1 Introduction
1.1 General
On the 31 May 2024, Lawrence Williams defended his doctoral thesis at the Faculty of Law, Stockholm University, where I was the faculty opponent.
This is a review of his doctoral thesis “Transfer pricing and the World Trade Organization agreement. The compatibility of transfer pricing principles contained in bilateral and multilateral tax treaties with the WTO non-discrimination norms”.
All in all, the thesis covers 330 pages and is published by Stockholm University; it is publicly available online through the website of the Stockholm University Library.1
The thesis concerns norms relating to non-discrimination and equal treatment contained in different agreements within the World Trade Organization (WTO) framework, primarily the General Agreement on Tariffs and Trade (GATT), the General Agreement on Trade and Services (GATS) and the Subsidies agreement. In particular, it focuses on provisions that grants national treatment and the treatment as most-favoured-nation (“MFN treatment”). These norms are identified in a number of provisions in primarily the three previously mentioned treaties. They are subsequently analysed in relation to 1) transfer pricing principles, and 2) Amount A of the OECD Multilateral Convention (MLC).
Basically, transfer pricing involves the mechanisms for establishing the correct arm’s length price for goods and services sold and bought between entities within the same corporate group in cross-border situations. The underlying objective is that the arm’s length price should be upheld, i.e. the price that would have been used between independent parties. For decades, the OECD has issued authoritative guidelines on how to establish the arm’s length price. Amount A of the OECD MLC is an agreement under Pillar One of the OECD/G20 Inclusive Framework. In relation to the arm’s length principle, Amount A of the OECD MLC offers an alternative way for allocating corporate profits, taking into account where “value is created” and where consumers are resident. It is still unclear whether Pillar One and Amount A of the OECD MLC actually will be approved by national parliaments and enter into force. Not the least the lack of support in the US Senate makes it highly uncertain whether the US will join the agreement, and without US participation, the successful implementation of the agreement is questionable. Pillar One is primarily targeted on taxing corporate profits of large US tech companies such as Google, Amazon and Meta.
The thesis can be downloaded from the web address: https://su.diva-portal.org (search on the author’s name, Lawrence Williams).
1.2 Subject, purpose and delimitations
In my view the subject of the thesis is good. The interaction between WTO norms and rules on international tax law has, so far, not been the subject of any doctoral thesis in Sweden, even if there are some international publications. There are some scholarly contributions in this field of law, but there are many uncertainties still left to be explored, and research in this area is welcome.
Basically, the thesis concerns two different fields of law, which the author labels “silos” (p. 55). One silo consists of WTO law and its non-discrimination norms, and the other silo is tax law, including transfer pricing and the calculation of Amount A according to the OECD MLC. A general objective of the author is to connect these two fields (“silos”) of law, and to investigate whether the arm’s length principle and the Amount A of the OECD MLC are in breach of WTO requirements, primarily the national treatment principle and the most-favoured-nation clause.
More precisely, the general purpose of the thesis is (p. 14) “…to explore the potential interactions, overlaps, and conflicts between bilateral tax treaties based on the OECD Convention and the OECD MLC with the WTO agreement”. In order to achieve this purpose the author (p. 14) “…will conduct a legal analysis on the compatibility of the transfer pricing principles contained in the OECD Guidelines, OECD Convention and the OECD MLC with the most favoured nation principle and national treatment principle contained in the General Agreement on Tariffs and Trade (“GATT”), the Subsidies Agreement, and the General Agreement on Trade in Services (“GATS”).”2
From a theoretical perspective, the topic of the thesis is both interesting and sophisticated.
A terminological issue should be addressed already at this stage. The author interchangeably refers to “the arm’s length principle” in the singular and “the transfer pricing principles” in the plural. In the list of abbreviations (p. X, Roman numeral) it is explained that the term “transfer pricing principles” should be understood as “the arm’s length principle as contained in [a] bilateral tax treaty based on the OECD Convention and Amount A of the OECD MLC” (see also on p. 48). Throughout the thesis, both “the transfer pricing principles” and “the arm’s length principle” are used, and I think that this use is somewhat confusing. In my view Amount A of the OECD MLC is a step away from transfer pricing principles and instead expresses another form of allocating corporate profits between jurisdictions, namely the method generally known as “formulary apportionment”. It is unclear to me why the author uses the terms “arm’s length principle” and “transfer pricing principles” interchangeably. A brief description of the arm’s length principle is contained in sub-chapter 2.6 (pp. 59–63), and Amount A of the OECD MLC is described in sub-chapter 2.7 (pp. 63–70). There is no mentioning of “formulary apportionment”, despite it being an accepted term used by the OECD in the Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (2022). According to the OECD global formulary apportionment would allocate the global profits of a multinational corporate group on the basis of a predetermined and mechanically working formula.3 The OECD considers global formulary apportionment as a “non-arm’s length approach”. Since Amount A of the OECD MLC embodies formulary apportionment, I think it is wrong to label it as a transfer pricing principle. It is, in fact, a different creature altogether.
Footnotes have been omitted from this quote.
OECD Transfer pricing guidelines for multinational enterprises and tax administrations, Paris, 2022, para. 1.16.
1.3 Outline – and some critical aspects of the thesis
The thesis contains eight chapters. Chapter 1 is the introduction and contains the scholarly starting points such as background, purpose, research questions, delimitations, legal method and materials, as well as previous research and formalities.
Chapter 2 provides a general background to, on the one hand, WTO and non-discrimination norms, and, on the other hand, the arm’s length principle and Amount A of the MLC. The arm’s length principle serves as a model for establishing the price on goods and services sold and bought within corporate groups across borders. Amount A of the OECD MLC is a recent model for allocating corporate profits between entities among large multinational corporate groups. An agreement exists between governments but has yet to be passed by national parliaments. The future of Amount A the OECD MLC is still uncertain.
The author describes the general methods for resolving conflict situations between different norms in international law in Chapter 3. I would say that this chapter is rather descriptive and relies much on the research of others. There is a good article on this subject in the Bulletin (2019) by Alireza Salehifar who discusses many of the issues raised in Chapter 3 and also refers to many of the sources mentioned by Lawrence Williams in this chapter.4 To my surprise that article is not referred to in Chapter 3 but rather included in the enormous reference list in footnote 368 on p. 120 in Chapter 4. That article should, of course, have been mentioned in Chapter 3.
Chapter 4 provides an historical account of the development of the GATT, GATS and Subsidies Agreement. This chapter is titled “The history”. Considering that it only covers 16 pages and that the subject is considerable, the use of an indefinite article (“A History”) would have been more appropriate than the definite article (“The History”). Chapters 5, 6 and 7 are the main analytical chapter of the thesis. In Chapter 6 the overlaps and conflicts between on the one hand transfer pricing principles and Amount A of the OECD MLC, and on the other hand the national treatment principle and the most-favoured-nation principle of the GATT, GATS and Subsidies Agreement are considered. There is a lot of repetition across Chapters 4, 5 and 6 when the contents of the referred international trade law principles are discussed from a historic perspective (Chapter 4), analysed concerning their scope (Chapter 5), and explored for overlaps and conflicts (Chapter 6). These chapters cover more than 100 pages of the thesis, and I think that they should have been organised in an other way in order to avoid the many repetitions of the actual content of the non-discrimination norms in the trade treaties considered.
Chapter 7 is the best part of the thesis and it concerns different justifications for non-discrimination norms found in the different trade treaties. Here, the author addresses actual trade treaty norms which cover issues on taxation. There are also a few concrete cases from WTO panels and appellate bodies that concern tax law issues including transfer pricing matters, which the author discusses in an eloquent way.
At this stage, it is possible to identify the main drawback of the thesis – disregarding the many repetitions and descriptive parts – which is the meagre discussion of tax law in relation to WTO law. This is a doctoral thesis submitted in the field of tax law (the term “financial law” is used in the thesis, but that is an incorrect translation of the Swedish term “finansrätt” which in Swedish terminology pertains to tax law and customs law; “fiscal law” could also have been used to label the subject in English).
The author discusses both the arm’s length principle and Amount A of the OECD MLC in very general terms. In my view, the discussion on the meaning of the arm’s length principle is so vague that it raises questions about whether the author really tests it in relation to the WTO non-discrimination rules. For example, on p. 176 and the following pages, the author discusses whether the transfer pricing principles are in line with the most-favoured-nation principle contained in Article 1:1 of the GATT. Here, the author states (p. 178): “The arm’s length principle as interpreted in the OECD Guidelines does not favour one country over another. Rather, the arm’s length principle, in its purest form, ensures that profits should be taxed in the jurisdiction where value has been created…”. However, the whole idea behind Pillar One and its step away from the arm’s length principle is largely motivated by the idea that the arm’s length principle does not meet the requirement of taxing where value is created, especially for the large multinational tech companies where consumer activity has altered the concept of value creation. The author should, of course, have explained on what grounds he makes this statement, but he does not. In defence of the author, it should be noted that he occasionally states that Amount A of the OECD MLC pertains to the allocation of the tax base rather than setting the price on intercorporate transfers of goods and services (see, for example, on p. 130 and p. 199). Even if he is correct here, it is not in line with his use of the term “transfer pricing principles” (p. X, Roman numeral, and p. 48) to cover Amount A of the OECD MLC, which does not concern transfer pricing but rather (global) formulary apportionment of corporate profits.
Chapter 8 has the title “Conclusions” and it only spans 14 pages (pp. 277–290). Rather than focusing on the general conclusions of the study, the author primarily provides a summary of the research questions and the contents of the different chapters. In other words, a lot of repetition, again. In the introductory section of this chapter the author states (p. 277): “From spending years analysing hypothetical transfer pricing situations, I was able to build upon this research and include Amount A of the OECD MLC”. I do not question that the author over the years has considered a number of hypothetical transfer pricing situations, but it is a pity that there are so few such examples provided in the thesis, and that the few he actually provides are repeated a number of times.
Furthermore, the author states that it is probably Article III:4 of the GATT (national treatment principle) that has the broadest scope and may also covers transfer pricing principles (p. 284). When countries apply or interpret transfer pricing principles differently, there will be the highest risk of a breach of WTO norms. He also believes that potential conflicts between countries in relation to transfer pricing or Amount A of the OECD MLC will probably be resolved according to dispute settlement mechanism in the tax law field, rather than in the WTO field of law. I agree with him. Finally, I sense that the credo of the whole thesis is for the WTO, OECD/G20 and the Inclusive Framework (IF) to cooperate more on tax as a potential trade barrier, and to work further on this issue and openly publish the results arising from such research and cooperation.
Salehifar, Alireza, The overlap between dispute settlement mechanisms in World Trade Organization agreements and tax treaties, Bulletin, 2019, pp. 50–56.
2 Additional comments on the form and substance of the doctoral thesis
In general, the book is written in a clear and readable English. However, there are language errors, misspelled names and, more seriously, some phrases that are so complex they are difficult to understand.5 An example of the latter can be seen in the formulation of the third research question of the thesis, specifically the fourth element of this research question. This part of the third research question (p. 16) concerns which of the transfer pricing principles contained in the OECD Guidelines, OECD Convention and the OECD MLC that “[c]onflicts with the GATT, the Subsidies Agreement, and / or the GATS”. The use of the expression “and / or” complicates the understanding of the sentence. To illustrate this problem, if one inserts the neutral components A and B, the phrase “A and / or B” means: A or B or both, but it becomes even more complex when it concerns three components (which is the case here), namely “A, B and / or C”, which (I think) should be interpreted as one of A or B or C; or A and B; or A and C; or B and C; or all three of A, B, and C. It is problematic to encounter such complex interpretative issues already in the formulation of the research questions of the thesis.
There are other places where the author’s writing is difficult to understand. In section 1.3, on delimitations, he states (p. 20): “Descriptive text is minimised when analysing the fragmentation of international law, whereby the sociological or political backgrounds that has led to the emergence of a special or specialist rule-systems and institutions is [should be “are”, reviewer’s note] omitted.” Presumably, the author wants to express that he has excluded political or sociological aspects from the study, but this is explained in a very round-about way.
If one takes the author’s view on the general purpose of the thesis seriously, the most important parts of the thesis should be when the arm’s length principle and the calculation and levying of Amount A under the OECD MLC are tested with the different WTO Agreements. However, the reader will be disappointed when reaching these parts of the thesis. On a number of occasions, the same very brief general descriptions of the arm’s length principle and Amount A of the OECD MLC are tested against different WTO norms.6 Here I think that the author could have made use of his experience as a tax adviser on transfer pricing issues and constructed different examples which he could have tested. Now, the analysis is very similar in relation to the different WTO norms, and since the results are almost identical in many situations, it could have been shortened.
The way in which sources are used is important for the evaluation of the quality of a scholarly work. In this respect, footnotes 367 and 368 should be mentioned. They cover large parts of pages 119 and 120 and include refences to a number of cases as well as a large number of articles and books on the topic of WTO law and taxation. However, the items referred to in these two footnotes are not intended to serve as sources for specific statements or views even though some of them are used elsewhere in the thesis. The footnote text is small and therefore difficult to read, and this long list of cases, books and articles seems to serve little purpose, other than to beef-up the Bibliography at the end of the thesis. In an article without Bibliography at the end, such footnotes might be justified, but now they have no legitimate purpose.
The author refers to some academics by their titles, for example Professor (Wolfgang) Schön (p. 46), while some are just addressed with their last name, for example (Michael) Lang (p. 17) and (Hans) Kelsen (p. 73). Of course, references should be consistent, and the easiest way is to ignore the academic titles in the main text of a thesis. The quality of the reasoning is the important thing, not the academic title of the person expressing a view.
The Summary is written in Swedish (“Sammanfattning på svenska”). It is required that a doctoral thesis written in English has a Swedish summary. The Swedish summary merely covers two pages (plus three lines on a third page).7 Considering that this is a doctoral thesis covering 330 pages, this Summary is way too short. It does not cover a full description of the objective of the thesis and its major results. An even stronger weakness concerns the quality of the Swedish in which it is written. It is very hard to understand.
The following sentence from the Summary in Swedish is an example of this enigmatic prose (p. 292):
“’Horizontal coordination’ som tar tillvara på kopplingar mellan liknande eller relaterade aktiviteter som förekommer i autonoma organisationer, som OECD och WTO, och kultiverar en integrativ approach till att angripa en utmaning borde vara ännu mer transparant och genomföras med en kritisk lins på hur reformer kan påverka olika områden inom internationell rätt.”8
Even with the best of intentions I do not understand the quoted sentence, or for that matter large parts of the Summary. A minimum requirement for any thesis is that the text can be understood. During the defence I asked the respondent how the Swedish text in the Summary was produced, however, I did not get a clear answer.
It is common to describe the research process in humanities and social sciences with the dichotomy between the process of discovery and the process of justification. In legal scholarship, the written word of the thesis constitutes both the means and goal of the research. In writing the thesis, the doctoral candidate investigates the research area and makes the analysis of the legal sources, identifies patterns and provides answers to the research questions. When the analysis is performed and the author reaches a better understanding of the subject, the objective, research questions, the analysis and the general outcome of the thesis changes. That is only natural with good research. The author has indeed worked with his objective and research questions and come to several findings. However, he has not amended the text in order to make the reading as easy as possible. Chapters 2 and 3 are largely descriptive and Chapters 4, 5 and 6 are more analytical but contain many significant repetitions. A thorough re-writing of these chapters would have made the thesis considerably more readable.
Another major problem with the thesis is that the author’s considerable competence as a skilled tax law adviser is not sufficiently used. There is very little discussion on the meaning of the arm’s length principle and how its application in different situations could result in a breach of the national treatment obligation and the most-favoured-nation obligation under investment treaties within the WTO framework. The author provides one brief example of how different states apply the measure of corresponding adjustments under Article 9.2 of the OECD MTC (and tax treaties based on it) in different ways. However, more can be expected from a doctoral thesis in tax law. The author works part-time as a tax adviser with an accountancy firm, and here – I think – he could have made use of practical examples inspired by his part-time work. The area of transfer pricing is to a large extent a practical affair, where specific facts of the individual cases play an important role in establishing the arm’s length price. However, that aspect is hardly taken into account in the thesis.
The other tax legislation that is considered is Amount A of the OECD MLC. An agreement on this matter exists within the OECD and its Inclusive Framework. However, it has still not been decided and implemented by the signatory jurisdictions. In a doctoral thesis, however, this does not pose a fundamental problem. Even hypothetical treaty provisions can be analysed and tested, with the results being open to discussion.
For example: On page 29 he refers to “author(s)” where he probably is referring to himself, and the reference should of course be to the “author”. On page 30 he uses the adjective “necessary” which should be the adverb “necessarily”. On page 31 he states that Article 31 of the Vienna Convention “directs” which probably should be “stipulates”. On page 37 he refers to “selection basis” which probably should be “selection bias”.
See for example the description of the arm’s length principle and Amount A of the OECD MLC on pp. 126–130, pp. 135–137, pp. 178–179 and pp. 198–199.
Op.cit. pp. 291–293.
Op.cit. p. 292.
3 Overall evaluation of the doctoral thesis
The subject of the thesis is too large. The research questions should have been more focused.
The subject matter is important and there is not that much research in this area so far, partly because the subject is very demanding. The thesis is submitted as a thesis in fiscal law (tax law) but the elements of a fiscal law discussion in the thesis are very limited. Instead, the thesis reads more like a survey of different parts of WTO law with a rather brief discussion on some tax law aspects at the end of the discussion of the WTO agreement at issue.
Research in this field of law is strongly required. However, I think the topic is too sophisticated a subject for a doctoral thesis in fiscal law, at least for someone that works part-time as a tax law adviser in parallel with undertaking his doctoral studies.
In conclusion, it is a weak doctoral thesis, basically because the subject is too broad and the analysis of the interaction between WTO law and fiscal law is meagre and repetitive. While the thesis reaches the level for approval, it does so by a very narrow margin. The thesis was unanimously approved by the examination board.
From the respondent’s perspective, however, I think he has made an achievement by managing to complete his doctoral studies and the doctoral thesis, and I congratulate him to that. Not many can achieve this while at the same time maintaining a professional career as a tax law adviser.
Mattias Dahlberg is professor of fiscal law with the Faculty of Law, Uppsala university.