Transfer pricing: changing views in changing times

ABSTRACT Transfer pricing for tax purposes has long been contentious, but recent political and public concerns about tax avoidance have energised critiques of current rules and debates about proposals for change. Transfer pricing tax rules are underpinned by the arm’s length principle and we consider how the common understanding of this principle has developed and changed, as criticism of it has grown and there have been increasing calls for change. In this qualitative study we draw on Bourdieusian concepts: we focus on the views of senior transfer pricing professionals relating to the UK and the US and consider how their views and transfer pricing practices have changed over a period of field disruption. This is important because calls for transformation of the field need to be cognizant of the extent to which existing practices are firmly embedded and thereby resilient to change. We find that over the period of our longitudinal study the senior transfer pricing professionals show a degree of adaptability to the use of the arm’s length principle, which continues to dominate.


Introduction
The corporation tax practices of multinational enterprises (MNEs) have been brought into the public gaze in recent years; much of the publicity has focussed on transfer pricing. Increased media and political attention has turned tax transfer pricing from an arcane technical issue, not well understood outside the close circle of practitioners for whom it is everyday practice, to an issue now appearing in public discourse, fuelled by the attentions of civil society organisations. Recent developments provide us with an opportunity to explore the transfer pricing field from a different perspective, providing insights into wider questions of how fields of practice shift and adapt over time.
The arm's length principle (ALP) is the basis of the tax transfer pricing rules in most countries and is used to determine an arm's length transfer price: the price that would be used if the same transaction were undertaken by unrelated third parties. Transfer pricing has become a sub-specialism within tax practice which is increasingly significant: it requires specialist expertise relating to different jurisdictions. Therefore, in addition to being local professionals, transfer pricing specialists occupy transnational spaces . These experts possess practical experience that enables them to apply the relevant transfer pricing tax rules and procedures to determine and demonstrate whether or not there has been compliance with them and with the ALP (Picciotto, 2015), in the allocation of profits between countries. Concerns about the uncertainty of the arm's length price under this principle have called into question the robustness of the ALP, particularly in relation to new business structures and practices. This has led to stronger calls for a fundamental change to formulary apportionment, (Rogers & Oats, 2019) whereby the combined profits of a group of MNEs are allocated, according to a fixed formula, to each taxing jurisdiction. 1 A shift away from the ALP, and the implementation of change, requires the involvement of transfer pricing practitioners.
We study the work of senior transfer pricing tax professionals over a 10-year time frame. The period of time for this analysis is anchored by fieldwork undertaken by one of the authors, comprising interviews with senior transfer pricing professionals in 2006/07 and 2016/17. This captures key moments in the shifting perceptions of transfer pricing, allowing us to consider changes in attitude over a period in which transfer pricing uncertainty has been heightened, and increased in its legal and political salience (Dallyn, 2017). Our analysis focuses on the response of elite transfer pricing professionals to one specific aspect of transfer pricing, the mechanism for allocating taxable profits between countries, and is grounded in the understanding of transfer pricing as a field of practice, inspired by Bourdieusian concepts. In this recent time frame, there have been many changes in the field, yet the ALP appears to remain impervious to these changes. To explore this phenomenon, we ask two questions: firstly, how has the common understanding and knowledge of the ALP developed? and secondly, linked to this, how have senior transfer pricing professionals' views of the ALP changed? These questions matter because any suggestions for change need to be cognizant of the extent to which existing practices are firmly embedded and thereby resilient to change. This paper thus contributes to the ongoing debate about changes to the ALP, recognising tax as a social practice, at a time when there have been calls for more empirical work drawing attention to the experience of tax professionals (Radcliffe et al., 2018) and for qualitative studies of transfer pricing (Cools, 2014;Rossing, 2013). There have been calls for research looking at the ways in which field level changes impact on professionals (Radcliffe et al., 2018;Suddaby & Viale, 2011) and the longitudinal nature of the study offers insights into the impact of significant changes in the field on the professionals we study. Our analysis therefore adds to the understanding of the dynamics of professional knowledge under conditions of uncertainty and disruption.
The remainder of the paper is structured as follows. In the next section we set out our theoretical position. We then discuss the increasing importance of tax transfer pricing rules and the ALP and discuss the transfer pricing literature. Following this we present the research method. The next section presents and discusses our interview findings and is followed by our conclusions.

Theoretical considerations
In order to make sense of the shifts taking place within the transfer pricing field, we draw on Bourdieusian concepts to understand how logics and practices evolve over time (Suddaby et al., 2007) resulting in new conflicts as the field simultaneously becomes more porous to external influences and more clearly defined as dominant agents cement adherence to pre-existing logics. Bourdieu argued that in order to understand an event or social phenomenon, it is not enough to look only at what is said and done (as cited in Thompson, 2008, p. 65); we also need to consider the social space within which interactions between agents are located. In the light of this we start by considering the field of transfer pricing.
Transfer pricing for tax purposes is a social practice within the broader field of international taxation. We conceive of "field" as being dynamic arenas of struggle, occupied by actors, individuals, organisations and even institutions, with varying degrees of dominance (Everett, 2002;Gracia & Oats, 2012). As a field of practice, transfer pricing is intertwined with other fields; the broader field of taxation, the juridical field in which laws are constructed, implemented and practiced, the bureaucratic field inhabited by those who work for the tax agencies and the professional field which enables and constrains the actions of actors who carry professional qualifications and dispositions.
Each field has certain regularities and field actors, who, as with players in a game, concur in their belief in the game (Bourdieu & Wacquant, 1992), and the pursuit of a specific goal (Bourdieu, 2005). As Oats and Morris (2015, p. 34) observe: the purpose of rugby union as a game is to win through scoring tries, conversions and penalties in accordance with the laws of rugby union, in contrast football (soccer) as a game provides a set of rules in accordance with which teams play to score goals. Rugby league and football are different "fields" with different "doxa"; each field and the players on the field abide by "different rules of the game".
Fields are not, however, deliberately constructed, but rather emerge as actors coalesce around valued capital and become "taken in by the game", having an investment in it (illusio) and concurring "in their belief (doxa) in the game and its stakes" (Oats & Morris, 2015, p. 34). As players, field actors have to believe in the game and recognise what is at stake, agreeing (collusio) "albeit sometimes tacitly, that resources are worth competing or fighting over or the game would simply cease to exist" (Ibrahim, 2015, p. 43). Ilusio thus ensures "an unreflexive commitment to reproducing the rules of that game" (Lupu & Empson, 2015p. 1312. The interplay between the rules of the game and the playing of the game is mediated by habitus (Spence & Carter, 2014). Field actors carry with them a habitus which is a "set of attitudes, values and behaviours that dispose agents to behave in particular ways" (Gracia & Oats, 2012). Bourdieu posits that habitus is central to the "continual reproduction of belief in the game, interest in the game and its stakes" (as cited in Oats & Morris, 2015, p. 33). Habitus generates behaviours that are suited to the conditions under which it operates; its effect is largely invisible.
The specific efficacy of habitus can be clearly seen in all the situations in which it is not the product of the conditions of its actualisation..[for example] when old people quixotically cling to dispositions that are out of place and out of time … Such effects of hysteresis, of a lag in adaptation and counteradaptation mismatch, can be explained by the relatively persistent, though not entirely unchangeable, character of habitus. (Bourdieu, 2005, p. 86) As Anesa et al. (2019) observe "Bourdieu suggests that actions result from the embodiment of social structures that suggest pre-reflexively to an individual what is legitimate" (p. 3). According to Bourdieu (2005) habitus leads to "reasonable (not rational) expectations" (p. 86), as a result of the internalisation of experience of recurring situations, leading to practical mastery but importantly also, practical anticipation of future situations. The emergence of a common body of understanding allows field actors with a habitus that closely aligns with the field to not only display mastery in the present, but also to anticipate shifts and adapt accordingly.
The dominant within a field exert pressure on the dominated: "they define the regularities and sometimes the rules of the game, by imposing the definition of strengths most favourable to their interests and modifying the entire environment" other actors and the "systems of constraints that bear on them or the space of possibles offered to them" (Bourdieu, 2005, p. 76). "The forces of the field orient the dominant towards strategies in the perpetuation or reinforcement of their domination" (Bourdieu, 2005, p. 80). In the transfer pricing field, a small number of senior professionals form a superordinate group: the transfer pricing elite.
Institutions "structure the struggles that characterise fields" (Ahrens & Ferry, 2018, p. 13), such that institutional arrangements, being vulnerable to political action, can shift with repercussions for specific fields. Fields are thus vulnerable to changes taking place in external realms. The structure of fields can also be affected by the appearance of new, effective agents (Bourdieu, 2005) commonly agents who are dominant in other fields, such that "changes within a field are often linked to changes in the relations with the exterior of that field" (p. 80). Having said that, the interrelationship between different fields is complex and constantly changing; the extent to which changes in one field precipitate changes in another will vary over time.
Field struggles are over forms of capital and agents jockey for positions within the field (Rey, 2007). Scholars have identified a wide variety of forms of capital as refinements of Bourdieu's four forms of economic, social, cultural and symbolic capital. Each individual field has forms of capital specific to it, such as "sacraments and divine sanction in the religious field" (Rey, 2007, p. 50). As Maclean et al. (2014) note, "capital formation is an ongoing, dynamic process, subject to accumulation or attrition" (p. 828). For the purpose of this paper, an important form of capital in the transfer pricing field is practical knowledge which confers on its holders a power that reinforces professional power. In the hands of the superordinate elite within the field, extensive practical know-how becomes symbolic capital in that the field as a whole and confers legitimacy on it as a marker of status. Holders of this form of capital draw on the professional field where control of knowledge is valuable. Indeed, Covaleski et al. (2003), suggest that control of knowledge is a crucial element of professional power: in particular abstract knowledge is a key factor in maintaining professionals' overall power and this power is particularly strong in respect of "problems to which treatment can be applied only once … .performance can never be directly compared … powers to define a problem, to measure its treatment, and to escape comparison are preliminary weapons of professional force" (Abbott, 1988, p. 137).
Within the wider body of literature on professional power, the importance of knowledge and control over knowledge is a recurring theme and knowledge based on theory is seen as a defining characteristic of professions (Abbott, 1988;Covaleski et al., 2003;Leicht & Fennell, 2001): in particular it is command of abstract, tacit knowledge that gives rise to professional authority and credibility. At the level of the individual, professional power is increased when knowledge is valued by different organisations and is portable. A sign of the most powerful professionals is the ability to control the portable knowledge they possess by "blackboxing" it (Fincham et al., 1994), that is, selecting when and where to transfer the knowledge, taking it to other environments where it is highly prized. It has been noted that "the most politically and economically successful expert groups are those who are able to blackbox their expertisethat is compartmentalise key elements of their knowledge base" (Reed, 1996, p. 583) and take it to where it is particularly valued. The ability to switch sites of practice with the blackboxed knowledge serves to further enhance the value of the individual professionals, as they then acquire knowledge and experience of different organisations which they can blackbox and take elsewhere. In the next section we explore the historical backdrop to the transfer pricing field which allows us to trace the growing importance of blackboxed, tacit knowledge in securing the dominance of the transfer pricing elite.

Transfer pricing as a dynamic field
Transfer pricing as a technical term refers to the pricing by an MNE of its intragroup transactions in order to attribute revenue and costs to each component (Groot & Selto, 2013). The term is increasingly, however, being used as an example of a tax avoidance technique, particularly in recent years in the context of reports of the transfer pricing arrangements of particular large MNEs that involved income shifting to low tax jurisdictions (for example Addison & Mueller, 2015;Apostol & Pop, 2019). While this shift in perception of the meaning of transfer pricing forms part of the dynamics of the transfer pricing field, in this paper we focus on transfer pricing in the former, technical, sense; as a necessary mechanism for determining the taxable profit of an MNE in each taxing jurisdiction, in order to comply with the different tax rules and practices of different countries. Our first research question is: how has the common understanding and knowledge of the ALP developed? We divide the discussion of this into sections to provide context on the transfer pricing environment at the time of the research interviews. We then discuss other aspects of the transfer pricing literature.

The ALP and the background to the first round of interviews
The OECD has been one of the most important influences in diffusing the ALP, through transfer pricing guidelines (Eden, 2019). However, tax rules based on the ALP were not widely enforced until the 1990s, when the US enacted new transfer pricing rules, supported by a stringent enforcement effort (Durst & Culbertson, 2003), triggering a response by the OECD (Radaelli, 1998): in 1995 final revised Transfer Pricing Guidelines were reissued (1995 OECD Guidelines), affirming the ALP as the predominant profit allocation mechanism and setting out methods for determining an appropriate transfer price. The preferred method in the 1995 OECD Guidelines is the comparable uncontrolled price method, although in practice, with a very small number of exceptions, there is not a directly comparable price (Durst & Culbertson, 2003). In the absence of an appropriate arm's length comparable, the 1995 OECD Guidelines listed possible methods to determine the arm's length price. The fifth and final method listed was the profit split approach, which allocates profits, described as the method of "last resort" (OECD, 1995, p. 3.5). Some countries objected to the presence of the profit split method, seen as akin to formulary apportionment, (Avi-Yonah & Pouga Tinhaga, 2017), but the risk of failing to harmonise the OECD Guidelines and the US regulations, so losing international consensus, was a driving force in its inclusion (Radaelli, 1998). Accordingly, in Bourdieusian terms, collusio emerged and "apparent consensus was maintained, by the insistence that all these methods comply with the arm's length principle" (Picciotto, 2015, p. 179).
The 1995 OECD Guidelines recognised that the ambiguity of the ALP requires judgement, acknowledging that its guidance on the ALP is supplemented by practice: Experience under the arm's length principle has become sufficiently broad and sophisticated to establish a substantial body of common understanding among the business community and administrations. (para1.14) This undocumented, informal "body of common understanding" which becomes established over time, is described as being "of great practical value in achieving the objectives of securing the appropriate tax base in each jurisdiction" (para 1.14). Therefore there is explicit acknowledgment that the 1995 OECD Guidelines offer an imprecise principle, which is underpinned by the emergence of an external, undocumented, common understanding shared by those working in the field, and senior transfer pricing professionals are a channel for this. At this point then, the OECD is complicit in cementing the ALP as transfer pricing's cornerstone, and thereby the power of those professionals in the field whose habitus aligns with the field illusio and are dominant by virtue of their superior capital in the form of practical knowledge of the ALP's application; and this dominance is linked to the imprecise ALP.
There was then a rapid increase in transfer pricing rules and requirements, often inconsistent, in different countries. This "arm's race of new international transfer pricing obligations" (Collier & Andrus, 2017, p. 86), could be thought of as an opening for professionals to advance by moving into a new space, expanding their expertise for new areas of practice (Anand et al., 2007) "which enable pioneers of the professions to colonize new territories of influence" (Dezalay, 1995, p. 334). The transfer pricing profession emerged as a sub-specialism of the wider corporate tax field, with an increasing number of journals and conferences, used to develop and communicate interpretations and understandings of recent developments, (Sakurai, 2002) thus contributing to the development of a "body of common understanding." Transfer pricing advisers offered their experience to help taxpayers determine appropriate transfer prices. Failure to agree that transfer prices are arm's length, can give rise to disputes: these can be between taxpayers and tax authorities, or between different tax authorities. 2 For example, in September 2006, Glaxo, the pharmaceutical MNE, and the IRS agreed to a $3.1 billion settlement, following a lengthy transfer pricing dispute, relating to the level of profits Glaxo reported in the US after the US company made payments to its UK parent in respect of intangibles and trademarks. This largest tax settlement in IRS history (Kollewe, 2006) demonstrates the significant impact that can arise from different views of the appropriate transfer prices in complex situations. It was against this backdrop that the first round of research interviews took place.

Disruption in the field and the background to the second round of interviews
Throughout the period between the two sets of research interviews the international business environment continued to evolve and change, impacting on the transfer pricing environment, since theoretically the transfer prices should reflect the way the business is structured and operates. There has been a continuing rise in integrated MNEs, increasing the challenge of applying the ALP, as within an integrated MNE there are synergies that are not present in unrelated companies (Avi-Yonah et al., 2009). Over this period the economy became increasingly digital, significantly impacting the business environment and companies,  with innovations such as robotics and cloud computing, where it is particularly challenging to apply the ALP (Mazur, 2016). Digitalisation is seen as opening up borders so it is easier for companies to sell products in a particular jurisdiction with minimal physical presence there, and it can lead to significant economies of scale and scope for integrated companies . The challenge of determining the location of a particular activity for tax purposes becomes harder in a digital world (Devereux & Vella, 2018).
The financial crisis is thought to mark a turning point after which the regulatory environment for large businesses became more intensely scrutinised and entered the public imagination (Anesa et al., 2019;Picciotto, 2020). Initially a key focus for media coverage was tax planning by large technology companies such as Apple, reported to have achieved non-taxation of profits of US$44 billion (Ting, 2014), but subsequent publicity extended to different sectors (Brauner, 2014), for example in the UK, publicity of Starbucks' low corporation tax payments in the UK, and subsequent calls for public boycotts (Thompson & Houlder, 2012) was a viewed as a "UK tax public relations disaster" (Kleinbard, 2013(Kleinbard, , p. 1516. While public interest in corporate tax matters was starting to increase, the OECD issued revised Transfer Pricing Guidelines: much was unchanged, apart from the removal of the priority for comparable price and profit methods no longer being a last resort (Avi-Yonah & Pouga Tinhaga, 2017). This shift demonstrates how the formal rules were changed to reflect the body of common understanding of the application of the ALP in practice. Thus the "international consensus" (OECD, 2010: B.2) on the ALP was reinforced.
The increasing media reports of alleged corporate tax avoidance, led to the calling by the G20 for work on this, and the initiation of the Base Erosion and Profit Shifting (BEPS) project, carried out through the OECD, began in in 2013, with a key aim of preventing non-taxation (Picciotto, 2017a). This can be viewed as an incursion on the autonomy of the transfer pricing field and the growing porosity of its boundary with the political field. The OECD published the final BEPS report in 2015: the stated aim was to create transfer pricing outcomes in line with value creation. The invocation of yet another ambiguous concept, value creation, to form part of the corpus of ALP practices arguably serves to buttress the continued dominance of ALP but at the same time heralds a subtle shift in the expertise required to navigate the ALP maze. 3 It was against the backdrop of the BEPS final report that the second round of interviews commenced.
Subsequently, the OECD released amended OECD Transfer Pricing Guidelines (OECD, 2017) consolidating the BEPS amendments and including a restatement of support for the ALP. Individual countries then began modifying their domestic tax laws to take account of the BEPS recommendations; these revised rules remain subject to varied interpretations (Picciotto, 2020), preserving the overarching logic of existing struggles in the transfer pricing field.

Transfer pricing literature
Transfer pricing is explored in the literature of different scholarships. Here we consider literature from management accounting corporate tax avoidance literature and the growing body of literature calling for an alternative to the ALP, specifically formulary approaches.

Management accounting
The growing body of management accounting literature on transfer pricing reflects the changes and developments in the field. Previously much of the literature was focused on management control systems for example resource allocation, how to set transfer prices optimally and performance evaluation (Eccles, 1985) and goal congruence (Emmanuel & Mehafdi, 1994). Following the new tax rules from the mid-1990s onwards there were calls for research to take account of tax rules and requirements, (Cools & Emmanuel, 2007). Subsequent management accounting research highlighted the importance of being cognizant of the tax rules when setting transfer prices in a "tax compliant" MNE (Cools et al., 2008) including transfer prices for services (Rossing & Rohde, 2010). The implementation of management control systems needs to take into account the tax requirements for MNEs to comply with transfer pricing tax rules and defend their transfer prices (Jost et al., 2014) and avoid double taxation (Cools et al., 2008). A detailed risk and functional analysis of the different parts of the MNE group is required to be undertaken, requiring the maintenance of large amounts of documentation; this is "costly and time-consuming work" (Cools et al., 2008, p. 606). There is much technical literature that is concerned with guiding MNEs on how to comply with different rules and defend their transfer prices to the tax authorities (for example, Anderson & Miall, 2004;Zetter et al., 2009).
Research studying "a transfer pricing tax compliant multinational enterprise" (Cools et al., 2008;Rossing & Rohde, 2010) is suggestive of a situation where corporate taxpayers can be identified as either tax compliant, or non-compliant, rather than one where there is pervasive uncertainty as to how to fully comply. However, within this management accounting research there is a recognition of the difficulty in ensuring the operation of "tax-compliant transfer pricing": whereby companies "do not experience transfer pricing tax compliance as an easy task" (Cools et al., 2008, p. 606).

Corporate tax avoidance
Against the backdrop of media coverage and increasing public concern, corporate tax avoidance is a topic gaining increasing academic attention. Until recently much of the research had a quantitative approach, focussing on statutory corporate income tax rates, seeking to identify profit shifting (Finér & Ylönen, 2017) (for example Barterlsman & Beetsma, 2003;De Boyrie et al., 2005;Hines & Rice, 1994;Huizing & Laeven, 2008). However, recently there have been studies using a qualitative approach (for example Anesa et al., 2019).
There are differing views on the meaning and acceptability of tax evasion, tax avoidance and tax minimisation (Hanlon & Heitzman, 2010) and accordingly "there is no consensus on how much corporate tax should properly be paid in relation to the statutory corporate tax rates of particular countries" (Ylönen & Laine, 2015, p. 9). However, the issue of companies' ability to reduce their corporate tax liabilities has been receiving increasing media attention, with an increasing role for the work and reports of NGOs such as Oxfam and Tax Justice Network (Dallyn, 2017). Accordingly, while the risk of double taxation for companies is acknowledged, this body of the literature focusses on non-taxation, where some of an MNE's profits are not taxed in any jurisdiction (Kleinbard, 2013;Wells & Lowell, 2011), the incidence of which was seen by some to be "rife" (Devereux & Vella, 2014, p. 451). Although many continue to regard tax minimisation as legitimate (Anesa et al., 2019), it is increasingly being presented as a moral issue  with calls for increased transparency (Oats & Tuck, 2019) and increasing concern about reputational damage in the face of "an emerging ethical logic" (Apostol & Pop, 2019, p. 2).
Within the literature on corporate tax avoidance, recent studies have looked at the role of tax professionals (Addison & Mueller, 2015). Tax advisers are seen as occupying a significant role in facilitating taxpayers' tax avoidance (Sikka, 2010) or taxpayers' compliance (Anesa et al., 2019). There has been increasing focus on transfer pricing and the part it can play in tax minimisation (Anesa et al., 2019;Radcliffe et al., 2018;Sikka & Willmott, 2010). There is discussion of the difficulties of considering hypothetical pricing arrangements and the distorting fiction of separate entities, rather than the reality of a single integrated economic unit (Finér & Ylönen, 2017). Within this literature there is discussion of the ambiguous nature of the ALP, which can facilitate tax avoidance, and recognition that "the BEPS initiative has not reduced the room that tax professionals as a whole have for manoeuvre" (Radcliffe et al., 2018, p. 49).

Calls for formulary apportionment
Many writers saw the BEPS project as a missed opportunity for radical change (Devereux & Vella, 2018;Picciotto, 2017a). The ongoing debate on BEPS has been seen as resulting in two broad proposals for transfer pricing reform, either amending the system based on the ALP, and a formulary apportionment approach (Eden, 2019;Sikka, 2017). Making amendments to the system is seen as problematic as it becomes increasingly complex, requiring complex detailed supporting evidence to justify transfer prices, for example with reference to economic analysis of databases (Avi-Yonah et al., 2009). Even after this, uncertainty remains as to whether the transfer prices will be judged to be appropriate by each relevant tax authority: this is particularly problematic within highly integrated MNEs, where there are interdependencies and transactions that would not arise between unrelated third parties (Clausing & Avi-Yonah, 2007).
Against the backdrop of the rapidly changing transfer pricing environment, including more complex rules and requirements there has been an increase in calls for some form of apportionment of profits to replace the ALP (Rogers & Oats, 2019) including for example : Clausing & Avi-Yonah, 2007;Durst, 2010;Devereux & Vella, 2014;Picciotto, 2017b. The ALP can facilitate profit shifting, viewed as having a particularly detrimental impact on developing countries (Hopper et al., 2017;Sikka, 2010). It is posited that these countries need easily administrable tax rules, rather than a system based on the judgement of highly trained professionals (Picciotto, 2017a). Furthermore, increasing globalisation and digitalisation is put forward as a further reason for a shift from the ALP to a formulary approach, as digitalisation has led to an increase in the ease with which profits can be shifted to low tax jurisdictions (Mazur, 2016). Within this stream of the literature there is recognition of the difficulties in implementing a formulary approach, including the difficulty in obtaining political agreement (Picciotto, 2017b) and the risk of factor shifting instead of profit shifting, (Fleming et al., 2014); however these difficulties are seen as a preferable alternative to the problems of the ALP.
The different strands of the transfer pricing literature speak of the increasing challenges and difficulties of applying the imprecise ALP in the context of inconsistent national approaches. The proliferating rules meant that there was a move from discussion of the MNEs that are tax compliant to MNEs that were grappling with an increasingly complex system, sometimes as a route to profit shifting, giving rise to increasing calls for change. Against the context of a changing transfer pricing environment subject to increasing scrutiny and calls for change, our second research question is: how have senior professionals' views of the ALP changed? To explore this, using a Bourdieusian lens, we draw on interview data that reveals ways in which the attitudes of senior transfer pricing professionals to the ALP have altered during this period of disruption.

Research method
There have been many new transfer pricing rules introduced from the 1990s onwards and extensive documentary analysis was undertaken of official documents and statements, including the OECD Guidelines, OECD statistics, national legislation, and guidance statements from tax authorities. Beyond the formally documented rules, the OECD Guidelines refer to a "body of common understanding" and we were seeking to gain an insight into how this was drawn on and shaped in practice. This paper adopts a qualitative approach, as this facilitates understanding the perspective of those being studied (Pratt, 2009). We draw on data from semi-structured interviews; this approach is consistent with previous studies of tax professionals and their work (Anesa et al., 2019;Mulligan & Oats, 2016;Radcliffe et al., 2018). Interviewees were asked the same open-ended questions; responses could then be followed up as the interview took place (Gendron, 2009). The interview data we present has a longitudinal dimension, and so has the advantage of exploring the historical context and appraising the data within that context (Morrell & Tuck, 2014). A total of 34 interviews were conducted during 2 rounds of 17 interviews each. The first round commenced in 2006, most interviews were conducted in that year, with 2 conducted in 2007, (for convenience here we use the phrase the 2006/07 interviews to refer to this round). The second round commenced in 2016, with most conducted in 2016 and 2017, with 3 conducted at the start of 2018 (for convenience here we use the phrase 2016/17 interviews to refer to this round). At the start of each interview interviewees were asked for their individual perspectives and advised that quotations could be drawn on but that these would be unattributed. Therefore, interviewees have been assigned a random interview number which is used as a reference when direct quotes are provided in the subsequent analysis and discussion (e.g. R1-1 denotes an interviewee from round 1, randomly assigned the number 1). The term arm's length principle, ALP, is preferred and used by interviewees from the UK and the OECD, while the term arm's length standard, ALS, is preferred by those from the US. In the quotes presented in this paper, the term ALP has been substituted for ALS, where necessary, both for reasons of consistency and for preserving anonymity.
In 2006 the focus on the UK and the US was chosen as both the US and UK had well developed transfer pricing laws and administrative practices (Barterlsman & Beetsma, 2003). Although the national approach to transfer pricing differs between the UK and US (Sakurai, 2002), we were seeking the views of interviewees in their capacity as transnational professionals. Interviews were sought from some of those working in transfer pricing in respect of the UK and the US at the most senior level: from tax authorities, the OECD, advisory firms and corporate taxpayers. It was recognised that is not an easy group to gain access to (Currie et al., 2015;Mulligan & Oats, 2016). Those working on transfer pricing in the most senior levels of these organisations were identified through studying international transfer pricing practitioner journals and transfer pricing practitioner conference proceedings and contacted with a request for an interview for academic research. Participation from some further interviewees was obtained, as a consequence of snowballing, following suggestions from interviewees.
Given the significant changes in the international tax and transfer pricing environment in the years after the first interviews (Anesa et al., 2019;Radcliffe et al., 2018), we sought to revisit the issues discussed a decade later, by conducting further interviews, to explore the perceptions of changes in the international tax environment. Requests were made to the initial interviewees who were still working in international tax and transfer pricing for a further interview. Nearly half, 8, of the 2006/07 interviewees agreed to be interviewed again, providing an element of continuity to the study. The other 2016/17 interviewees were individuals identified as occupying roles that had been vacated by some of the 2006/07 interviewees who had moved on. In some cases, access was facilitated by the 2006/07 interviewee who had moved on, suggesting the contact.
The table in the Appendix shows the organisation that interviewees were based in at the time of the interview. Most interviewees had previously worked in a different key organisation, indicating much movement of staff in transfer pricing at senior levels (Borkowski, 2005) which is suggestive of individuals blackboxing knowledge and experience, which they take to different organisations. Accordingly, most interviewees had experience and understanding of transfer pricing across different organisations, indicating a good understanding of different contexts with respect to transfer pricing, and bringing a further richness to the interview data.
Our research approach involves the analysis of qualitative data, transcripts from semi structured interviews, in an interpretive manner (Silverman, 2010), from an identification of themes from the data, (Miles & Huberman, 1994). The transcripts were read line by line to identify initial themes, noted by using open codes. Phrases, or paragraphs were reviewed to study the interviewee's meaning, rather than breaking them down into individual words and phrases, making use of Word software and NVivo as a tool to assist data management. These open codes facilitated the development of categories for further attention. This was an iterative process and rereading the transcripts and coded reports facilitated the emergence of many themes; we draw on this rich data, including on interviewees' own words, analysed by reference to the theoretical concepts of Bourdieu.

Findings and discussion
Transfer pricing is a sub-field of the international tax field in which MNEs struggle with different tax authorities from the countries in which they operate to secure agreement as to the share of profit falling to be taxed by each country. Given the lack of a clear rule, those working in the field of transfer pricing require substantial practical knowledge and experience accumulated over time, including not only the theoretical application of ALP but also agreements reached in practice. This practical knowledge is valuable capital in the transfer pricing field and the holders of this knowledge are able to exert professional power by blackboxing it and securing dominance. Collectively, the transfer pricing profession is able to offer knowledge and experience to deal with new requirements, and expertise to help to mediate transfer pricing uncertainty and operate a workable system. As the transfer pricing field solidifies over time, becoming more autonomous vis a vis the wider tax field as a specialised area, the struggles for domination within the field increased.
Although from different backgrounds, the senior transfer pricing specialists possess shared knowledge and experience and therefore habitus. This knowledge and understanding, and the means to disseminate it, form part of the capital of senior transfer pricing professionals which serves to embed their dominance in the field: this is important because, if there no illusio of "appropriate" transfer prices, the system would become unworkable. Transfer pricing rules are based on an ambiguous principle, supplemented by an acknowledged unwritten body of common understanding, which is drawn on and applied to complex, fact intensive situations. This is an important source of power, as power is enhanced, both at the level of the individual professionals and the community as a whole, when linked to a body of technical knowledge that is not easily mastered. If the body of knowledge is ambiguous and uncertain, requiring complex techniques and nuanced judgements, this serves to protect and buttress the professionals' power and authority, and for the elite professionals, field domination. Furthermore, the application of the ALP in fact intensive situations, to determine the appropriate transfer prices, means each particular case will be slightly different and so cannot be directly compared, a key weapon of "professional force" (Abbott, 1988).
In the following section, we discuss, drawing on senior professional interviewees' observations and perspectives, how a discernible shift has occurred in the decade of disruption under consideration, and interviewees' attitudes to the disruptors.

2006/07 The ALP is seen as bedding down
Just over a decade after the publication of the 1995 OECD Guidelines, interviewees spoke positively about the ALP. A key advantage of the ALP put forward by interviewees was its flexibility, enabling it to take into account the different circumstances of each case. It was seen as a principled way of determining the prices to be used between related parties, reflecting economic reality by considering what unrelated parties would do in similar market conditions. In this regard, the ALP was not presented as a difficult approach "what we're after here is to replicate the market, it's not the most technically difficult" (R1-4). One interviewee saw increasing acceptance of the ALP, envisaging that it would work better in the coming years: the ALP is really bedding down … I think the general understanding of it is improving so that you get far less sort of outliers, I think it's maybe less political than it's been. (R1-12) This "bedding down" is indicative of a strengthening consensus among the dominant field actors. This suggests that over time, tacit knowledge, the body of common understanding, was developing into a shared view of how the flexibility should be accommodated in practice. The flexibility of the ALP in the OECD Guidelines was spoken of positively; an interviewee commented: they permit you to do anything … ..I find them to be pretty flexible in terms of they don't dictate much of anything. (R1-4) The use of a profit split method was reported as becoming increasingly commonplace, despite this approach not being based on arm's length comparable prices and being described by the OECD, 1995 Guidelines as a last resort: when the guidelines were first introduced, back in the 1990s, there was a very strong push to reject so called profit-based methods in favour of traditional based methods and that shows up in the guidelines. The guidelines ultimately approved of them but in relatively narrow circumstances. The kind of methods of last resort as it described them. Well, with time having gone by, and the evidence is that these are not methods of last resort. (R1-16) While this was a move away from the use of comparable transactions, interviewees still considered it to be consistent with the "pretty flexible" ALP: profit splits can be arm's length, you're saying if two third parties were making this contribution, how would they split it? It's kind of arm's length, but it's hard to find comparables for it. (R1-2) The increasing use of the profit split method since 1995 could be viewed as an example of the flexibility of the ALP. The subsequent 2010 OECD Guidelines removed reference to profit methods being the method of last resort, according them equivalent status to the traditional pricing methods (Avi-Yonah & Pouga Tinhaga, 2017), showing how the body of common knowledge can develop and be applied in practice, ahead of being formally documented by the OECD in its Guidelines.
A disadvantage of the ALP is the potential for double taxation. In addition to high profile cases such as the Glaxo case settled in 2006, it was commented that: minor double taxation occurs every day because a lot of those remedies are uneconomic … at that point double taxation becomes a cost of doing business. (R1-17) Despite acceptance of this disadvantage, in the context of reiterating the ALP's advantages of flexibility, interviewees expressed opposition to formulary apportionment; it was not considered to be a principled approach, departing as it does from individual market and economic circumstances, in addition to being inflexible. Several interviewees rejected it without further discussion: "I have nothing against global formulary apportionment, it's just that it doesn't work" (R1-15), and "the formula doesn't really attempt to measure economic value" (R1-3). It was also posited that formulary apportionment would not resolve the problem of double taxation: you get double taxation because countries have different views of the economic value of things, which is why you get double taxation, that's one of the reasons you won't get formulary apportionment because it's the same, you will get far more double tax, one, you've got to agree on the base to be split as well, that is also a nightmare, whose accounting rules do you use? so, actually you could end up, as is common with profit splits, you can end up with more than one, even if you've agreed profit split, you can end up with double taxation because both people measure the, the profit to be split differently. (R2-10) The details and formulas for a formulary method would require agreement between countries that had differing pre-existing approaches, and achieving that agreement was not viewed as realistic: what are the chances of actually getting global consensus on the factors used in the formula apportionment and the weight to be given to each? (R1-9) In 2006/07, interviewees recognised that the ambiguity of the ALP gave rise to some problems, for both taxpayers and tax authorities, particularly uncertainty and the risk of double taxation. Nevertheless, it was put forward as a workable approach, based on a common understanding of how it should be applied in practice, underpinned by an international consensus. The ALP has long been a central component of the rules of the transfer pricing game. Being seen to apply the ALP forms a doxic practice within the field and field members. This is not to say that field actors blindly accept the efficacy of the ALP, but rather this broad principle is tacitly agreed to be a pragmatic solution to the intractable problem presented by the need to find a mechanism by which sovereignty can be asserted over MNE profits for the purposes of taxing them. In this sense field participants collude to maintain the illusio of the field: the ALP is "the root of the competition which pits them against each other and which makes the game itself" (Bourdieu, 1996, p. 228;cited in Gracia & Oats, 2012, p. 307).

Disruption in the field
There was considerable change in the transfer pricing field after the first round of interviews including the increase in integration within MNE groups, the financial crisis, the impact of digitalisation and media scrutiny of transfer pricing, which was seen as one of the triggers for the BEPS project. The extent of the change in the transfer pricing environment was acknowledged by all interviewees during the 2016/17 interviews, such that the transfer pricing world was now "a completely different place" (R2-16).
There has been considerable change in the structure and organisation of MNE groups, and in technology since the ALP, with its emphasis on comparable transactions, had been set out in the 1995 OECD Guidelines. During this period, transfer pricing professionals who had been working with the "bedding down" ALP found this area of taxation was becoming the subject of media attention. The financial crisis was seen as a trigger for increasing concerns about the international tax affairs of large companies, including established transfer pricing practices now receiving attention by being reported in the media, for example: It's been going on for a long time. But people who aren't familiar with it, it's 'Wow this is like some weird structure they've created to hoodwink us and use all these loopholes'. So, people aren't even speaking the same language. (R2-9) There was a view that media coverage did not convey the challenges of transfer pricing management, and the complexities of the ALP approach, while acknowledging that is not easy to do: "You can't explain it to everybody, it's too complicated" (R2-6). There was recognition of the declining public confidence following media reports of transfer pricing manipulation and tax avoidance by high profile MNEs for example: "to the average guy on the street, it sort of seems unfair" (R2-11). This disruption was seen as something to be treated with care "people have to be super careful of what they say" (R2-15) and weathered until it disperses "people are going to get tired of it, they already have" (R2-9). The media attention was also seen to have impacted tax authorities' conduct, with concerns about reputational risk extending to tax authorities This led to more challenges for the workability of the system as it meant it took longer to resolve issues, for example: it's also meant there's much, much higher awareness with tax authorities and governments making sure that they can actually pass their reputational risk test that they have internally … they don't want to end up making any agreement with the taxpayer that's going to be disclosed in the newspaper and they haven't done their homework. So that has actually slowed down things where that need to get though more hurdles just to manage their reputational risk themselves. (R2-3) Since the publication of the 1995 OECD Guidelines, the business environment has undergone significant change, not least with the rapid growth of the digital economy, which has transformed modern business enterprises, for example: "the internet changed everything. The boots on the ground is not required to get sales in the company" (R2-2). This change has opened up new avenues for value creation, thereby posing new challenges for transfer pricing. The challenge is for all companies beyond those with digital products or a significant online presence, and also for tax authorities, to deal with these changes: the big challenge is digitalization in all industries, computers and algorithms are making a lot of decisionsit's not really taking into account this is happening, with the speed that it's happening. (R2-10) There was a sense that over this decade, current tax rules had not kept pace with the changing business environment, particularly technology: "it's like the world of tax is so far behind the way the world is working" (R2-6). The period following the first round of interviews was one of considerable change; against this backdrop, how have the senior transfer pricing professionals' views of the ALP changed?

2016/17 The ALP is seen as challenging
The second round of interviews began in 2016, following the BEPS final report in 2015. Interviewees acknowledged that the BEPS project encompassed a substantial amount of work, undertaken in a short time frame, which was seen as a result of a heightened level of urgency as, previously, "the history of the OECD has always been it takes forever to get anything out" (R2-14). Some BEPS points relating to transfer pricing were viewed as still offering considerable room for different interpretations such that, in practice: "any position you want I'm sure you could find support in BEPS language for doing it" (R2-13) This is because, in echoes of the negotiations over the 1995 OECD Guidelines, the BEPS negotiations involved international organizations trying to herd people into a room to get consensus on something, but then they have to go away and whether they do anything based on that consensus agreement or not, is up to them … … nobody's quite sure what it really means, or how it's going to be interpreted, because in order to get people to agree you kind of had to agree on something that people can interpret in different ways. (R2-1) A decade after the first round of interviews, interviewees raised that it was increasingly challenging to apply the ALP, giving rise to a less workable, more uncertain system. Previously the ALP was perceived as flexible by interviewees: this view of it as an advantage was not stressed in 2016/17. The flexibility that doesn't "dictate much of anything" (R1-4) was now viewed as challenging. The view in 2006/07 that the ALP was bedding down had not led to much clarity over the following 10 years, for example: I have people come into the tax team and they've started doing transfer pricing and they go "gasp, but I don't understand, there's no rule". There isn't really a rule. (R2-6) Taking into account all the different circumstances, functions and risk of each entity was seen as increasingly elaborate, requiring ever-increasing levels of documentation and analysis work: it's huge, auditors have to wade into some document room full of boxes and just a little thumb drive can have zillions of pages worth of stuff. (R2-2) By 2016/17, just over 20 years after international consensus was being built up for the OECD Guidelines, support for the ALP was more equivocal. The view expressed positively in 2006/07 that the ALP permitted and accommodated many different possible approaches was seen much less as a strength and more of a challenge. Problems that had previously been identified in 2006/07 were reiterated more strongly, particularly the lack of certainty and comparables: The comparables were hard enough to get in the old world. They're not going to be any easier to get in the new world. (R2-6) And: It's just so subjective, inherently subjective. And the economics are getting more extreme. If you look at reports that are done by economists, they're all over the place. A lot of formulas and adjustments. (R2-11) With regard to intellectual property: you never know which brand is going to become the super-valuable hot brand, that's going to suddenly produce huge profits, so transfer pricing can't predict that with any reliability. So you can end up with a huge amount of profit not taxed anywhere, even though the arm's length methods have been followed in all the relevant countries where that global group is conducting its affairs. (R2-12) Compared with 2006/07, when there had been an emphasis on double taxation, there was more recognition and discussion of non-taxation. This was acknowledged as an area of public concern and a force behind the BPEPS project: "the concept of stateless income got them to the table" (R2-5).
Interviewees reported further increased use of profit splits in practice, and that this was frequently the most appropriate and pragmatic transfer pricing approach. Therefore, in the years between the interviews, there is a reported wider acceptance and increased use of profit split methods in practice, operating under the aegis of the ALP: "arm's length principle plus is effectively what's happening" (R2-2). Over time, with repeated use, aspects of a profit split approach, such as a certain percentage markup under certain circumstances, become accepted as appropriate and common practice. Determining transfer prices is thus not always undertaken on a case-by-case basis, but rather in line with an emerging corpus of acceptable profit split practices, an important part of transfer pricing professionals' body of knowledge.
In 2016/17, interviewees expressed concerns with the problems of the ALP and a much less strident rebuttal of formulary apportionment compared with the previous round of interviews. There was now a recognition that this was on the agenda and potentially worthy of further consideration given the increasing problems stemming from the ALP. Some interviewees reiterated that formulary apportionment is not a principled approach; however, rather than rejecting formulary apportionment out of hand, there was a recognition that it was now on the agenda and there was a willingness to consider it: globalization has moved on to the point where groups are running an integrated global business, it just gets really hard to figure out in advance how to tax whatever the profits will end up being. So, it'll have to move to some kind of backward-looking methodology or some variation of formulary approach. (R2-1) Nevertheless, while there was an acceptance it was on the agenda, there was much suspicion of it, and misgivings, particularly with relation to the factors to use and the difficulty in obtaining international consensus on those factors. A shift to an increased openness to the possibility of formulary apportionment in 2016/17 was also associated with an increased discussion of its practical challenges, which were discussed in much more detail than 2006/2007, including some that were reiterated: People won't agree. Formulary apportionment would be great except for the one flaw that no one can ever agree on what the formulas were. (R2-8) One interviewee when asked about opposition to formulary apportionment commented: Why do I think there's resistance? Because we're used to the Arm's Length Principle, I think, is the only reason. It's like gravity. (R2-11) This indicates that the ALP is an engrained given for those working in transfer pricing and speaks to the resistance of fields to change as dominant actors struggle to maintain the illusio. The important role of consensus surrounding this longstanding given was drawn on as an argument to dismiss formulary apportionment: "I don't think it's stood the test of time that the arm's length principle has" (R2-14). Now the existence of the consensus around the ALP was presented as a reason for continuing with it: the long-standing consensus for the established principle was seen as making it difficult to move away from it.
The findings from our fieldwork supports the proposition that transfer pricing professionals are adapting to field level changes by altering their view of ALP as the only sensible mechanism for allocating taxable profits between jurisdictions. There has been a shift to more formulary approaches in practice but seen as coming under the umbrella of a very flexible ALP. The changes we observe in the transfer pricing field emanate from both within and outwith the field itself. Institutional shifts include the politicisation of transfer pricing as an immoral practice linked to profit shifting, leading to the intervention of the OECD through the BEPS project. Transfer pricing practices are no longer behind closed doors but are becoming issues of public discourse and civil society scrutiny. The capital held by the dominant transfer pricing specialists in the form of abstract, blackboxed knowledge of extant practices potentially loses its potency as the efficacy of the ALP is called into question. Prompted in part by civil society activists and media interventions, formulary apportionment morphs from an obscure, impractical, inefficient and dysfunctional mechanism into a viable alternative to ALP, mediated by the interim rise to prominence of the profit split method countenanced by the OECD. The elite transfer pricing specialists find themselves faced with having to adapt to the new field conditions, to realign their habitus to the new rules of the game. To maintain a dominant position within the transfer pricing field, these professionals leverage their sens practice, their feel for the game. Whether such dominance can be maintained in a future radical shift of approach remains to be seen.

Conclusion
This paper addresses two interrelated research questions: firstly, how has the common understanding and knowledge of the ALP developed? and secondly, how have senior professionals' views of the ALP changed? We answer these questions drawing on Bourdieusian concepts, which allow us to focus on the practices and relational interactions in the field. In so doing we respond to calls for fieldwork looking at transfer pricing in practice (Cools, 2014;Cools et al., 2008;Rossing, 2013) and the experience of tax professionals, (Radcliffe et al., 2018). This study of elite transfer pricing professionals differs from previous studies by offering a longitudinal view, drawing on interview data that shows changing views on issues raised first in 2006/07 and revisited in 2016/17, with an overlapping group of interviewees, contributing new insights into the experience of elite tax professionals and into transfer pricing practices.
The first research question allows for a new conceptualisation of the transfer pricing field, shedding light on the power struggles taking place within the field. The answer to this question then allows us to interrogate the role of the elite professionals whose power is partly rooted in the prevailing ALP, while at the same time allowing some movement in its boundaries bringing new struggles to the fore. Much of the transfer pricing accounting literature focuses on the transfer pricing rules and the challenging steps to follow in the pursuit of tax compliance, and the body of common understanding is generally unremarked. We therefore contribute a focus on the importance of unwritten "rules of the game" and accepted practices: the body of common understanding. This study took place over a time of considerable field disruption, (Anesa et al., 2019), including increasing digitalisation and media attention, when commentators highlighted deficiencies in the international tax system, and there has been increasing scrutiny and calls for change (Picciotto, 2020). Against this backdrop our empirical data shows that concerns raised in the academic literature, for example regarding digitalisation (Devereux & Vella, 2018;Eden et al., 2019;Mazur, 2016) and non-taxation of income (Kleinbard, 2013;Wells & Lowell, 2011) are echoed by the senior transfer pricing professionals.
Alongside the increasing media attention during the period, proponents of formulary apportionment, drawing on their expertise and positions in other fields such as the scholarly or advocacy fields, (Dallyn, 2017) sought to disrupt the transfer pricing field doxa with a view to precipitating radical change, and shelving the ambiguous ALP. Despite these pressures, those at the most senior levels of transfer pricing, while sensitive to increasing public concern, continue with their work, navigating complex, inconsistent rules, to determine profit allocation between different nations, weathering the media scrutiny. The ALP prevails, and our study contributes insights into this continued dominance by discussing the views of those working at the most senior levels of transfer pricing. We find that these senior professionals are sensitive to field disruptors, willing to recognise and discuss non-taxation in addition to double taxation, and also, now, to consider some formulary approaches. We find that in the opening position taken by these professionals, the ALP is held to be the preeminent mechanism for allocating taxable profits between the countries in which MNEs operate and formulary apportionment is effectively dismissed as an alternative. By the end of the period of our study, following significant disruption impacting on the transfer pricing field, the ALP is no longer held in such high regard and the possibility of formulary apportionment no longer an anathema. As a halfway house between ALP and formulary apportionment, profit split becomes increasingly prevalent, potentially facilitating greater tolerance towards formulary apportionment approaches. These perspectives and insights are important and relevant to those looking to change tax policy and governance, because the increasing calls for transformation of the field need to be cognizant of the extent to which existing practices are firmly embedded and thereby resilient to change.
Our study looks at these perspectives and practices through a Bourdieusian lens. As Spence et al. (2015, p. 782) observe in the context of cross-national professional service firms, "a Bourdieusian perspective allows us to understand how the contours of a field actually work; it helps understand the 'rules of the game'". Ongoing debates about the workability of the ALP, and the potential alternative approach of formulary apportionment provide us with the opportunity to unpack the transfer pricing field and consider how dominant actors within the field are able to shift their position as the field illusio shifts, with the result that their dominance is maintained. Importantly, the ALP is not defined by bright line tests and requires a significant level of discretion. This vagueness has allowed it to evolve and be reinterpreted over time, against a changing backdrop of intersubjective political-economic ambiguities and interaction, more recently through the intense scrutiny afforded by the BEPS project (Radcliffe et al., 2018). In contrast to Lupu and Empson (2015), who find that experienced professionals lose their ability to question the game, becoming captured by the illusio, we find that the elite transfer pricing specialists display adaptability in being responsive to changes external to the field. The increasing practical difficulties in applying the ALP in practice and the acknowledgement of an undocumented body of common understanding, concentrates power in the hands of those who dominate the transfer pricing field, whose position is preserved, and the value of their capital in the form of know-how of otherwise hidden, undocumented practices maintained.
Meanwhile "the system keeps stumbling on" (R2-9). Consensus and the need for agreement is a recurring theme in the interview data, highlighting the significance of illusio as an organising logic of the field. In 2006/07, the ALP was presented as a principled, flexible approach, with a strengthening consensus building around it. In 2016/ 17, we find that, despite increasing problems with the ALP in practice, and the principled approach not working as well as had been envisaged at the start of our study, the consensus itself was now presented as a reason to retain the ALP. The existence of that consensus was put forward as a key argument to retain the ALP, even though the business environment is different, and the ALP is working differently than envisaged when the consensus was being established; illusio preserved. Continued.
Tax authority Tax advisory firm Cindicates category of organisation where the interviewee was based at the time of the interview. Xindicates category of organisation where the interviewee had previously worked. Panel 1 -All but three interviews were at the workplace of the interviewee and all but one of the interviews was face to face. All but two of the interviews were recorded (for those two detailed notes were taken during and after the interview) and were later transcribed. Panel 2 -All but two of the interviews were at the workplace of the interviewee and were face to face. All interviews were recorded and later transcribed.