Note: The following is an excerpt from the introduction to the article as published in The Tax Lawyer. Author citations have been omitted for brevity. Tax Section members may read the article in its entirety in Adobe Acrobat format.
THE TRANSFER PRICING PROBLEM: A GLOBAL PROPOSAL FOR SIMPLIFICATION
*Associate Professor of Law, Universidad Torcuato Di Tella, Buenos Aires, Argentina; Universidad de Buenos Aires, LL.B., 1990; Harvard Law School, LL.M. (Constitutional Law), 1993; London School of Economics and Political Science, LL.M. (International Tax Law) 1999. I have benefited from comments on previous drafts of this Article by Benjamin Alary, Reuven Avi-Yonah, Richard Bird, Yariv Brauner, David Duff, Marcelo Ferrante, Edward Iacobucci, Emiliano Marambio Catán, Diane Ring, Eduardo Rivera López, Thomas Ulen, Lode Vereeck, and participants in workshops at the University of Toronto, University of California at Berkeley, and Universidad Torcuato Di Tella Law Schools. All errors are my own responsibility.
The world has experienced two globalization booms and one bust over the past two centuries. The first boom started about 1820 and lasted until the advent of World War I. The second began at the end of World War II and has continued since. The interwar years witnessed a retreat from this otherwise continuous shift towards greater global integration. 1 One major consequence of the globalization movement was the emergence, in the late 19th century, of a novel strategic problem among nations: how to divide the international income tax base in the absence of a centralized authority. Developed countries eventually reached a fundamental consensus on how to solve this problem. 2 That consensus is currently embodied in the Organization for Economic Cooperation and Development (OECD) Model Tax Convention on Income and on Capital (OECD Model). The OECD Model is the foundation for a network of over 2,500 bilateral tax treaties, which is referred to as the international tax regime. 3 The OECD Model is based largely on a web of standards, as opposed to rules, whose precise meaning can only be determined with certainty ex post via case law or a functional equivalent to case law. 4 In effect, key norms of the OECD Model do not have precise meanings before the taxpayer acts; their precise meanings are determined on a case-by-case basis after the taxpayer acts. 5 The central role of the OECD Model is to minimize international double taxation by establishing some structural legal fictions to guide the division of the international income tax base. 6 The international taxation of multinational enterprises (MNEs) is a case in point. Since about 1928, by consensus, developed countries created the legal fiction of the separate entity approach. Under this approach, the different profit units of a given MNE should be deemed independent enterprises. For example, pursuant to the OECD Model, IBM parent company (based in the United States) and each of its subsidiaries (based in a number of other countries) should be considered separate taxpayers for purposes of national corporate income taxes. Developed countries also agreed that the fiction of the separate entity approach should be enforced via the arm’s length standard (ALS). 7 The ALS provides that national tax jurisdiction over income produced by an MNE should be allocated among countries on the basis of how comparable unassociated enterprises would have realized income in comparable circumstances. For example, the ALS provides that the transfer pricing method (agreed to by IBM and one of its subsidiaries) for a certain type of hardware must be consistent with the market price of a comparable type of hardware that independent companies would agree to in similar circumstances. If the ALS is not met in a given case, national tax authorities are normally vested with the power to adjust the transfer pricing of associated enterprises to make it consistent with the ALS. The concept of associated enterprises, a structural element of the ALS, is standard based because its meaning is uncertain ex ante. 8 Scholars have devoted considerable energy to determining the precise meaning of associated enterprises. 9 Predictably, that effort has been largely unsuccessful due to the lack of case law with public good features in this area. 10 For certain strategic reasons, developing countries of all legal traditions and cultures have been systematically importing the international tax regime since the early 1960s. The adoption of the international tax regime has not been limited to developing countries with open economies. For example, treaties based on the OECD Model have been concluded by developing countries from both the Islamic world, such as Iran, 11 and the Communist world, such as North Korea and Cuba. 12 The massive importation of the international tax regime by developing countries has been problematic in areas like enforcement (consider, for example, the experience of Argentina in the context of transfer pricing). 13 Case law in developing countries has less relevance with respect to predicting court decisions because of the political instability and the relatively weak observance of the rule of law in the respective countries. 14 For instance, since 1947, Argentina’s entire Supreme Court has “been replaced en masse five times,” 15 which has lead to sudden changes in the Argentine transfer pricing case law. The unstable case law implies that the precise meaning of the international tax regime’s standard based norms, such as the ALS approach, remains unknowable in much of the developing world. Since the late 1970s, developed countries have also faced severe problems with the enforcement of the ALS. This is because the emergence of confidential advance pricing agreements (APAs) 16 has made litigation rare in the context of transfer pricing. Moreover, the case law that is available is not a public good because the holdings are typically too fact specific to allow a representative taxpayer to predict the probable outcome of a future court’s decisions, especially when no comparables are readily available. 17 In sum, for different reasons, both the developed and developing worlds are facing the same fundamental problem: the meaning of the ALS is increasingly unknowable because of the absence of case law with public good features in this area. Therefore, the ALS is unable to provide taxpayers with a clear sense of how they are expected to behave in the legal system in which they operate. In part, this scenario accounts for the worldwide ALS crisis. 18 As an unfortunate product of the ALS crisis, a wave of transfer pricing litigation has emerged in both the developed and developing worlds since the beginning of the 21st century. 19 For example, GlaxoSmithKline (Glaxo), a British pharmaceutical giant, filed suit against the Commissioner of the Service early in 2004; with $5 billion at stake, it is the largest transfer pricing litigation in world history. 20 The purpose of this Article is twofold. First, it seeks to provide an analysis of the problem faced by the ALS when the legal system in which it operates is unable to produce case law with public good features ( i.e., case law capable of providing reliable guidance with respect to the precise meaning of the ALS in other cases). The U.S. and Argentine transfer pricing experiences are used as case studies in order to identify normative lessons. Second, it suggests a procedural method for inducing the legal systems of representative developed and developing countries to produce a proxy for transfer pricing case law with public good features. The proposal is the product of six normative lessons that have been inferred from the U.S. and Argentine experiences in the enforcement of the ALS. The proposal is written to facilitate addition to the OECD Model Tax Convention on Income and on Capital, article 9. The suggested proposal is arguably superior to its competition ( i.e., formulary apportionment and consolidated base taxation, as elaborated by the European Commission in October of 2001). This is largely because the procedural proposal, given the relatively minor demands it places on local institutions and human capital endowments, can be applied readily in both the developed and developing worlds. Alternative options lack such versatility. This Article is divided into five parts. After this introduction, Part II outlines the distinction between rules and standards from a law and economics perspective. It shows that the OECD Model is primarily a standard based legal framework. The OECD Model presupposes a decentralized network of domestic courts that are capable of producing case law with public good features ( i.e., case law that provides precise meanings to the model). Finally, Part II explains how this failure to produce case law with public good features is at the root of the transfer pricing problem. Part III explores the normative lessons to be drawn from the Argentine and U.S. experiences in transfer pricing. It concludes that, for different reasons, both the developed and developing worlds are facing the same problem: the meaning of the ALS is largely unknowable because of the absence of case law with public good features in this area. Therefore, the key assumption of the OECD Model, as referred to above, is violated in the transfer pricing area. The result is that the ALS is unable to provide taxpayers with a clear sense of the law’s demands with respect to transfer pricing in the legal systems in which the taxpayers operate. Part IV argues that a representative transfer pricing contest is an unsolved, one-shot prisoner’s dilemma between the tax authority and a taxpayer. Consequently, the parties to a transfer pricing case lack the incentive to provide the courts with all the available information, which are necessary to produce transfer pricing case law with public good features. As a result of this strategic scenario, most current transfer pricing precedents are private goods: they are typically applicable only to the case at hand. Conversely, APAs may trigger an iterated prisoner’s dilemma between contracting states that might be solved spontaneously. Hence, the Article proposes that the bilateral or multilateral APA, provided for in article 25 of the OECD Model, be applied in such a way as to achieve two fundamental goals: (i) to promote the emergence of tit-for-tat between contracting states in the transfer pricing arena, and (ii) to produce a proxy for case law with public good features capable of providing meanings to the ALS in the wide array of contexts in which that standard operates ( e.g., e-commerce and derivative financial instruments). Part V concludes that the procedural proposal is the best of the available options because it applies relatively readily to both developed and developing countries and it is consistent with the present structure of international taxation.