FLORIDA TAX REVIEW VOLUME 7 2006 NUMBER 9 COMMENTS ON THE OECD PROPOSAL FOR SECRET AND MANDATORY ARBITRATION OF INTERNATIONAL TAx DISPUTES by Michael J. Mclntyre* I. INTRODUCTION .......................................... 622 II. EVALUATING THE NEED FOR MANDATORY ARBITRATION ....... 624 A. Potential Costs and Benefits of Mandatory Arbitration .... 624 B. Unsettled Cases of Potential Double Non-Taxation ....... 626 C. Addressing the Rare Cases of Double Taxation .......... 628 III. SECRET ARBITRATION: OECD ENDORSEMENT OF OPAQUENESS . 630 IV. ADDITIONAL FLAWS IN THE OECD PROPOSAL ............... 637 A. Potential Bias ofArbitrators ......................... 637 B. Participation by Non-Party Taxpayers ................. 638 C. Unrealistic Time Limits ............................. 639 D. Lack of Procedural Rules Governing Arbitration ......... 640 E. Some Additional Criticisms .......................... 641 V. CONCLUSION ............................................ 645 * Professor of Law, Wayne State University Law School. Comments on the OECD Proposal I. INTRODUCTION The Organization for Economic Co-operation and Development (OECD) has proposed amendments to its Model Tax Convention and Commentary that would establish a system for the mandatory arbitration of tax disputes between two treaty countries when the tax officials of those countries have been unable to resolve those disputes within a two-year period.' The proposal is undoubtedly well-meaning and does address a small but significant problem - the "rare cases" (as characterized by the OECD)2 of potential double taxation that are unresolved through the existing tax-treaty mechanism.3 Nevertheless, as discussed below, the public policy goals of this proposed system are at best obscure, and the risks to sound administration of national tax systems are great. The OECD's goal seems to be to please the international business community when the goal ought to be to advance the public interest in a transparent and unbiased system. Although the OECD has aggressively sought comments from the international business community on its proposal and has redesigned its system in accordance with those comments, it has been far less aggressive in obtaining comments from academics and other people more likely to represent the public interest. Moreover, the many developing countries not within the OECD's orbit have played little or no role in the development of this proposal. The OECD Proposal includes the following three features that have been on the wish lists of multinational companies for a very long time: (1) A forum outside the control of the tax authorities where they can litigate tax disputes in secret; (2) A club they can use to compel the tax authorities to resolve international tax disputes within very tight time deadlines (typically six months); and 1. See OECD, Proposals for Improving Mechanisms for the Resolution of Tax Treaty Disputes, (2006), http://www.oecd.org/dataoecd/5/20/36054823.pdf(hereinafter "OECD Proposal"). 2. Id. at 6, 91, 45 ("This paragraph provides that, in the rare cases where the competent authorities are unable to reach an agreement under paragraph 2, the unresolved issues will, at the request of the person who presented the case, be solved through an arbitration process."). 3. See OECD, Articles of the Model Convention with Respect to Taxes on Income and on Capital (2005, http://www.oecd.org/dataoecd/50/49/35363840.pdf (providing that the contracting states shall endeavor to resolve tax disputes arising under the treaty) (hereinafter "OECD Model Tax Convention"). 2006] Florida Tax Review (3) Direct involvement by their legal staffs in the competent-authority process. No wonder the arbitration system proposed by the OECD has the strong support of the international business community. The draft proposal barely acknowledges the features listed above and makes no serious effort to defend them. In particular, the OECD fails to explain why a system of dispute resolution that contained these features would advance the public interest. All of these features are ones that many OECD member states have opposed in the past on the ground that they are inconsistent with their national sovereignty.4 The OECD apparently believes that it is in the last stage of developing a robust and workable proposal for resolving international disputes in an appropriate manner. At this point, it seems to be interested simply in fine tuning its proposal to make it even more user friendly to the international business community. This apparent belief is unfounded. The proposal ought to be examined, perhaps for the first time, in terms of its contribution to the public good. In these brief comments, I seek to provoke such an examination. Section II, below, examines the argument that adopting mandatory arbitration is needed in order to guarantee that international tax disputes under tax treaties get resolved. I argue in that section that the costs of resolving all cases through arbitration may exceed the benefits. In many cases, the benefits are modest in that the taxpayer is not seeking to avoid double taxation but instead is seeking to make use of flawed tax-treaty rules to avoid taxation in both of the treaty countries. Section III, below, argues that the secrecy in the arbitration process contemplated in the OECD Proposal is contrary to public policy and is inconsistent with the OECD's strong support in other contexts for transparency in international tax matters. I suggest that the arguments against blanket secrecy are so strong and so obvious that the OECD appears to be acting in the interests of the multinational companies rather than in the public interest. Section IV provides a brief discussion of a variety of other flaws in the OECD Proposal. Some of these flaws are relatively minor and easily fixed. 4. If the legislative body in a country concludes that the OECD Proposal infringes inappropriately on national sovereignty or otherwise contravenes public policy, it is unlikely to approve a tax treaty that embodies that proposal. Several U.S. Senators, in a letter to the U.S. Secretary of the Treasury, recently expressed their concern about various features of the OECD Proposal, including its provision for a totally secret proceeding. The letter cites and attaches my comments provided to the OECD. See Letter from Senators Byron L. Dorgan, Russ Feingold, and Carl Levin to Treasury Secretary John W. Snow, May 16, 2006. [Vol. 7:9 Comments on the OECD Proposal Others are more fundamental. A conclusion, with some recommendations for modification in the OECD Proposal, is provided in Section V. II. EVALUATING THE NEED FOR MANDATORY ARBITRATION In the materials the OECD has distributed for public comment, the OECD has not explained why it believes that arbitration is an appropriate mechanism for resolving disputes in the "rare cases" in which the appropriate tax officials of the treaty countries (the "competent authorities") are unable to resolve those disputes on their own. The OECD seems to believe (1) that some resolution of the unresolved disputes is needed, and (2) that secret and mandatory arbitration is the only available alternative for resolving them that would be acceptable to the international business community. This practical approach of accommodating the international business community may have some appeal to beleaguered international bureaucrats. It is not a sensible basis, however, for designing a coherent dispute resolution system that will obtain public legitimacy. In subsection H,A, below, I look at the costs of attempting to resolve through mandatory arbitration the relatively few cases that currently are not being resolved by domestic courts or by the competent authorities. Subsection II,B explains why a significant percentage of the unresolved cases are likely to involve situations in which the taxpayer is seeking to avoid taxation in both of the countries that are parties to the dispute. Subsection II,C discusses those cases in which the taxpayer is facing a significant risk of double taxation. I suggest that many of these cases are likely to be transfer-pricing cases - cases involving the proper price to charge on transactions between related persons or between branches of a single entity. I argue, inter alia, that the proper solution to transfer-pricing problems is to improve the OECD's transfer-pricing rules, which currently do not yield definitive answers in almost all cases involving the sharing of intellectual property. A. Potential Costs and Benefits of Mandatory Arbitration Resolution of all tax-treaty disputes is not necessarily worth the costs that the attainment of that goal would entail. If the OECD is correct that the cases of non-resolution are rare, then the better part of wisdom may be simply to declare victory. Of course, the competent-authority mechanism is needlessly opaque; consequently, those of us on the outside do not have the data needed to evaluate the OECD claim that non-resolution is rare. Because the people in position to challenge that claim do not appear to be objecting to it, I am assuming for purposes of this report that the OECD claim is well-founded. No dispute resolution system is perfect, nor can it be expected to be perfect. If perfection could be achieved at little or no cost, then the OECD's 20061 Florida Tax Review purported goal of perfection (resolving all disputes) would make lots of sense. But the OECD has developed a complex, expensive system that undermines national sovereignty and presents serious risks of corruption and unfair dealing. This endeavor is not cost-free. Before embarking on a search for perfection, the OECD should have examined the costs of its proposal and compared them to the costs of living with imperfection. One important negative consequence of the OECD Proposal is that it would give multinational companies an opportunity to bypass domestic courts and still get their case adjudicated by a tribunal independent of the tax authorities. Under the OECD Proposal, only cases that bypass domestic courts would go to arbitration because a matter already decided by a domestic court is not subject to arbitration.5 Bypassing domestic courts is a serious matter, not to be permitted without compelling reason. Many tax-treaty issues depend on an interpretation of domestic law, and that interpretation ought to come from domestic courts.6 The OECD seems to recognize that an international tribunal should not act as a court of review for the decisions of domestic courts. Yet, it is prepared to offer multinational companies an opportunity to supplant the domestic courts with a secret tribunal of their own choosing. Another negative consequence is the potential for corruption presented by a secret adjudicative system. In a democratic society, the basic check on the integrity of the judiciary is transparency. The OECD would eliminate that check under its proposed secret system. In private arbitration, some check on the integrity of the system is provided by the strong self-interest of the parties. That self-interest is decidedly less strong when the money at stake is the people's money, not the private fortunes of the parties. Corruption is always a concern within a tax department, and well-run departments have strict internal procedures to combat it. Those internal procedures cannot work effectively to police decisions made outside the department. I do not suggest that corruption is inevitable under the OECD Proposal, only that the possibility is cause for serious concern. If the OECD believes that resolution of all treaty disputes referred to the competent authorities is necessary for some reason, it should not favor its 5. See OECD Proposal, supra note 1, 14, 51, 53, 62. Ideally, an arbitration system should operate only if the taxpayer has agreed to be bound by it. Achieving that ideal system is difficult in countries that do not allow for the waiver of court remedies. See Hugh J. Ault, "Improving the Resolution of International Tax Disputes," 7 Fla. Tax Rev. 137, 146-47 (2005) (hereinafter "Ault"). Professor Ault is a senior advisor to the OECD Centre for Tax Policy and Administration. 6. The OECD suggests that the arbitration panel might refer certain matters of domestic law to some other arbitration panel having competence in local law. See 91 of OECD report. The likely effect of such referral would be to prevent the arbitration panel from reaching a final resolution of the matters before it. [Vol. 7:9 Comments on the OECD Proposal own proposal. That proposal, by its own terms, will provide for resolution of a dispute only when the taxpayer, at its sole discretion, elects to invoke arbitration. I assume that the OECD did not propose to make arbitration mandatory in all cases because it did not want to provoke opposition from taxpayers. That solicitude for the interests of taxpayers is both unwise and unwarranted. It is unwise because it gives taxpayers one more lever to game the system. It is unwarranted because taxation is not about doing nice things to please taxpayers. If treaty partners have an interest in resolving their treaty disputes, as the OECD alleges, then they should not refrain from doing so simply to please some disgruntled taxpayers. As suggested above, the potential costs of perfection - resolving all competent-authority cases - are high. In contrast, for reasons discussed below, the costs of tolerating some imperfection in the competent-authority mechanism may be fairly small. Not knowing the particulars of any of the actual cases that do not get resolved, I obviously cannot make anything close to a full assessment. Still, I can suggest a general framework for analysis. B. Unsettled Cases of Potential Double Non-Taxation The non-resolved competent-authority cases necessarily fall into two categories: (1) cases of potential double taxation, and (2) cases of potential double non-taxation. I suspect that the majority of the unresolved cases fall into the second category. I have no data on unresolved cases, of course, because the whole competent- authority process is secret. But I have known a lot of tax administrators over the years. In general, the ones I have known do not like double taxation, and they do not like double non-taxation. So, it seems plausible to me that they generally would be inclined to compromise with their counterparts to settle double taxation cases but might occasionally dig in their heels when asked to facilitate double non-taxation. Double non-taxation cases are themselves common and are often the goal of sophisticated tax planning. As an example, assume that Country A exempts capital gains and Country B does not. Country B, however, has a tax treaty with Country A that exempts some but not all capital gains earned in Country B by a resident of Country A.7 The taxpayer, a resident of Country A, 7. See OECD Model Tax Convention, supra note 3, Art. 13, 5 ("Gains from the alienation of any property, other than that referred to in paragraphs 1, 2, 3 and 4, shall be taxable only in the Contracting State of which the alienator is a resident."). 2006] Florida Tax Review earns a capital gain of $100 million in Country B, which it claims is exempt from tax in Country B under the tax treaty. The tax officials of Country B disagree and have a reasonable basis for that disagreement. The taxpayer asks the competent authorities of Country A to intervene on its behalf, claiming that taxation by Country B is "not in accordance with" the treaty.' If the matter were to go to arbitration and the taxpayer were to win, the result would be international double non-taxation. No one would contend that taxpayers should be denied treaty benefits to which they clearly are entitled on the ground that granting the benefit would result in tax avoidance. If a taxpayer believes that it is entitled to a treaty benefit and the government believes otherwise even after consultation with its treaty partner, the taxpayer has the right to litigate the matter in the domestic courts. The question is whether the taxpayer should be given the additional option of bypassing the domestic courts and bringing the matter to an international arbitration panel. I think that the OECD should require a compelling reason for giving an affirmative answer to that question. Facilitating international tax avoidance does not strike me as a compelling reason. The example above illustrates one of the major flaws of the OECD Model Tax Convention - its unfortunate role in promoting international double non-taxation. A model tax treaty designed to prevent international tax avoidance should provide that the source country would not relinquish its right to tax unless the residence country, in fact, was exercising its right to tax. The proper long-term solution to that double non-taxation problem is not to devise an arbitration system that guarantees taxpayers the right to avoid taxes even in disputed cases. The better course of action is to revise the model treaty to prevent it from presenting taxpayers with opportunities for double non-taxation. The United Nations Model Tax Convention,9 developed to take account of the economic, political, and social circumstances of developing countries, is the only significant competitor to the OECD model. Although it is based in part on the OECD model and has incorporated many of its flaws, it does limit to some degree the opportunities for double non-taxation provided by the OECD model. For example, its default rule is that income not addressed by the treaty remains taxable in the source country, whereas the OECD model provides that such income is not taxable in the source country and may not be taxable anywhere. 8. See OECD Model Tax Convention, supra note 3, Art. 25, 1. 9. United Nations Model Double Taxation Convention between Developed and Developing Countries, ST/ESA/PAD/SER.E/21 (2001) http://daccess-ods.un.org/ access.nsf/Get?OpenAgent=&DS=ST/ESA/PAD/SER.E/21 &Lang=E (last visited, October 16, 2006). [Vol 7:9 Comments on the OECD Proposal Many of the double non-taxation cases that arise under tax treaties probably are transfer-pricing cases - that is, cases involving a controversy over the prices charged by the taxpayer in dealings with related persons. The OECD Proposal indicates that some of the unresolved cases are transfer-pricing cases, although it offers no details. A transfer-pricing case could result in double non-taxation when a high-tax country has a tax treaty with a low-tax country. In such circumstances, a resident of the low-tax country would be inclined to allocate as little income as possible to the high-tax country and as much income as possible to the low-tax country. The tax officials in the high-tax country might disagree with that allocation, and if the tax officials in the low-tax country supported the position of the taxpayer in a competent-authority proceeding, there is at least some chance that the matter would not get resolved. The public policy cost of a failure to resolve such a conflict does not appear to be significant. C. Addressing the Rare Cases of Double Taxation The OECD, in its discussion of its arbitration proposal, totally ignores the issue of double non-taxation. It seeks to defend its arbitration proposal as a method of preventing double taxation. In general, eliminating such double taxation is a good thing. Not all good things, however, are worth the costs of attaining them. One policy question is whether the elimination of double taxation in the rare cases in which it occurs is a good enough thing toj ustify the costs inherent in the OECD's Proposal. Another policy question is whether some other approach might be taken that would address the problem at lower cost. Neither of these policy questions can be answered fully without some information about the rare cases that are not being settled by the competent authorities. As already noted, that information is not being shared with the international tax community. My best guess, nevertheless, is that many of the unsettled cases are transfer-pricing cases. Anyone familiar with the transfer-pricing rules, which are promoted with great enthusiasm by the OECD, would anticipate that they would result in some conflicts between taxpayers and tax officials and between the tax officials of different countries. The simple fact is that the transfer-pricing rules are broad guidelines, similar in some respects to the broad guidelines that accountants receive from various accounting-standards boards. The end result is that the applicable "law" is indeterminate - it is less like law and more like an invitation to negotiate. The official transfer-pricing guidelines promulgated by the OECD favor three transactional methods - the comparable uncontrolled method, the resale-price method, and the cost-plus method. These methods, by their own terms, rarely are applicable to transactions involving intellectual property. 2006] Florida Tax Review Because many multinational companies earn most of their income from the exploitation of intellectual property, those transactional methods have limited applicability. The other transfer-pricing methods, begrudgingly endorsed in the OECD guidelines, can be applied in more cases. Their application, at best, gives a range of possible transfer prices, not a definitive arm's length price. Moreover, the taxpayer is not compelled to pick any particular method and has discretion to invent its own method in many circumstances. In this legal setting, disputes over transfer prices are inevitable. And the stakes are high, sometimes measured in the hundreds of millions of US dollars."0 The OECD finding that unresolved competent-authority cases are rare is exceptionally good news, given the fundamental flaws in the OECD's transfer-pricing rules. Still, the OECD wants to eliminate the remaining unresolved cases through its arbitration proposal. As noted above, transfer-pricing cases present problems under tax treaties because the basic rules are indeterminate. The preferred long-term solution is not to make up numbers out of whole cloth. It is to fix the rules. I will not discuss here the many methods that might be employed to reform or replace the transfer-pricing rules." The basic solution is to provide unambiguous default rules that would apply whenever the other rules would produce an indeterminate result. An alternative solution, attractive in the short-term, is simply to leave unresolved the rare cases that are not resolved through the competent-authority procedure. The result would be some possibility of unfairness. No system, however, can expect to eliminate all possibilities of unfairness. Moreover, the risk of serious unfairness is small. Multinational companies that play clearly within the rules, without engaging in aggressive tax avoidance, are not likely to end up with unrelieved double taxation due to their choice of transfer prices. It is mostly the companies that play the transfer-pricing game aggressively that find themselves at risk. They undoubtedly have calculated that they end up better off by running the 10. The well-publicized case of GlaxoSmithKline PLC, the British pharma- ceuticals giant, illustrates the huge amounts that can be at issue in a transfer-pricing case. The case, involving tax years 1989 to 2005, was settled in 2006, with Glaxo agreeing to pay the IRS $3.4 billion. The U.K. did invoke the competent-authority procedure in the U.S.-U.K. tax treaty. It does not appear, however, that Glaxo claimed that it was subject to double taxation. Robert Guy Matthews and Jeanne Whalen, "Glaxo to Settle Tax Dispute With IRS Over U.S. Unit for $3.4 Billion," Wall Street Journal, Sept. 12, 2006 at A3. 11. The preferred solution is to abandon the arm's length method entirely. See Michael J. McIntyre, "The Use ofCombined Reporting byNation States," 35 Tax Notes Int'l 917 (Sept. 6, 2004). Many less radical steps can be taken, however, to reform the transfer-pricing rules. [Vol. 7:9 Comments on the OECD Proposal risk of the occasional unrelieved double tax in order to maximize their opportunities for double non-taxation. I am prepared to assume that these very clever people, who have millions of tax-planning dollars at their disposal, have figured the odds correctly. In most transfer-pricing cases, a failure of the competent authorities to reach agreement does not leave the taxpayer without a remedy. The taxpayer can always go to court in the country assessing the allegedly unfair tax and challenge the assessment. It is possible, of course, that the domestic courts in the two countries will produce inconsistent outcomes, ultimately resulting in unrelieved double taxation. From reading many transfer-pricing cases, however, I think that the risk is a modest one. Transfer-pricing cases are exceedingly difficult cases for the government to win, for a variety of reasons. Among those reasons are that tax departments are typically "outgunned," they have difficulty discovering relevant facts under the taxpayer's control, and they have to contend with rules that were designed to give maximum flexibility to the taxpayer. Il. SECRET ARBITRATION: OECD ENDORSEMENT OF OPAQUENESS The single most objectionable feature of the OECD Proposal is its provision for total secrecy. Even the names of the taxpayers engaging in the arbitration proceedings would be kept secret. No information would be released about the results of the arbitration or the basis for the decision of the arbitrators without the expressed, written consent of the affected taxpayers. Even with taxpayer consent, which can be expected to be withheld in many cases, only a bare outline of the case is likely to be released to the public. The taxpaying public, nevertheless, would be asked to pay all of the costs of the arbitration proceeding. 2 12. Tax officials in the U.S. and Germany recently negotiated a protocol that would amend the U.S.-Germany treaty to provide for arbitration. The arbitration is not fully mandatory in that the parties by mutual agreement may decide not to arbitrate a particular matter. The proposed protocol provides for full secrecy of the proceedings and does not require the arbitrators to prepare a report. Instead, the arbitrators simply pick between the settlement proposals made by the two parties. The outcome of the arbitration, however, is disclosed only to the concerned taxpayer and its representatives, again under a requirement ofconfidentiality. Only enumerated issues, including transfer- pricing disputes, are subject to arbitration. See Protocol Amending the Convention Between the United States of America and the Federal Republic of Germany for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and Capital and to Certain Other Taxes, June 1, 2006. Available at http://www.treas.gov/press/releases/reports/germanprotocolO6.pdf(last visited Oct. 17, 2006). 2006] Florida Tax Review In recent years, the OECD has been a champion for transparency in international transactions. A much bloodied champion, I might add. Its efforts to open up the secret books of financial institutions that have been enabling tax fraud have brought it under attack in many countries. 3 So, this step back into opaqueness is an unwelcome development. The reason for it, however, is not difficult to divine. My best guess, with no insider information, is that early drafts of the OECD Proposal provided for reasonable disclosure. After all, the OECD surely understands that an arbitration system that is viewed by the international tax community as a "black box" will not achieve much credibility and will be useless in providing guidance in related cases. It should be aware of the substantial risks of fraud from a secret procedure. And the obvious model for the OECD to look to in fashioning an arbitration procedure for international tax matters is the highly-regarded arbitration system developed by the World Trade Organization (WTO). The WTO arbitration procedures deal with issues at least as sensitive as those to be addressed in resolving international tax disputes. The WTO procedures are undoubtedly more refined than would be required for dealing with the rare cases that do not settle through the competent-authority mechanism. These procedures are particularly noteworthy, nevertheless, for providing a commendable degree of transparency. I certainly recognize that multinational companies love secrecy. They pressured the U.S. Internal Revenue Service (IRS) to keep its Advanced Pricing Agreements secret, although the law required that redacted version of those agreements be made available to the public. 4 I do not know whether the OECD 13. For example, a contingent of lobbyists representing certain U.S. business interests recently recommended to President George W. Bush that the United States eliminate its financial support of the OECD as punishment for the OECD's efforts at curtailing tax-haven abuses. See Coalition for Tax Competition, Coalition for Tax Competition Urges White House to Defund the Paris-based OECD, at http://www.freedomandprosperity.org/press/p02-09-06/ p02-09-06.shtml (Feb. 9,2006). 14. The IRS litigated the issue, resulting in a decision against it and in favor of transparency. See Bureau of Nat'l Affairs, Inc. v. Comm'r, 24 F. Supp. 2d 90, 93 (D.D.C. 1998). The multinational companies then launched an aggressive lobbying campaign that ultimately led a subservient Congress to overtum the courts and provide total secrecy to Advance Pricing Agreements. See IRC ยง 6103(b)(2)(C). For discussion of the importance of transparency in the APA process, see Michael J. McIntyre, "IRS Affirms Plans for Developing Secret Tax Law on Transfer Pricing," 3 Tax Notes Int'l 267 (Mar. 1991); Michael J. McIntyre, "What's in a Name?" 3 Tax Notes Int'l 29 (Jan. 1991); and Michael J. McIntyre, "The Case for Public Disclosure of Advance Rulings on Transfer Pricing Methodologies," 2 Tax Notes Int'l 1127 (Nov. 1990). See also Joel D. Kuntz & Robert J. Peroni, 1 U.S. International Taxation A3-275 (2005) ("Judging by the fear of sunshine, it seems quite possible that this body of secret law contains some special treatment for certain taxpayers."). For an unnerving account of the abuses of a [Vol. 7:9 Comments on the OECD Proposal staff simply anticipated the opposition of the multinationals and compromised in advance or gave ground bit by bit. In any event, the victory for the multi- nationals has been complete. The normal starting point, in designing an international arbitration system for public institutions, would be total transparency. Following the WTO model, the hearings should be open to observers, the submissions of the parties should be available publicly, and the decision of the deciding body should be available publicly. By the WTO standard, the OECD Proposal is wrong in every particular. An alternative model that the OECD undoubtedly looked at is the EU Arbitration Convention, ," which applies only to transfer-pricing disputes. That convention provides for a secret procedure, with the report of the arbitration panel to be published only with the consent of the taxpayer. I find this model to be objectionable for the same reasons I object to secrecy in the OECD model. Few cases have been decided under the EU Arbitration Convention. 6 Taxpayers are not entitled under the convention to obtain the benefits of double non-taxation. 7 That limitation may explain, in part, its lack of use. As noted above, nothing - not even the names of the affected taxpayers - can be made public under the OECD Proposal without the consent of the taxpayer. The taxpayer, of course, is a non-party; the parties to the arbitration are the affected governments. Yet, the credibility of the arbitration procedure is held hostage to the desire of the non-party for total secrecy. The rest of the secrecy rules are of limited importance, once the taxpayer is given total control over the release of information. The fact that these additional rules are even in the proposal has led me to speculate that secret tax law, see Martin Lobel, Lee Ellen Heifrich, Henry M. Banta, and JeanAne M. Jiles, U.S. Transfer Pricing and Oil Royalties: A Cautionary Tale, 19 Tax Notes Int'l 177 (July 12, 1999). 15. The official name of that convention is Convention on the Elimination of Double Taxation in Connection with the Adjustment of Profits of Associated Enter- prises. It was signed in 1990, went into effect for a five-year period at the start of 1995, and lapsed at the end of 1999. It was renewed retroactively from 2000 to 2004 and then lapsed again. In 2004, it was renewed and extended to new member states of the EU. 16. See EU Joint Transfer Pricing Forum, Report on the Re-entry into Force of the Arbitration Convention, DOC: JTPF/019/REV5/2004/EN (May 30, 2005) at p. 4, 14 (reporting that only two cases have been referred to an arbitration panel and only one case, between France and Italy, has been decided by a panel). 17. See EU Arbitration Convention, Article 14 (providing that double taxation is considered to be eliminated if the profits of the taxpayer are included in the computation of income in one contracting state only or a credit is given in one state for the taxes imposed in the other contracting state). 2006] Florida Tax Review earlier drafts contemplated fuller disclosure. Because the taxpayer can prohibit the release of any information, the information that might trickle out with the taxpayer's consent is unlikely to be the least bit interesting to the international tax community. Still, it is telling that the OECD is unwilling to foster transparency even with the consent of the taxpayer. All that can be released even with consent is a bare bones report that does not disclose the name of the concerned taxpayer or any identifying information about the taxpayer. With taxpayer consent, a sterile summary of the legal issues, totally useless in a transfer-pricing case without the relevant facts, can be released. Surely, the OECD contemplated something more useful when it formulated its initial proposal. I am truly astonished that the OECD believes that the taxpayer has a legitimate interest in maintaining the secrecy of its own identity. Even the client-lawyer privilege does not protect a lawyer from having to disclose the identity of his client. The OECD is acting as if a public arbitration procedure, financed with public funds and charged with the obligation to decide major issues of public policy, is actually some kind of internal settlement procedure within the tax department. As a result of that flawed perspective, it has promulgated a proposal that is completely inconsistent with the public nature of the arbitration procedure it hopes to establish. I cannot help thinking that the OECD, in endorsing a secret procedure, is acting under heavy pressure from the multinational companies. I do not discount the possibility, however, that some pressure is also coming from member states. Government officials do not always welcome public scrutiny of their work. They may feel that such scrutiny not only may be embarrassing in some cases but also may inhibit them from cutting the kinds of deals that they feel they must make to manage their case load. In addition, there are some important OECD countries where the multinational companies are particularly influential. Given the potential bias of tax officials and the international business community in favor of secrecy, they should not be the only people at the table when the degree of secrecy in international tax arbitrations is being decided. The transparency of international tax arbitration is a public policy issue that ought to be decided after full public debate. The OECD is to be commended for inviting public comment on its proposal. The challenge now is to ensure that the desirable debate actually occurs and that the final proposal of the OECD embodies the results of that debate. An argument often made for secrecy in arbitration is that the publication of the details of a multinational company's international tax dispute would provide an advantage to the companies with which it competes. This claim is mostly bogus. There occasionally may be some sensitive material, and some measures might be taken to sanitize it. As Holden Caulfield would say, however, most of the material that would enter the public domain is as sensitive [Vol. 7:9 Comments on the OECD Proposal as a toilet seat. The reason is that the material is likely to be hopelessly dated by the time it needs to be released in a tax dispute. Tax disputes ripe for arbitration typically are not about tax issues for the current year, or the prior year, or the year before that. A typical transfer- pricing case for a multinational company may involve transactions that occurred many years ago. For example, Bausch & Lomb,"8 the famous U.S transfer-pricing case, was decided by the Tax Court in 1989. It dealt with an assessment issued in 1985 with respect to tax years ending between 1979 and 1981. The 1980s were the good old days; the lag between the year of the transaction and the resolution of the dispute in court has increased significantly in the last decade, at least in the United States. Although speculating about motives is a bit risky, I strongly suspect that the aversion of multinational companies to public litigation has far more to do with its concerns for public relations, push-back from tax reformers, and audit exposure in third countries than any concerns about competitive advantage. These fears may be rational. After all, the disclosures coming from the Bausch & Lomb case helped shape new transfer-pricing regulations in the United States, and that regulation project gave impetus to the development of the OECD transfer-pricing guidelines. The OECD, however, should not seek to accommodate these fears of bad publicity, political backlash, and third-country audit exposure. On the contrary, the possibility of these consequences highlights the importance to public welfare of a transparent procedure for resolving public tax disputes. The reason for the OECD's Proposal for secrecy is not craven subservience to the multinational corporations. It is a concern that an open procedure would be shunned by the multinational companies. That concern is well-founded. Indeed, I strongly suspect that the major motive of the multinationals in promoting arbitration is to obtain a forum outside the control of the tax departments for resolving international tax disputes in secret. If the multinational companies were willing to make their case in public, they would do so in the domestic courts of the country assessing the challenged tax. 9 The basic issue for the OECD to decide is a simple one. Should the multinational companies be given a secret forum in which they can litigate their tax disputes? The clear answer, from a public policy perspective, is "no." If that answer means that the OECD's arbitration initiative ends up having no 18. Bausch & Lomb, Inc. v. Comm'r, 92 T.C. 525 (1989), aft'd, 933 F.2d 1084 (2d Cir. 1991). 19. Public litigation of complex transfer-pricing cases has virtually ceased in the United States. To the best of my knowledge, no major transfer-pricing case dealing with rich factual issues has been litigated in the United States since the publication of the revised transfer-pricing regulations in 1994. 2006] Florida Tax Review caseload, so be it. The OECD frankly acknowledges that the cases not resolved under the existing mechanism are rare. As discussed above, it is highly unlikely that grave injustice is occurring with any frequency even in those rare cases. In the exceedingly rare case that presents the potential for grave injustice, a public procedure could serve as a useful safety value. Citizens have a right to know what their government is doing in their name. Everyone understands the need for secrecy in some cases. The public's right to know public business is not absolute. In addition, taxpayers do have legitimate privacy rights in settling their tax obligations, although the legitimate rights of public corporations to privacy are frequently overstated. Those privacy rights obviously are not absolute. They end when the taxpayer challenges the determination of the government in an independent forum. That is the universal rule for domestic courts, and that rule is fundamental to the legitimacy of the decisions reached by those courts. The rationale for that rule is even stronger when the independent adjudication is undertaken by an international body, due to the greater problems that such a body is likely to have in achieving legitimacy. Defenders of secrecy are likely to argue that secrecy is a normal feature of an arbitration procedure. It is certainly true that most domestic arbitration proceedings are secret. The analogy to domestic arbitrations, however, is inappropriate.2" Domestic arbitrations are typically between private parties, the costs of the arbitration are privately financed, and the issues being resolved are private disputes. In sharp contrast, the OECD is proposing an international body that would be charged with the responsibility of deciding the amount of tax due to sovereign states. That dispute is a public dispute, the costs of resolving it are charged to the public, and the parties to that dispute (the governments) are public bodies accountable to their citizens. Those who find force in the analogy between the OECD's Proposal and private arbitration might consider the factors typically considered in deciding whether private arbitration is an appropriate mechanism for resolving a particular dispute. I list below the factors that specialists in dispute resolution 20. See William W. Park, "Income Tax Treaty Arbitration," 10 George Mason Law Rev. 803, 823 (2002) ("Without exception, all major institutional rules for international commercial arbitration (ICC, LCIA, AAA International, UNCITRAL, ICSID and Geneva Chamber of Commerce) require arbitrators to state the grounds for their decision unless the parties explicitly opt out of a reasoned award... .On balance, tax arbitration probably should require arbitrators to explain themselves. While this will make their job harder, and in practice mean exposure to a greater degree of judicial scrutiny, the end product will be a better decision.") (hereinafter "Park"). [Vol. 7:9 Comments on the OECD Proposal typically would take into account in deciding whether to recommend private arbitration.2 1. Neither party to the dispute has an interest in a public resolution of the dispute that would have value as a precedent in resolving similar disputes. 2. Arbitration would reduce the risk for one or both parties of an unpredictable and catastrophic result. 3. The case in arbitration would be adjudicated by people better able to handle difficult technical matters. 4. One or both parties needs to be able to control case-scheduling issues and would be able to do so more effectively in the arbitration proceeding. 5. A complete and final resolution of the dispute can be achieved in arbitration and cannot be achieved otherwise within a reasonable time. 6. One or both parties wishes to limit discovery in the case. 7. The costs of arbitration are lower than the costs of the alternatives. I If all or most of these factors are present, then arbitration is likely to be appropriate, whereas if many of the factors are not present, arbitration is inappropriate. I think it fair to say that all or most of these factors are not present for the international arbitration proceedings contemplated in the OECD Proposal. In many cases, the two countries would welcome a public precedent that interpreted the provision of a tax treaty. The OECD proposes that the arbitration decision would not even affect the resolution of an identical issue with the identical taxpayer in a subsequent year.22 The parties to an OECD-type arbitration are not seeking to avoid risk, such as the risk of a high damage judgment in a tort case - indeed, the OECD's arbitration procedure is more likely to present special risks. Although the arbitrators are likely to be competent technically, so also are the likely adjudicators in any alternative proceeding. Case scheduling is likely to be more difficult in arbitration, given the tight deadlines and the likely scheduling problems of busy arbitrators. The OECD does make some effort to provide 21. See Jay Folberg et al, Resolving Disputes: Theory, Practice, and Law, Aspen (2005) at 460-461 (drawing on guidelines for arbitration prepared by the CPR Institute for Dispute Resolution). 22. See OECD Proposal, supra note 1, 57. 2006] Florida Tax Review finality to an arbitration result, by requiring the taxpayer to agree to forgo domestic appeals. The OECD procedures, however, clearly anticipate that many arbitrations would not give finality. Indeed, it would appear that only a small part of a taxpayer's tax liability might be at issue in the arbitration, and no requirement is provided for limiting counterclaims or collateral defenses. Limiting discovery, which is often an issue in the context of U.S. private litigation, is unlikely to be a significant reason for seeking international arbitration in a tax case. Finally, the costs of an OECD-type arbitration almost certainly would be significant in any transfer-pricing case. High costs are almost guaranteed as a result of the OECD Proposal to allow the taxpayer to participate actively in the proceedings. IV. ADDITIONAL FLAWS IN THE OECD PROPOSAL The total lack of transparency in the OECD's proposed procedure is sufficient to justify strong opposition to it. Even if the secrecy problem is addressed and fully rectified, however, the proposal is still far from acceptable. Fortunately, most of the remaining flaws are more easily corrected. A. Potential Bias ofArbitrators One serious flaw, easily corrected, is the failure to provide measures designed to guarantee the impartiality of the arbitration. For example, the arbitrators are not required to disclose possible conflicts of interest, or appearances of such conflicts, nor are they required to refrain from engaging in remunerated activities on behalf of the taxpayer after the conclusion of the arbitration. Each side is expected to name one arbitrator, typically from its own tax office. Neutrality is not expected from that arbitrator. The proposed Commentary states that the third arbitrator should "act in total neutrality and independence."23 It provides no mechanisms, however, for ensuring such a result or for providing redress if that result is not obtained. Indeed, the rules proposed by the OECD would not prevent the OECD or the parties from appointing tax counsel for the taxpayer to serve as the "neutral" arbitrator. Although such a result is unlikely in the extreme, the absence of even minimal anti-conflict rules would undermine public confidence in the system. In private arbitration, conflict issues are generally managed by requiring full disclosure by a potential arbitrator of all real or apparent conflicts.24 The basic idea is that the parties have elected to arbitrate and have 23. OECD Proposal, supra note 1, 70. 24. The seminal case in the US requiring disclosure of all apparent conflicts of interest is Commonwealth Coatings Corp. v. Cont'l Casualty Co., 393 U.S. 145 (1968). [Vol 7:9 Comments on the OECD Proposal full autonomy to decide by contract how they wish to proceed. As long as they receive the relevant information, they are free to act as they wish. They may decide, for example, to run some extra risk of having a biased arbitrator in order to obtain the services of an arbitrator knowledgeable about their business. Full disclosure of conflicts is certainly a minimum for an international tax arbitration. The OECD, nevertheless, does not mandate any such disclosure. Full disclosure, even if mandated, would not be sufficient. The arbitrators in the OECD Proposal are expected to do the public's business, not the private business of the non-party taxpayer. The public interest needs to be protected by imposing strict conflict rules, buttressed by enforcement measures designed to give them bite. In particular, it is important that the arbitrators have no significant prior financial dealings with the taxpayer and that they agree not to have any such dealings for a significant period after the arbitration is concluded. To minimize potential conflicts of interest and the appearance of such conflicts, it may be best to exclude tax advisors working for major law firms or accounting firms from serving as the neutral arbitrator. Taxpayers undoubtedly would prefer a representative from the private sector to serve as the deciding arbitrator. Private-sector arbitrators, however, are unsuitable in many cases because they or their partners and associates are unlikely to be able to avoid all potential conflicts of interest. In addition, many potential arbitrators from the private sector are likely to view the taxpayer as the real party to the dispute and may seek a solution that focuses too heavily on the concerns of the taxpayer and too lightly on the concerns of the actual parties. In principle, the parties could select as the neutral arbitrator a person who is not expert on international tax matters but is experienced in running an arbitration proceeding. It is unlikely, however, that most governments would be willing to trust a complex international tax case to a professional arbitrator, whatever the merits of doing so might be. If the parties conclude that expertise in international tax is required, they probably should pick as the neutral arbitrator a tax official or an academic specialist from some third country. B. Participation by Non-Party Taxpayers Another serious flaw in the OECD Proposal is the provision allowing the taxpayer to submit a brief and to participate, with the permission of the arbitrator, in the actual proceedings. Non-parties to an arbitration proceeding should be treated as non-parties. This provision again suggests that the multinational companies are using the OECD as a wedge to obtain a secret and quasi-judicial hearing of their tax dispute outside the control of the tax departments of the taxing states. This backdoor creation of a special forum is both unwarranted and highly dangerous. The only valid goal of the OECD's arbitration procedure is to force the parties to the treaty to resolve their dispute 2 006] Florida Tax Review on how the treaty should be applied in the case before them. They do not need a non-party to direct them how to do so. And no country should be asked to run the risk that its court system will be bypassed in favor of some secret international forum. If the non-party taxpayer is allowed to participate in the arbitration proceeding, its legal staff may end up dominating that proceeding in many cases, simply due to the resources the taxpayer will be able to bring to bear on the matter.z5 It may be precisely for that reason that the multinational corporations are pressuring the OECD to accede to their demand to be treated functionally as a party to the litigation. The informality and collegiality that ought to prevail in such a hearing would be lost. The proper rule should be that the non-party taxpayer is treated as an observer. If third party observers are permitted to present briefs or participate in the proceedings, the same rights should be extended to the taxpayer. Otherwise, the taxpayer should be allowed to participate only through its correspondence with the competent authority of its country of residence. C. Unrealistic Time Limits Under the OECD Proposal, a case is sent to arbitration if the competent authorities do not resolve it themselves within two years, and the arbitrators are required to decide the case before them within six months.26 If the arbitrators do not decide the case within the 6-month time limit, the parties may extend the time limit for another six months or replace the arbitrators.27 These time limits are realistic in some cases but are unrealistic for complex transfer-pricing cases. The interest of the OECD in imposing tight time limits for a decision by the arbitrators is understandable. In this one respect, the OECD has followed the WTO model. The arbitrators in a WTO panel hearing, however, are facing a substantially different situation than the one that the arbitrators in an OECD arbitration would face. The goal of the WTO arbitrators is to judge the conduct of the parties. They can put pressure on the parties to cooperate fully and can sanction them in various ways for non-performance of their obligations. The OECD arbitrators, however, are asked to judge the conduct of the non-party taxpayer. In some simple cases, the taxpayer can be expected to give full 25. For a close-to-home example, the OECD might ponder what happened with its e-commerce initiative when it allowed the private sector to assume a major organizational role. To the outside world, it appeared that the private sector took over the effective management of that project, even to the point of getting material inserted into the Commentary that would be helpful to multinational companies in avoiding the CFC rules of the United States. 26. See OECD Proposal, supra note 1, 93. 27. OECD Proposal, supra note 1, 93. [Vol. 7:9 Comments on the OECD Proposal cooperation; and, in cases of recalcitrance, the arbitrator may have the ability to impose sanctions, perhaps by making evidentiary presumptions against the taxpayer. In complex transfer-pricing cases, however, the OECD arbitrators will be heavily dependent on the taxpayer for providing relevant information and will be hamstrung if that information is not provided in timely and usable form.28 The OECD Proposal anticipates the problem by providing for a stay of the time clock for any delays caused by the taxpayer. That solution is inadequate because multinational corporations are fully prepared to provide a few hundred cartons of records on request. What the arbitrators need is usable information. That means they need full access to the electronic books and records of the taxpayer and its affiliates, plus access to the software used by the taxpayer in manipulating those books. They need all of the reconciliation data showing the relationship of taxable income to income on their financial statements. The OECD Proposal, however, gives no guidance on the data that a taxpayer is expected to produce or the form in which it must be produced. Although I favor speedy resolution of disputes, I do not favor strict deadlines that provide procedural benefits to taxpayers. Most tax administrators are anxious to resolve cases expeditiously. They fail to achieve that goal in some cases for a variety of good reasons, including a heavy workload, taxpayer recalcitrance, and the general complexity of the issues under review. Those problems are not even addressed by the imposition of strict time deadlines. In a complex transfer-pricing case, the matter should not go to arbitration until the factual record has been established and stipulated to by the parties and the taxpayer. Otherwise, the arbitrators are required to determine the facts - a process that takes many years for tax officials to accomplish. An arbitration board cannot be expected to do that type of work in six months. With the short deadline contemplated by the OECD, all that the arbitrators can be expected to do is pull some compromise number out of the air. D. Lack of Procedural Rules Governing Arbitration An arbitration proceeding, to function smoothly, needs to have rules of engagement. In private arbitrations, the parties frequently incorporate by reference the detailed rules developed by various arbitration groups, such as the American Arbitration Association, the International Chamber of Commerce, the London Court of International Arbitration, the CPR Institute for Dispute 28. In the Glaxo case, discussed supra note 10, the company had turned over 5 million pages of internal documents to the IRS by 2002. Glenn R. Simpson, "Glaxo in Major Battle With IRS Over Taxes on Years ofU.S. Sales," Wall Street Journal, June 11, 2002, page 1. 2006] Florida Tax Review Resolution, and the Judicial Arbitration and Mediation Services. The OECD notes that the arbitrators in an international tax dispute might incorporate such rules, mentioning by name the ICC. Its general rule, however, is that the arbitrators are to develop their own procedural rules on a case-by-case basis. Leaving to the arbitrators the job of developing ad hoc rules for each arbitration proceeding is a decidedly bad idea in my view. I support a full grant of authority to the arbitrators to set their own rules. Any other approach might compromise the finality of the decision of the arbitrators. But the arbitrators should not be left without detailed guidance in selecting the rules of engagement. In the typical case, the arbitrators are likely to be tax professionals, not professional arbitrators. Having them formulate their own rules of procedure as they go along will severely compromise their ability to meet their strict time deadlines. The experience of private arbitration in the United States clearly shows that inexperienced arbitrators should not be devising procedural rules on the fly. They should adopt an approved set of rules developed by professionals, which they can then adapt as circumstances warrant. E. Some Additional Criticisms The OECD Proposal is long and complex, and this brief report does not do itjustice. I have emphasized the features of the proposal I find objectionable or unworkable and have largely ignored the features that deserve praise. My purpose is not to give an overall evaluation of the proposal but to point out areas that I believe need fixing or, in the case of the secrecy rules, a fundamental change in direction. In closing, I offer the following additional brief criticisms. 1. Reimbursement for Wasted Expenses. I understand the rationale for requiring the parties to foot the bill for the arbitration proceedings. It is inconsistent with that rationale, however, to treat the taxpayer as if it were also a party. In any event, the rationale is inapplicable when the taxpayer, after instigating the arbitration procedure, takes advantage of domestic law to challenge the decision of the arbitrators in the domestic courts. In such an event, the taxpayer has wasted the time and resources of the two governments and should be required to reimburse them for the full cost of the arbitration.29 29. For a similar suggestion, see Ault, supra note 5 at 147 (2005) ("It might also be possible to stipulate that the taxpayer would have to bear the costs of the arbitration procedure if he subsequently refused to be bound by the procedure which [Vol. 7:9 Comments on the OECD Proposal The recoverable costs should include a reasonable allowance for the time of the tax officials involved in preparing their government's case. 2. Respect for Tax Officials. The negative tone directed at tax officials in the OECD report is palpable. Portions of the report sound like they were drafted by the International Chamber of Commerce. For example, in paragraph 9, the report suggests that tax officials have "no incentive" to settle disputes in the absence of the threat of binding arbitration.30 That claim is obviously wrong and demeaning. Tax officials have very strong incentives to settle disputes and achieve fair results, and they do so, according to the OECD, except in rare cases. 3. Government Avoidance of Mutual Assistance Procedure. The claim of the OECD that the existence of an arbitration procedure will encourage governments to use the competent-authority mechanism3' is almost certainly wrong in most cases. The OECD arbitration procedure is one-sided, secretive, expensive for the governments, and an intrusion on national sovereignty. Most governments will want to avoid arbitration. The OECD believes that this aversion to arbitration will lead to settlement in the rare cases not already being settled. That speculation may be well-founded.32 But the aversion to arbitration also is likely to cause many governments to nip the threat of arbitration in the bud by simply declining to engage in the mutual agreement procedure.33 Many governments are already cautious in the use of that procedure. I would anticipate that many governments offered him a solution to double taxation"). 30. OECD Proposal, supra note 1, 9. 31. See OECD Proposal, supra note 1, 11. 32. Professor Park makes a similar prediction. See Park, supra note 20 at 804 ("Like Dr. Johnson's proverbial hanging, the prospect of arbitration would serve to focus the minds of the administrative authorities to find a sensible solution, for fear of seeing the matter taken out oftheir hands altogether for decision by a neutral tribunal."). 33. Article 6(2) of the EU Arbitration convention permits a party to avoid invoking the competent-authority procedure in the first instance unless "the complaint appears to it to be well-founded." Article 25(2) of the OECD Model Tax Convention imposes an obligation on a party to invoke the competent-authority mechanism only "if the objection [of the taxpayer] appears to it to be justified." OECD Model Tax Convention, supra note 2, Art. 25, 2. 20061 Florida Tax Review will find an added reason for caution if they face the risk of mandatory arbitration. 4. Lack of Reciprocity. Arbitration is usually a reciprocal arrangement. The OECD Proposal lacks that feature. The taxpayer, and the taxpayer alone, can initiate arbitration. I would not give that authority to the taxpayer - the power to invoke arbitration should be reserved for the parties. I understand the OECD's motive for the asymmetrical rule. It wants to give the taxpayer a club to force tax officials to resolve its issue. That rule, nevertheless, is foreign to the structure of arbitration. In private arbitration, the usual rule is that arbitration is by consent of the parties. Some courts do mandate arbitration, but then the results of the arbitration typically are not binding on the parties. At a minimum, if the taxpayer is given the right to invoke arbitration, then the parties should have the same right. The one-sided feature of the OECD Proposal is not tenable. 5. Selection of Neutrals. The OECD proposed to make itself the agency for appointing arbitrators when the parties fail to do so. In my view, that "mission creep" on the part of the OECD is unwise. The OECD has no particular competence for accomplishing thatjob. More importantly, it does not have the necessary credibility, with many governments or with many taxpayers. It is a highly political organization, by design. That characterization is not meant as criticism -just descriptive of its history and purpose. I concede that the OECD is a better choice than some of the alternatives likely to be suggested by the international business community. Still, I would anticipate that a small country entering into an arbitration with a country that has a dominant political position in the OECD would not believe that the appearances of neutrality were being observed if the OECD picked the deciding arbitrator. 6. Legal Status of Commentary. I would respectfully suggest that the OECD might consult with experts on international law before assuming that the OECD Commentary, as periodically amended, would be viewed as highly relevant in interpreting a tax treaty under the Vienna Convention on the Law of [Vol. 7:9 Comments on the OECD Proposal Treaties.34 The longstanding position of the OECD is that the Commentary, as amended, should be viewed as part of the context of the treaty in interpreting a treaty based on the OECD Model Tax Convention, even with respect to the changes in the Commentary occurring after the treaty came into force. This so-called ambulatory view of treaty interpretation is sound policy; unfortunately, it is not supported by some commentators and some domestic courts. The arbitrators need to reach their decision based on the actual status of the Commentary under local law, not on the basis of the OECD's preferred treatment of the Commentary under local law. 7. Taxpayer Treated as Party. The OECD recognizes that the parties to a dispute under the mutual agreement procedure are the contracting states, not the taxpayer that invoked the procedure. It then asserts that "when the process moves to arbitration, the person who presented the case [the taxpayer] is more of a direct participant."35 If the OECD is simply being descriptive of the procedure it envisions, I agree. That de facto treatment of the taxpayer as a participant is one of the many flaws in the proposal. If the point of the assertion, however, is that the move to arbitration justifies the treatment of the taxpayer as a participant, I strongly disagree. The taxpayer simply is not a party to the dispute and should not be treated as a de facto party. 8. Language of Arbitration. The OECD suggests that an arbitration proceeding might be conducted in multiple languages. Experience in private arbitration shows that the disadvantages of that approach are significant. I understand the motive for that proposal - governments naturally want to have a proceeding conducted in the language of their country. It is not a problem, aside from cost, to provide for translation. But every effort should be made to get the parties to agree on one official language for an arbitration. 9. Premature Termination ofArbitration. Assuming that the most basic flaw in the OECD Proposal - its total secrecy - is 34. See OCED Proposal, supra note 1, 89. 35. OECD Proposal, supra note 1, 76. 2006] Florida Tax Review corrected, then the OECD ought to rethink its proposal for allowing the parties to terminate an arbitration in mid-stream. The argument made is that, because the parties allegedly control the arbitration, they should be allowed to end it at will if they can agree on a settlement.3 6 That argument has some merit; the parties should be allowed to reach a settlement whenever they wish to do it. It would be bad policy, nevertheless, to allow termination of the proceedings without requiring publication of the resolution of the case. Once a proceeding has been initiated, it is a public activity, requiring some minimum public disclosure. V. CONCLUSION For the reasons given above, I do not favor the OECD Proposal for mandatory arbitration of international tax disputes and probably would not favor it even if the many flaws in that proposal were corrected. If some form of mandatory arbitration is unavoidable, however, I would fix the various problems discussed above. In addition, I would redesign the OECD Proposal as follows: 1. Only Double Taxation Cases. I would limit mandatory arbitration to cases involving international double taxation and would exempt from mandatory arbitration all cases involving international double non-taxation.37 Given the high costs of the program, including the costs of reduced sovereignty, I see no good reason for mandating arbitration when the end result would be to increase international tax avoidance.3" 2. Finality Required. I would limit the cases of mandatory arbitration to those cases that can be settled with finality by the arbitrators. Finality is one of the hallmark features of arbitration. If the domestic law of one of the parties or the particular circumstances of the taxpayer prevent finality, then 36. See OECD Proposal, supra note 1, 98. Actually, the parties do not have control of the arbitration under the current OECD Proposal; on the contrary, arbitration is triggered by an election of the taxpayer. 37. This is the position taken in the EU Arbitration Convention. 38. Defenders of mandatory arbitration invariably point to the public benefit from a reduction in double taxation as the primary justification for their position. See, e.g., Park, supra note 20 at 863. Double non-taxation is harmful to the public interest by distorting trade flows and reducing government revenues inappropriately. [Vol. 7:9 Comments on the OECD Proposal I would not allow the taxpayer (or a government) to invoke the procedure. To achieve finality, the arbitrators need to be able to determine (1) the amount of tax the taxpayer must pay and (2) the government that should be allowed to collect that tax. 3. Certainty of Collection. I would require the taxpayer, as a condition for invoking the arbitration procedure, to provide a surety for the higher of the taxes assessed by the parties. That is, the surety should be enough to guarantee payment of the maximum tax that would be due if double taxation is eliminated and the party claiming the higher tax prevails. 4. Transparency. Finally, as noted numerous times in this report, I would require that the arbitrators prepare a full report that sets forth the controlling facts and explains in appropriate detail the legal basis for their decision. Most importantly, this report should be made public, redacted only to protect the confidentiality of trade secrets. As the OECD has recognized in other contexts, transparency is a condition for legitimacy. An adjudicative system that lacks legitimacy is simply unacceptable in a democratic society. I do not suggest that a revised arbitration proposal embodying the above suggestions would be ideal. I do suggest that it would be a significant improvement over the proposal under discussion by the OECD. I recognize that my proposed system would be less appealing than the OECD Proposal to the international business community. But as Edmund Burke famously noted, "To tax and to please, no more than to love and to be wise, is not given to men." 2006]