 7920 (3 November 1983), 1983–2 Cumulative Bulletin 69.
 A survey of Tax Court decisions over the period from 1955 to 1987 actually identified 26 separate factors that went into the analysis. Paul Robertson, Zoel Daughtrey & Daryl Burckel, ‘Debt or Equity? An Empirical Analysis of Tax Court Classification During the Period 1955–1987’ (1990) 47 Tax Notes 707.
 Notice 94–47, 1994–1 Cumulative Bulletin 357.
 The other listed factors are whether the instrument gives holders a right to participate in management of the issuer; whether the issuer is fully capitalised; whether there is identity between the holders of the instrument and the owners (stockholders) of the issuer; the label placed upon the instrument, and whether the instrument is intended to be treated as debt or equity for non-tax purposes. See also IRS Revenue Ruling 2003–97, 2003– 34 IRB 380, adding ‘the intent of the parties,’ which would seem difficult to establish independently of other factors.
 IRS Technical Advice Memorandum 199910046 (16 November 1998).
 Litton Business Systems v Commissioner 61 TC 367 (1973), 373.
 Hartman v Commissioner 17 TCM 1020 (1958), 1023.
 Notice 94–47, above n51, lists as a factor the identity between holders of an instrument and shareholders.
 Again, the decision in Laidlaw, above n39, provides a hint of an exception. See above n42.
 See, for example, IRS Private Letter Ruling 8538034 (21 June 1985), involving the question of whether short-term discount paper issued by a US corporation to its foreign affiliates would qualify for exemption from tax at source under IRC ss881(a) and 871(g): ‘Taxpayer has represented that [its] ... foreign subsidiaries will be free to exercise their own independent judgment ... and will be under no obligation either to purchase or reinvest in commercial paper issued by the Taxpayer.’ The private ruling was withdrawn by IRS Private Letter Ruling 8612023 (18 December 1985), as ‘too dependent on subsequent facts’ — namely, whether the subsidiaries would ‘exercise their own independent judgment.’
 See Laidlaw, above n39 at 2616: ‘If a transaction is controlled by related entities, the form and labels used may not signify much because the parties can mold the transaction to their will.’