No job name

(C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
It’s Not Just Academic: The OECD Should ReevaluateTransfer Pricing Lawsby Michael C. Durst Michael C. Durst is a tax lawyer practicing in Washington, and during the mid-1990s served as director of the Advance Pricing Agreement Program, the Internal Revenue Service advance ruling program for The author wishes to thank those colleagues who have reviewed prior drafts of this article and provided helpful comments. The views expressed herein are those of the author only and should not be attributed to any other person or any organization.
Asonewhohasspentmuchof alegalcareerwork- annuallyaroundtheworldongovernmentalenforce-
ing in international transfer pricing, in the public ment efforts that have little chance of success, and on and private sectors, I find this article not easy to write.
meeting expensive compliance requirements, including While in government, I took pride, as did those with the maintenance of ‘‘contemporaneous documenta- whom I worked, in doing what seemed feasible to help tion,’’ which are of little real use in promoting tax build and maintain a satisfactory international tax sys- compliance. Moreover, as the rules become more and tem, and as a private practitioner, I, like many others, more entrenched in an ‘‘international consensus,’’ not have sought to operate in a manner that would reflect only the wealthier industrialized countries but also de- well on my profession and my clients. I also have veloping countries face pressure to adopt the system, sought to study the history and practical operations of thereby imposing constraints on the successful develop- the transfer pricing rules, and to share thoughts through professional and academic publications.
Recently, in response to continued criticism of the Like others inside and outside government, I have international regime, senior staff members of the perceived shortcomings in the underpinnings of the OECD Committee on Fiscal Affairs, while acknowl- transfer pricing laws and have offered suggestions for edging the need for improvement in transfer pricing remedying what seem to be serious defects, while re- rules, have taken the view that reform efforts should maining within the overall structure of the current continue to take place within the constraints of the arm’s-length system. However, despite many efforts at current system, which is based on the arm’s-length reform around the world during the 40 years or so in standard.1 Under this system, multinational groups are which the current system has played an important to divide their incomes for tax purposes, among affili- international role, governments have never been able to ates in the different countries in which the groups do administer the system effectively. Moreover, experience business, in the same way in which the income ‘‘would to date is sufficient to demonstrate that the current sys- be [divided in transactions] made between independent tem is based on faulty assumptions regarding the way multinational business is conducted, so that the system, no matter how hard one seeks to reform it, simply is not capable of functioning acceptably.
1Kevin A. Bell, ‘‘OECD’s Owens Rejects Unitary Idea, Fo- The resulting damage has been, and is, substantial.
cuses on Making Arm’s-Length Work,’’ 18 Tax Mgmt. Transfer Pricing Rep. 518 (2009); and Kevin A. Bell and Molly Moses, Governments around the world are systematically ‘‘Silberztein Defends Arm’s-Length Standard, Speaks to Restruc- hobbled in their ability to collect revenues from the turing Project, Other Issues,’’ 18 Tax Mgmt. Transfer Pricing Rep. corporate tax system. Billions of dollars are wasted TAX NOTES INTERNATIONAL
JANUARY 18, 2010 • 247
VIEWPOINTS
enterprises.’’2 This means, as a practical matter, that The inescapable problem, however, is that the failure (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
multinational businesses are required every year to en- of the arm’s-length system is not rooted merely in the gage the services of thousands of accountants, econo- particular way the system is implemented. The problem mists, and lawyers (like myself) to judge how they lies in the assumption, on which the entire system is would operate if they consisted of collections of inde- based, that the tax results of multinational groups can pendent companies instead of commonly controlled be evaluated as if they were aggregations of unrelated, groups, and that tax authorities each year must engage independent companies transacting with one another at their own thousands of accountants, economists, and arm’s length. Until that view is finally abandoned and lawyers to judge whether the businesses have con- replaced by one that is more attuned to practical reali- ties, the international corporate tax system will remain In limiting reform efforts to measures that stay within the bounds of the arm’s-length standard, OECD Flaws of the Arm’s-Length Approach
officials would reject consideration of a more objective means of dividing taxable income among affiliates that The history of the international transfer pricing sys- has been in operation for many decades among the tem is complex, and it is not possible to ascribe its de- U.S. states and the Canadian provinces, and which the velopment to any single influence or event. In review- European Commission is considering for adoption ing the historical record, however, it is impossible not within the European Union. Under this system, formu- to identify as pivotal the congressional deliberations lary apportionment, taxable income is apportioned leading to the Revenue Act of 1962 in the United among taxing jurisdictions not based on the theoretical States.4 Following the end of World War II, U.S. com- judgments of economists and other tax practitioners, panies, particularly in the pharmaceuticals industry, but on observable facts such as the extent to which had quickly developed a highly profitable industry multinational enterprises have incurred costs and gener- based on valuable patents, and they had established ated sales revenues in different jurisdictions.
‘‘base companies’’ in jurisdictions such as Switzerland and Puerto Rico to which interests in those patents The apparent intention of the Committee on Fiscal Affairs to foreclose consideration of formulary appor- tionment is disappointing. My own view, based on Some in Congress apparently believed that the years of observation, is that formulary apportionment, arm’s-length principle, which had nominally been in while far from perfect, operates much more effectively effect in tax treaties for several decades but had been of than transfer pricing under the arm’s-length standard, little practical significance during the prewar period, which has proven to be so subjective as to be unen- provided inadequate basis for ensuring that the non- U.S. base companies paid adequate royalties to their U.S. parents. In response to this concern, the House of While it is presumptuous to speculate about the rea- Representatives approved a measure directing Treasury soning of the OECD in excluding consideration of for- to devise a transfer pricing system based — at least in mulary approaches, I suspect that the decision is partly part — on a formulary system similar to that in use by influenced — as my own thinking has been in the past the U.S. states, in the House version of what became — by a fear that efforts to challenge the arm’s-length system head on are doomed to political defeat. Indeed, The Senate, however, dropped the formulary provi- no one who has worked seriously in transfer pricing sion from the act. As ultimately passed, the act did not over the decades can fail to be impressed by how include a specific transfer pricing measure, but instead, deeply the arm’s-length system is entrenched. Despite the conference report directed Treasury to devise an my belief in the deficiencies of the arm’s-length ap- approach to the income apportionment question. In proach, for example, I have suggested compromise ap- 1968 Treasury issued regulations setting forth the first proaches not in a belief that they are optimal, but in- ‘‘modern’’ transfer pricing rules under the arm’s-length stead in the belief that half a loaf is better than none.3 standard, based on searches for ‘‘comparables,’’ and detailed factual analyses of both taxpayers and alleg- Undoubtedly, many factors contributed to the deci- This is the language used in article 9 of the OECD’s model sions of the Senate to kill the formulary provision, and income tax treaty; substantially identical formulations are found in bilateral income tax treaties and national tax rules around the 3See Michael C. Durst and Robert E. Culbertson, ‘‘Clearing Away the Sand: Retrospective Methods and Prospective Docu- 4This and other developments in the history of the arm’s- mentation in Transfer Pricing Today,’’ 57 Tax L. Rev. 37 (2003) length standard referred to below are described and documented (suggesting greater use of safe harbors, and clearer articulation of in Durst and Culbertson, id., at 42-96; and Reuven S. Avi-Yonah, intragroup contracts, to facilitate operation of the transfer pricing ‘‘The Rise and Fall of Arm’s Length: A Study in the Evolution of U.S. International Taxation,’’ 15 Va. Tax Rev. 89 (1995).
248 • JANUARY 18, 2010
TAX NOTES INTERNATIONAL
VIEWPOINTS
of Treasury to adopt what might be seen as an aggres- plex and risky to be accomplished by unaffiliated (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
sively arm’s-length approach. It seems clear that at groups of companies transacting with one another in- least one motivation of Congress and Treasury was to dependently. For these reasons, many important indus- avoid upsetting the apple cart of a relatively low effec- tries are dominated, either entirely or almost entirely, tive tax rate that had been achieved by U.S. pharma- by multinational groups of commonly owned compa- ceutical companies through the use of base companies.
nies.5 In these industries — which are the only indus- The pharmaceutical industry was (and remains) central tries in which international transfer pricing matters — to the U.S. economy, and the government’s preoccupa- there typically will be no reasonably close comparables tion in 1962 was, as it is today, economic growth. To on which transfer pricing compliance and enforcement have subjected the industry to bright-line rules of in- come apportionment, rather than the murky bound- A second fundamental flaw in the arm’s-length sys- aries of the arm’s-length standard, would have risked tem, which has become increasingly evident over the economically hobbling the U.S. economy at a vulner- past decade, is that by treating different affiliates within the same group as if they were free-standing entities, Over the next several decades, as new industries de- the system respects the results of written contracts be- veloped that, like the pharmaceutical industry, were tween these related entities. These contracts have no centered on high-value, easily licensed intangibles, they real economic effects, as the same shareholders stand too became politically attached to the arm’s-length on both sides of them, but they nevertheless are given standard. The arm’s-length standard was perceived as effect under the arm’s-length standard.
so important to the after-tax well-being of the most Thus, multinational groups generally have been free strategic industries in the United States that any alter- to enter into internal contracts that shift interests in valuable intangibles to tax haven countries in which A political Gordian knot was thus created. There taxpayers conduct little if any real business activity.
may well be solid economic reasons to protect Also, more recently, tax professionals have become intangibles-intensive industries from the full statutory adept at designing contracts that treat specified mem- measure of corporate taxation. I have come to identify bers of commonly controlled groups, typically in low- strongly with these industries over the years, and I am tax countries, as ‘‘entrepreneurs’’ that bear all the busi- impressed not only by their role in driving innovation, ness risks of a set of transactions, thereby gaining but also their vulnerability to economic risks. As a re- rights to the lion’s share of income, with the activities sult of the decision of 1962, however, the means of in higher-tax countries designated under contract as effecting this reduction in effective tax rates has been ‘‘limited risk’’ distribution or manufacturing attracting the gradual build-up of a massive system of transfer relatively little income.6 Under the arm’s-length stand- pricing law that causes far more damage globally than ard, the question whether contracts among related par- could possibly have been foreseen in the 1960s. More- ties should be respected depends on whether the con- over, the system has become so ponderous that it is tracts are similar to those into which unrelated parties difficult if not impossible for the public, or even all but might enter — but because the activities of unrelated specialized tax practitioners, to understand. Difficult parties are systematically different from those of com- though it might be politically, the time has come for a monly controlled groups, there are never any plausibly fundamental redesign of the international tax system similar contracts against which to evaluate the con- around income apportionment rules that work better tracts among related parties, so as a practical matter, it is impossible for governments to second-guess them.
It is not surprising, then, that real-life transfer pric- Incremental attempts at reform are doomed to fail- ing examinations, no matter how well conducted, even- ure because the unenforceability of the arm’s-length tually dissolve in confusion and controversy. Anyone standard derives not from the details of its implementa- tion but from its central premises. First, at the center of the arm’s-length transfer pricing system is the idea that income from transactions among members of multina- 5For example, I and others who have worked in international tional groups should be benchmarked by the results of business over the past three decades have seen a great many in- comparable transactions among unrelated parties. It stances in which multinational companies have acquired previ- requires no sophisticated analysis, however, to recog- ously independent distributors in their larger markets, for the rea- nize that commonly controlled multinational groups son that it usually is not efficient for a manufacturer to seek to arise precisely because there are some transactions that distribute goods by means of continual negotiations with unre- do not occur, on an economically efficient basis, be- lated distributors in the larger markets.
tween unrelated parties. Thus, for example, the manu- 6These developments are sometimes discussed under the ru- facturing and marketing of expensive consumer du- bric of ‘‘restructuring,’’ and have been under active discussion rables on a global basis, or the exploitation of valuable within international tax circles in recent years. See, e.g., Kevin A.
Bell, ‘‘OECD Delegates Debate How to Price Business Restruc- intellectual property in such fields as pharmaceuticals, turings, Taxpayer Representatives Bemoan Non-Recognition Pro- software, and information technology, are far too com- posal,’’ 18 Tax Mgmt. Transfer Pricing Rep. 159 (2009).
TAX NOTES INTERNATIONAL
JANUARY 18, 2010 • 249
VIEWPOINTS
who has participated in a transfer pricing controversy, in which its potential exposure, based on the stated (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
from the standpoint of either the taxpayer or the gov- IRS position, was reported to be $14.5 billion.8 A sys- ernment, can see vividly that the system does not tem of taxation in which the government and tax- achieve its intended result of a reasonably clear meas- payers, presumably in good faith, can state positions ure of a company’s taxable income according to clearly that are this widely disparate does not deserve to be articulated, and hence practically enforceable, stand- ards. Words and numbers spew forth from both tax- payer and government representatives, but the words These cases, on which information is publicly avail- and numbers have little connection with economic real- able, represent just the tip of the iceberg. A great many ity. Controversies are resolved through the exhaustion additional controversies involve taxpayer and govern- of both sides, using the grossest forms of compromise, ment positions that are similarly divergent, but under rather than on a standard that can assure reasonably applicable taxpayer privacy laws are settled out of the similar results in similar cases. For decades the system public eye, usually in administrative appeals proceed- in operation has been characterized by chaos.
ings that, despite the good intentions of all involved, take on the character of negotiations over the sale of a The unenforceability of the
used car of questionable mechanical heritage. The re- arm’s-length standard
sults in different cases cannot be reconciled according derives not from the
Recent reform efforts, particularly within the OECD, details of its
have sought to address these problems by greater reli- ance on ‘‘income-based’’ transfer pricing methods.9 implementation but from
Such methods, which have been in use for many years, its central premises.
especially in the United States, do not seek to bench- mark prices in particular transactions (for example, a particular license of intangibles) against supposedly comparable prices, but instead seek generally to bench- mark the incomes of members of groups against the The extent of this chaos unfortunately can be fully appreciated only by those who work with the system incomes of arguably similar uncontrolled companies.
regularly, but a useful inkling can be seen from the Thus, for example, the income of a controlled U.S. dis- wildly disparate positions of tax agencies and taxpayers tributor of a foreign manufacturing group might be when entering into a controversy. For example, in the benchmarked against the net incomes of uncontrolled recently decided U.S. Tax Court case of Veritas Software distributors of roughly similar products.
Corp. v. Comm’r, 133 T.C. No. 14 (Dec. 10, 2009),7 the As an academic matter, such a method might have IRS originally sought to increase the taxpayer’s income some attraction — after all, we learned as undergradu- from a particular transaction by $2.5 billion, whereas ates that returns on capital tend, over the long run, to the court largely upheld the taxpayer’s own income equilibrate among firms in a competitive economy — inclusion of only several hundred million dollars. In but in practice it fails dismally. Because the arguable the recent decision of the Tax Court of Canada in comparables that can be found are always very ap- General Electric Canada, Inc. v. The Queen, Tax Ct. (Can.), proximate, the methods in the applicable rules generate 2006-1385(IT)G (Dec. 4, 2009), the Canadian govern- supposed arm’s-length ranges that are so wide as to be ment sought to disallow deductions of approximately useless. For example, a typical transfer pricing analysis, C $130 million, and the court rejected the attempted conducted according to best practices under the U.S.
adjustment in its entirety. In 2006, in a rare case in regulations, might conclude for a given manufacturer which the details of an out-of-court settlement in a that a net operating margin within the range of, say, 2 transfer pricing case have become matters of public to 6 percent should be accepted as arm’s length. That record, GlaxoSmithKline settled for $3.1 billion a case means, for example, that a net income for tax purposes anywhere between $200 million and $600 million 7Veritas and General Electric Canada, mentioned immediately below, both remain subject to appeal at this writing, so the final outcomes of these cases are not known. Even if their results are Tamu N. Wright, ‘‘Glaxo to Pay $3.4 Billion to Settle Larg- overturned on government appeal, however, these cases demon- est Tax Dispute in IRS History,’’ 15 Tax Mgmt. Transfer Pricing strate the lack of meaningful guidance available to both taxpay- ers and governments in seeking to resolve cases on a principled 9The Committee on Fiscal Affairs has incorporated its sug- basis under current law. (For the court opinion in Veritas, see Doc gested reforms in proposed revisions to the OECD transfer pric- 2009-27116 or 2009 WTD 236-42. For the court opinion in General ing guidelines. These proposed revisions are available online at Electric Canada, see Doc 2009-26729 or 2009 WTD 233-15.) http://www.oecd.org/dataoecd/1/57/43655703.pdf.
250 • JANUARY 18, 2010
TAX NOTES INTERNATIONAL
VIEWPOINTS
should be considered acceptable. Such a wide range is Alternatively, it might be possible to design profit- (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
almost meaningless for purposes of effective tax ad- split formulas for different industry groups that, while not precisely tailored to every taxpayer’s circumstances, nevertheless provide sensible results in most instances It might be possible to address the problem of wide and have the great advantage of being predictable and ranges by instead establishing income-based rules that understandable by taxpayers and governments alike.
incorporate specific income targets instead of ranges.
For example, it might be required that controlled distri- But a system based on this kind of profit split is, of bution entities earn taxable profits of at least X percent course, a formulary system of the kind in use by the of sales, or that controlled manufacturing entities earn U.S. states and Canadian provinces and under consid- a taxable markup of at least Y percent on some speci- eration by the European Commission. It also is the fied measure of costs.11 Such rules, however, would kind of income apportionment system that, I think, amount to nothing more than a very rudimentary — must be accepted as necessary if the international tax and dysfunctional — kind of formulary system. In par- ticular, these kinds of rules would require distribution or manufacturing entities to earn specified profits re- The practical costs of the unenforceability of arm’s- gardless of the profitability of the multinational enter- length transfer pricing rules are enormous.
prise as a whole, and such a system would quickly col- Governments’ Loss of Fiscal Control
lapse in the face of changing economic conditions.
Therefore, greater reliance on profit-based methods Current rules permit those taxpayers positioned to such as the comparable profits or transactional net shift income by contract, involving either intangibles margin methods does not appear to offer a promising ownership or, more recently, risk-shifting, to obtain dra- route for repairing the arm’s-length system.
matic reductions in their effective tax rates. The result This conclusion also applies, unfortunately, to profit- is to severely limit the effectiveness of the corporate tax split methods, the other group of income-based transfer as a means of raising revenue. This is not to say that at pricing methods that sometimes are used. In theory, this time, companies in high-technology industries or profit-split methods hold promise because they do not other economic sectors should face increases in their rely as heavily as other transfer pricing methods on effective tax burdens through transfer pricing reform or searches for uncontrolled comparables. In practice, other means. The problem with current transfer pricing however, profit-split methods cannot yield reasonably rules, however, is that they distribute tax reductions objective estimates of related parties’ appropriate net arbitrarily among different companies, even within the same industries, and the reductions are largely out of the control of the legislative process. The transfer pric- First, the kind of profit split that is most commonly ing rules should be replaced by a system that raises applied — the residual profit split — in fact relies revenues predictably according to the decisions of leg- heavily on searches for uncontrolled comparables for a islators; tax reductions for economic growth should be crucial step in its operation, causing the method to fail designed under the control of legislators and targeted for the same reasons as other transfer pricing methods.
where they are needed most, rather than distributed Some other forms of profit splits skip the step of reli- ance on searches for comparables, but these kinds of profit splits require that a means of dividing profits Recently, the ‘‘self-help’’ tax reductions available to among members of a group be designed individually in companies through transfer pricing have received criti- every case. The methods require virtually endless fac- cism from nongovernmental organizations concerned tual judgments to be made by both taxpayers and tax with prospects for economic growth in the poorer de- authorities, and they are not susceptible of reasonable veloping countries.12 These organizations may in some instances overstate the extent to which transfer pricing rules are impairing economic growth in developing countries, but their underlying observation is sound. To the extent developing countries follow the international consensus and adopt arm’s-length transfer pricing rules, Surprisingly, such wide ranges do not appear to have been unforeseen consequences of the applicable rules. The U.S. trans- fer pricing regulations themselves contain examples in which the arm’s-length ranges extend from $19,760 to $34,840, a range of 76 percent based on the lower bound; and from $15,500 to 12See, e.g., Christian Aid, ‘‘False Profits: Robbing the Poor to $30,000, a range of 94 percent based on the lower bound. Treas.
Keep the Rich Tax-Free’’ (2009), available at http:// reg. section 1.482-5(e), examples 1 and 3.
www.christianaid.org.uk/Images/false-profits.pdf; Greenpeace, 11A similar result might be reached by eliminating the prac- ‘‘Conning the Congo’’ (2008), available at http:// tice of accepting a result anywhere within a range of results, and www.greenpeace.org/raw/content/nederland-old/reports/ instead insisting that an entity’s taxable income be determined at conning-the-congo.pdf; and Duncan Green, From Poverty to Power (Oxfam International: 2008), at 314-317.
TAX NOTES INTERNATIONAL
JANUARY 18, 2010 • 251
VIEWPOINTS
their corporate tax systems will be hobbled in unin- ated. And this industry (of which I am a part) inevi- (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
tended and unpredictable ways. The OECD, as the tably has grown into something of a political force guardian of the international consensus, should design with an interest in retaining current rules. Finally, the costs of government attempts at administering the sys- Astronomical Compliance Costs
tem, generally futile though they appear to be, must amount to at least hundreds of millions of dollars per In a vain attempt to make transfer pricing rules year. Overall, the costs of trying to make the arm’s- more administrable, governments around the world length system work — or even just appear to work — now require taxpayers to maintain, annually, contem- easily reach billions of dollars per year worldwide.
poraneous documentation of their transfer pricing poli- In connection with the subject of compliance and cies. Such documentation typically is voluminous; it enforcement costs, a word should be said about proce- contains detailed descriptions of the taxpayer’s busi- dural innovations that are sometimes seen as improving ness, and pages and pages of computer output detailing the administrability of the arm’s-length system. These the financial information of arguably comparable com- include (i) advance pricing agreements by which tax- panies. The result of all this paperwork, however, is payers can enter into advance rulings, sometimes with almost invariably an arm’s-length range extending at more than one government simultaneously, on transfer least 100 to 200 percent above the bottom — a range pricing matters, and (ii) binding arbitration of transfer too wide to be of any real use. Contemporaneous pricing issues between governments, under income tax documentation therefore serves to put a thin veneer of treaties. I have personally been involved in the develop- scientific method over an analysis that is in fact highly ment of APA rules, and I do think that in some cir- subjective and imprecise — and the cost is enormous.
cumstances APAs avoid litigation or other protracted Assuming, very conservatively, that there are 5,000 disputes and thereby save valuable resources for tax- multinational enterprises in the world that must main- payers and governments. As a practical matter, how- tain contemporaneous documentation for use in multi- ever, APA negotiations often are plagued by the same ple countries, and that the average annual cost of cre- vagueness of legal standards that besets the arm’s- ating and maintaining such documentation (in internal length standard generally; APA negotiations often are personnel costs and outside consultants’ fees) is the protracted, difficult, and expensive; and APAs can ad- equivalent of $200,000, then annual costs to the dress only a tiny fraction of transfer pricing issues that worlds’ shareholders of maintaining the veneer of re- arise in multinational business each year. Similarly, spectability for current transfer pricing is the equivalent binding arbitration under tax treaties is a positive and cost-saving development; however, arbitration can as a practical matter barely scratch the surface of the great The practical costs of the
many double tax issues that the arm’s-length approach unenforceability of
spawns each year. It will be far better to reduce the incidence of double taxation by adopting a formulary arm’s-length transfer
approach that governments can use to resolve double taxation disputes in a predictable and uniform manner.
pricing rules are enormous.
Even if the particular formulas used by the different governments involved differ in some respects, there should be fewer axes of negotiation — and hence, quite possibly, greater ease in resolving double taxation And this is for documentation alone — billions issues — than under the arm’s-length system. In sum, more are spent annually to perform the elaborate legal while APAs and binding arbitration can save costs in and accounting work needed to accomplish shifts of some cases, they cannot transform the arm’s-length income, contractually, to low-tax jurisdictions. Simply system into one that is practically manageable.
look at the rosters of the large international accounting and law firms, and the many independent economic Loss of Respect for the Tax System
consultancies involved in transfer pricing, to get an The world has in recent years experienced the fail- understanding of the costly industry that has been cre- ure of some fiscal institutions that were supposed to be safeguarding the public interest but instead were cap- tured, to greater or lesser extent, by embedded interest groups pursuing narrower agendas. Although it is unre- 13This estimate is not, of course, based on a detailed analysis, alistic to expect the general public to gain a detailed and the actual annual costs of compliance are likely to be either understanding of the methods of international corpo- higher or lower than this estimate. I believe, however, that the rate taxation, the international movement of income to estimate in the text is very conservative, and that actual compli- tax havens is increasingly visible, and I believe the feel- ance costs are substantially higher. In any event, the underlying ing is growing around the world that the international problem is that substantial real resources are being expended on documentation that is of questionable value, at best, for purposes tax system is sacrificing the public interest in favor of embedded beneficiaries. This is, I believe, the source of 252 • JANUARY 18, 2010
TAX NOTES INTERNATIONAL
VIEWPOINTS
some of the animus displayed recently by nongovern- model of understatement, the committee claims, ‘‘A (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
mental organizations concerned with the situation of practical difficulty in applying the arm’s-length prin- developing economies, and I think reaction to this per- ciple is that associated enterprises may engage in trans- haps overstated animus has led in turn to overstate- actions that independent enterprises would not under- ment by those who continue to defend the arm’s-length take.’’15 This would suggest that generally, members of system. My overall — if reluctant — view, after years multinational groups function pretty much as would of practice in this field, is that the critics are funda- independent enterprises transacting with one another at mentally correct; the current system fails utterly in its arm’s length, and that departures from this comfortable public role, the appearances created are unseemly, and situation are exceptions rather than the rule. The real- ity, however, is just the opposite: Multinational groups exist precisely because it is impossible to conduct their The Search for an Alternative System
businesses other than under common control; members A fundamental alternative to arm’s-length transfer of multinational groups will rarely, if ever, transact pricing — formulary apportionment — not only has business with each other similarly to unrelated parties long been available, but it has been in effect for many acting at arm’s length. Similarly, the proposed revisions years among the U.S. states and the Canadian prov- would repeat the statement from the existing guide- inces. Many people (including myself) have analyzed in lines: ‘‘The arm’s length principle has . . . been found detail the manner in which formulary apportionment to work effectively in the vast majority of cases.’’16 could be applied at the international, in addition to While in political environments such as the OECD, state and provincial, levels of taxation.14 It is clear that people sometimes find themselves saying things they an international formulary system would operate only later find they cannot support, it is inconceivable to me imperfectly, just as formulary apportionment has oper- that any fair observer of transfer pricing practice over ated only imperfectly within the United States and the past 20 years could believe this statement to be cor- Canada. The admitted imperfections of the formulary system, however, are meaningful only in the context of a detailed comparison with those of the arm’s-length Some have expressed unwillingness to subject for- standard. Because the formulary system, unlike the mulary approaches to careful practical review based on arm’s-length approach, is not based on a flat fallacy — the view, which seems patently erroneous, that the for- namely that ‘‘comparables’’ and ‘‘functional analysis’’ mulary approach is merely an academic or theoretical can lead to a workable tax system — it seems clear construct, and that information on which it can be that a formulary system, despite all of its flaws, enters evaluated practically do not exist. To the contrary, the comparison with a key practical advantage over the there is probably more practical experience available of arm’s-length approach. The formulary system can yield formulary systems than of arm’s-length systems. Ex- a reasonably clear measure of a company’s taxable in- perience with formulary apportionment in the United come in a given jurisdiction, but the arm’s-length stand- States and Canada provides a body of detailed factual ard, both in theory and in practice, cannot.
experience extending over almost a century — far longer than the arm’s-length standard has been in Over the years, the Committee on Fiscal Affairs has widespread and intensive use internationally. Moreover, steadfastly declined to consider global formulary appor- a wealth of practical work has been done in connec- tionment as a viable alternative to the arm’s-length sys- tion with the European Commission’s development of tem, and if anything, this position has recently become a formulary system; and at least one example of statu- more insistent. For example, in recent proposed revi- tory and regulatory language based on a formulary ap- sions to the OECD’s transfer pricing guidelines, the proach has been prepared for evaluation by a practi- committee repeats generalizations concerning the tioner in the United States.17 There is plenty of arm’s-length standard that simply cannot be supported practical, detailed material available on which to base a by any fair evaluation of real-life experience. Thus, in a careful look at a possible formulary system.
It is against this background that the Committee on Fiscal Affairs, and particularly its professional staff within the Centre for Tax Policy and Administration, I am sure that my own contributions can be improved would seem to have an important evaluative role. On upon, and indeed I hope that they will be, but they do suggest that seeking to design an international formulary system is not an utterly hopeless task. I have provided suggested statutory and regulatory language in Michael C. Durst, ‘‘A Statutory Proposal for U.S. Transfer Pricing Reform,’’ Tax Notes Int’l, June 4, 2007, 15Proposed revised paragraph 1.11, based on existing para- p. 1041, Doc 2007-12446, or 2007 WTD 109-8; a second version, incorporating some technical refinement, is provided in Reuven S. Avi-Yonah, Kimberly A. Clausing, and Michael C. Durst, ‘‘Al- Proposed revised paragraph 1.9, based on existing para- locating Business Profits for Tax Purposes: A Proposal to Adopt a Formulary Profit Split,’’ 9 U. Fla. Tax Rev. 497, 540-553 (2009).
TAX NOTES INTERNATIONAL
JANUARY 18, 2010 • 253
VIEWPOINTS
its face, it might seem questionable to look for guid- demic analysis has provided useful insights regarding (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
ance to the staff of the same multinational organiza- both formulary and arm’s-length systems, it cannot tion that, 15 years ago, produced the OECD transfer substitute for practical comparisons of the operations pricing guidelines, and that continues today to argue of arm’s-length and formulary systems based on realis- that reform efforts should remain within the confines tic and detailed fact patterns, conducted by experienced of the arm’s-length construct. Despite whatever has tax administrators. What is needed is detailed, hard- happened in the past, however, it remains the case that nosed, and sophisticated staff work illustrating how the two systems compare in operational effectiveness; the results of this staff work then can be made public for scrutiny by academics as well as others.
Key to this effort will be assigning the task of build- ing the formulary model to those who are willing to • despite the undeniably political nature of the serve as vigorous advocates on its behalf. Only if that OECD as an organization, generally permitted to condition is met can a true debate be held, with the operate in an environment of independence from outcome unknown until the debate is conducted. It is possible that, as a result of the debate, the staff will Moreover, the staff has the ability to supplement its have reached sufficient consensus to formulate a rec- ranks, as needed, with high-quality consultants from ommendation to the governmental representatives who academia or other sources outside the OECD, and comprise the Committee on Fiscal Affairs.
Alternatively, views of the staff may remain suffi- ciently divided so that it is not feasible to provide a The member governments of the Committee on Fis- single staff recommendation. Even in that case, the cal Affairs should ask the staff, using whatever inde- governmental representatives will have in front of them pendent outside technical assistance is needed, to build vigorous defenses of competing views, compiled in in- a thorough model of a system of formulary apportion- sulation from political pressure and compiled according ment that could replace the arm’s-length system.18 to the high quality expected of OECD staff and con- Based on such a model, the staff should then conduct sultants. Also, once the staff work is done and public a detailed and vigorous debate, contrasting that model (and inevitably political) debate ensues, the Committee with the arm’s-length standard as reflected in the on Fiscal Affairs should ensure that not only business guidelines and the various national bodies of law that groups, but also nongovernmental organizations that incorporate it. The goal of the debate and comparison are likely to approach the issues from different perspec- should be a fresh comparative evaluation of the two tives, are given opportunity to participate fully in the systems, conducted in an environment as insulated as possible from political lobbying. Further, the debate should be transparent, with all working papers open to By insisting that the debate be held under the super- public scrutiny — but only after the final product of the vision of its staff, and in an adversarial fashion that is evaluation has been released, to provide additional in- insulated from outside politics, the Committee on Fis- cal Affairs would avoid the trap that has affected other recent reviews of the merits of arm’s-length and formu- It is important that the comparison be conducted on lary systems, including the debate over the U.S. Treas- a practical level by staff members and consultants fa- ury white paper in the late 1980s and early 1990s and miliar with the day-to-day operations of tax adminis- the ensuing debates over the OECD guidelines ending trators and other practitioners. The recommended in- in 1995. In both these instances, the din of political quiry therefore should be different from, say, a pressure was deafening. Although formulary ap- conference involving the presentation of academic pa- proaches were formally given some consideration, their pers. There is a large amount of academic literature on development was forestalled by political pressure before the formulary versus arm’s-length question. While aca- they could be articulated in any detail and compared with the arm’s-length approach point by point. The political order of the day was to bury formulary, not to analyze it seriously, much less to praise it.
18The European Commission has in recent years been devel- In particular, the problem with past official and oping a formulary system for possible use in the European Union as part of a proposed system for a common consolidated quasi-official reviews of the arm’s-length standard was corporate tax base, and the work done by the commission’s staff that because of political pressures to reject a formulary could assist the Committee on Fiscal Affairs in designing an ap- system as unworkable, the identification of likely flaws portionment model. The commission describes its work online at in the operation of a formulary approach — and they http://ec.europa.eu/taxation_customs/taxation/company_tax/ are, admittedly, significant — was viewed as sufficient common_tax_base/index_en.htm; see also Joann Martens Weiner, reason to reject adoption of the system as a dangerous Company Tax Reform in the European Union: Guidance From the United States and Canada on Implementing Formula Apportionment in ‘‘unknown quantity,’’ without considering carefully the EU (Kluwer: 2005). See also supra note 14.
whether a formulary system might function much more 254 • JANUARY 18, 2010
TAX NOTES INTERNATIONAL
VIEWPOINTS
fairly and predictably than the current system. Histori- Some have also criticized formulary systems, cor- (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
cally, reviews of formulary versus arm’s length were rectly, on the grounds that including such items as pay- similar in many ways to trials in which judgment was roll and property within the formula could influence rendered after the indictment was read, before the in- companies’ decisions concerning the location of em- troduction and serious consideration of the evidence.
ployment and plant construction. For many years, There is no need, however, for such summary rejection however, U.S. states have addressed this issue by basing of the formulary model. It should be possible for prop- their formulas predominantly on the location of sales, erly motivated, knowledgeable, and unbiased staff rather than employment and plant, and it should be members to build a realistic model of a formulary sys- possible to adopt a similar approach internationally.
tem at the international level and to compare its coher- Further along these lines, some critics point — again ence and workability point by point with the current correctly — to difficulties in determining where sales should be deemed made under a formulary system, for example, in instances of electronic commerce or when For example, in the past, some have sought to dis- sales of intermediate goods are made to be incorpo- miss formulary apportionment peremptorily based on rated in products for final sale outside the location of the observation that if countries choose to adopt differ- manufacture. On careful comparison, however, such ent formulas, differing, for example, in how heavily difficulties of factual determination might well be far sales or payroll factors are measured, a problem of po- more tractable than the countless factual uncertainties tential double taxation will arise. That certainly is true that arise in virtually every situation under the arm’s- — but it is also true that huge amounts of double taxa- tion result from the impossibility of conducting satisfy- A practical comparison will make it possible as well ing analyses under the arm’s-length system. Moreover, to address, in a balanced way, another alleged short- under current rules, which provide no clear standards, coming of formulary apportionment. In particular, it double taxation disputes between tax administrations sometimes is suggested that formulary apportionment can take many years to resolve, if they are resolved at would face serious difficulties of implementation be- all. Under a formulary system, even if different coun- cause it would require taxpayers and tax authorities to tries have adopted different formulas, they might de- review not only the financial information of the affili- velop hybrid formulas to resolve double taxation dis- ate of a multinational group located in a particular putes relatively easily and quickly. Thus, the double country, but also other components of the group. It is taxation problem might actually be less troublesome certainly correct that a formulary approach requires under formulary apportionment than under the current reference to information originating outside a particular system. In any event, this question must be looked at taxpayer’s jurisdiction, but that does not necessarily closely and the two systems compared carefully and as mean that the formulary approach raises more difficult specifically as possible if a meaningful judgment is to information requirements than does the arm’s-length approach. Even under the arm’s-length approach, tax- payers and tax authorities must look at the activities Similarly, some would seek to dismiss formulary and results of affiliates in many jurisdictions to apply approaches out of hand based on the observation that transfer pricing methods — for example, when apply- under formulary systems, disputes arise concerning ing profit split methods, and when searching around which business activities should be included in the ag- the world for a multinational group’s ‘‘internal compa- gregation of income that is subject to the formula. In- rables.’’ For these and other reasons, tax authorities deed, over the years, there have been many such dis- already routinely refer to financial results of out-of- putes under U.S. state tax laws involving the definition country affiliates in performing transfer pricing and of the unitary group, and such disputes can be trouble- other international tax examinations.20 More generally, some. Similar problems arise, however, every day under it must be remembered that under a formulary system arm’s-length transfer pricing rules, when it is necessary the tax administrator typically does not need to audit to determine which of a company’s business activities the taxpayer’s operations in every separate jurisdiction.
should be taken into account in conducting the func- What is needed instead is to compare the operations in tional analysis needed to try to identify appropriate the tax administrator’s home jurisdiction against the comparables. Indeed, those difficulties arise in nearly consolidated financial results of the multinational every transfer pricing question that tax administrators must address, and there is virtually never a means of resolving the question systematically. It seems to me that the question of which business activities should be included in the analysis is probably far more troubling My own less than perfect, but I hope still potentially useful, suggestions for addressing the problem of location of sales are under the arm’s-length approach than under a formu- contained in Avi-Yonah, Clausing, and Durst, supra note 14, at lary system. The member governments of the OECD should see this question addressed carefully in an open 20See, e.g., IRC sections 6038A and 6038C, and the regula- TAX NOTES INTERNATIONAL
JANUARY 18, 2010 • 255
VIEWPOINTS
group. Multinational groups around the world typically ate them practically and comparatively, rather than use (C) Tax Analysts 2010. All rights reserved. Tax Analysts does not claim copyright in any public domain or third party content.
maintain both consolidated and entity-level financial them as a reason to dismiss formulary approaches out statements on a global basis, and it seems unlikely that of hand. Let the battle of arm’s-length versus formu- moving to a formulary system would impose prohibi- lary be fought energetically, under the auspices of the tive accounting or information-gathering needs.21 In Committee on Fiscal Affairs, by respected and commit- sum, it seems quite possible that a formulary system ted advocates for both sides. There is more than would impose fewer, rather than greater, information enough indication of dysfunction in the current system, needs on taxpayers and governments than does the cur- with serious harm being caused, to justify such a com- rent attempt to divide income based on the arm’s- parative inquiry. The cost of this work will be very length standard. In any event, the comparison should small compared with the hundreds of billions of dol- be made carefully and in detail; the result should not lars of tax revenue, including that of the poorest coun- be assumed, in advance, in favor of either approach.
tries, at stake, as well as the billions lost each year to Other defects of formulary apportionment undoubt- excessive costs of compliance and attempted enforce- edly can be pointed out,22 but it is important to evalu- ment. This is not an abstract intellectual exercise; if we are in fact operating under a dysfunctional system, it is real people who are ultimately being harmed. Let’s get on with the job of resolving this issue with the inten- 21In fairness, this might not have been the situation, say, 40 years ago when current policy views relating to apportionment methods were first being formed; at that time, multinational groups had less experience in comprehensive global accounting 22As an additional example, those intent on preserving the what the economic and social advantages of such changes might arm’s-length approach point correctly to the need for numerous be. What is needed is to develop a model for the necessary legis- changes to national tax laws and to tax treaties, and for poten- lative and treaty changes and transitional rules, then consider tially complex transitional rules, if formulary apportionment is to their cost in light of the advantages that might ensue from re- be implemented. Those concerns, however, could be used to placement of the current system, which surely involves massive avoid serious consideration of all kinds of tax reforms, no matter administrative difficulty of its own.
(Footnote continued in next column.)
256 • JANUARY 18, 2010
TAX NOTES INTERNATIONAL

Source: http://www.taxjustice.net/cms/upload/pdf/Durst_1001_OECD_-_not_just_academic.pdf