ato logo
Search Suggestion:

Keynote address to The Tax Institute's 2021 Transfer Pricing Conference

Last updated 9 September 2021

Hector Thompson
Deputy Commissioner – International

The Tax Institute – 2021 National Transfer Pricing Conference
9 September 2021

Good morning – I would like to start by respectfully acknowledging the traditional owners and custodians of the land on which we meet today, and pay my respects to their elders, past and present. I would also like to extend that respect to other Aboriginal and Torres Strait Islander people here today.

It is a pleasure to have been asked to provide some opening remarks to the Tax Institute of Australia’s Annual Transfer Pricing Conference.

First, I have a couple of confessions to make. I’m a bit of a latecomer to transfer pricing, having only recently re-joined the ATO after a 24-year absence. If it is any consolation, I am an economist and more importantly (at least in my opinion) am fascinated by how and why businesses make decisions. So hopefully that means that I might have something useful to offer you today. The other thing to confess before we get started is that before training as an economist, I was a historian. So, I feel particularly lucky to have the opportunity to become better acquainted with transfer pricing at a time when the global tax community is experimenting with some changes to the transfer pricing landscape. Hopefully that will explain why today I intend to cover a slightly longer time-period than I expect is the norm at conferences such as this.

Of course, I had also hoped that today would provide me with the opportunity to get to Melbourne for the first time in 2021 to meet with the many ATO staff who have been doing such a remarkable job during the COVID-19 pandemic, as well as an opportunity to meet many of you in person. Regrettably it is not to be. But I will continue to look for the chance to get to Melbourne and when I do, hopefully catch up with many of you in person. Finally, before I kick-off I would like to say a special thanks to the ATO’s Economist Practice and in particular Carol Kim for assisting me in putting together today’s remarks including the excellent research into the early history of transfer pricing.

While COVID-19 has understandably dominated the news over the last 18 months, it has also been an eventful year in the world of international tax – most notably with 134 jurisdictions coming together and agreeing to work together on a solution to the tax challenges arising from the digitalisation of the economy. That is quite a mouthful, so for ease of reference going forward I’ll just refer to this work as BEPS 2.0. While there remains considerable uncertainty and much technical work to be done, a key lesson of BEPS 1.0 is that where the international will for change exists, change will occur. So, it is never too early to take stock of how the international tax system is evolving, including reflecting on the drivers that have shaped the system to date.

Today I am planning to start by offering some reflections on the history of the international tax system, including the choices made around source and residency and the establishment of the arm’s length principle as the bedrock of the transfer pricing system as we know it today. I will then offer some thoughts on how the international business environment has changed over the last twenty years, the impact of BEPS 1.0 and how the ATO has responded. I’ll conclude by looking at BEPS 2.0 and what some possible implications of that work might be for the ATO.

As outlined in an excellent 2019 paper from Sara McGaughey and Pascalis Raimondos, the origins of our current system for taxing multinational enterprises can be traced back around one hundred years to the end of World War 1.

Just like the economist in me wants to suggest that the answer to any question you might ask me is ‘it depends’, the historian in me always likes to caveat things by saying that of course things were very different one hundred years ago. On this occasion, I do so with a touch of irony given I am delivering this address virtually due to a global pandemic that has closed Australia’s state borders!

The Spanish flu aside, coming out of World War I governments around the world were concerned about financing post-war recovery and reconstruction efforts - broadening the tax base to raise revenues became a universal priority and so clear rules on where things should be taxed and to what extent also became a priority. Perhaps for the first time, double taxation (which arises when the same income is subject to tax in more than one tax jurisdiction) became a real risk for both individuals and businesses that had income sourced from jurisdictions outside of their place of residence.

The other key driver was of course the desire to avoid international disputes – today we talk about trade wars, but one hundred years ago there was a very strong (if ultimately unsuccessful) desire to avoid actual wars. The League of Nations was formed in 1920 to facilitate international co-operation and in the same year the issue of double taxation was referred to the League of Nations by the International Chamber of Commerce as a key policy agenda item. The International Chamber of Commerce specifically called for '…prompt agreement between the governments of the allied countries in order to prevent individuals or companies from being compelled to pay a tax on the same income in more than one country…'.

To facilitate the development of a global solution for double taxation, the League of Nations fundamentally needed to consider – firstly, where income should be taxed (in the source or residence jurisdiction), secondly, if subject to tax in multiple jurisdictions, how income ought to be allocated, and thirdly, how to resolve disputes and administer any resolutions going forward.

It is interesting to note that then (as now) the US largely set the tone and framing of the double taxation, with Thomas Adams, then the key economic advisor to the US Treasury department neatly describing the potential existence of '…a corporation whose owners live in jurisdiction A, whose factory is in jurisdiction B, whose main offices are in jurisdiction C, and whose principal sales department is in jurisdiction D…' and concluding that all jurisdictions could reasonably demand and ultimately succeed in collecting some tax. The US also had a bit of a head start given their experience implementing significant tax reform in 1918, which included the introduction of an unprecedented ‘foreign tax credit’ to individuals or businesses that had paid taxes on the same income in a foreign source jurisdiction.

In pursuit of a multilateral solution, in 1921 the League of Nations engaged four economists - Professor Edwin R.A Seligman of the US, Sir Josiah Stamp of Great Britain, Prof G.W.J Bruins of the Netherlands, and Professor Luigi Einaudi of Italy. They produced a report in 1923, and on the question of where income ought to be taxed, found that it was dependent on the classification of the income – allocating progressive personal income taxes to the residence jurisdiction, and all other impersonal taxes between the residence and source jurisdiction based on economic allegiance principles. On the question of how to allocate income where more than one jurisdiction had entitlement, they came up with the observation that ideally an 'individual’s whole faculty should be taxed, but that it should be taxed only once, and that the liability should be divided among the tax districts according to… the doctrine of economic allegiance.’ This doctrine of economic allegiance was tied to the understanding that different activities had varying contributions to value creation, and thus entities ought to be taxed based on those respective contributions.

There are a couple of useful points to make about this conclusion. The first is that it sounds a lot like the key headline from BEPS 1.0 to me – aligning taxation with the location of economic activity and value creation. The second is that it was a report written by economists and so lacked much in the way of practical guidance about how such an ideal solution could be achieved. Perhaps as a result, their findings were placed to one side by the League of Nations, and instead a 1925 report produced by a committee of seven technical experts that represented the key European governments was endorsed by the League as it provided a solution that was more practical.

By 1927, this expert committee eventually expanded to include experts from six additional jurisdictions, and they met over the next few years to draft the model tax treaties. The model tax treaties were finalised in 1928, and was ultimately submitted to government experts from 27 jurisdictions, endorsing taxation based on residence, and largely limiting taxation in the source jurisdiction to scenarios where a permanent establishment existed - thus the need for a ‘physical presence’ to trigger a taxing right was created. While this was not necessarily the optimal technical approach or based on any compelling economic logic it did do a pretty good job of balancing the interests of capital-exporting countries such as the UK, capital-importing countries in continental Europe, and the US’ unique interests, which despite being a net capital exporter, also supported source-based taxation. The US position was interesting - driven by the both the US’ dominant trading position at the time (and long-term foreign trade interests) which enabled them to concede tax revenues to the source jurisdiction. It also aligned with Thomas Adams’ principle-based preference for source based-taxation, as evidenced by his work to introduce a foreign tax credit in the US a decade earlier.

In 1933, Mitchell Caroll, who had assisted Adams in the 1927 and 1928 model tax treaty meetings from the US Department of Commerce, was engaged by the permanent Fiscal Committee formed in 1929, to review the allocation of income – the main outstanding issue which had not been adequately addressed in the 1928 model tax treaties. Caroll found that there were two main approaches being adopted in the jurisdictions he investigated, firstly, a separate accounting approach, which involved each entity reporting on its national profit that would ultimately constitute the tax base and secondly, a fragmented, formulaic approach which calculated profits on a consolidated basis, and then apportioned these amounts based on the activities of the various entities.

Critically, Caroll endorsed the separate accounting system approach, which he represented as being clearly preferred by most tax jurisdictions. This effectively affirmed the Arm’s Length Principle which has formed the bedrock of the transfer pricing system ever since. His words at the time stand up remarkably well today - “As long as the inter-company transactions are carried on under the same circumstances and conditions and on the same terms as they would be between two entirely separate and independent persons, dealing with each other in an open market, and in a manner which is graphically described as at ‘arm's length’, the tax authorities in general respect the separate legal existence of the subsidiary company and tax it on the basis of its own declaration as supported by its properly kept separate accounts.”

Caroll’s findings were ultimately reflected in the drafting of the multilateral tax convention in 1933, the revised version in 1935, then the Mexico model tax convention in 1943 (which favoured source-based taxation and evolved into the UN’s Model Tax Convention), as well as the London Model Tax Convention of 1946, which favoured the residence-based approach (and evolved into the OECD’s Multilateral Tax Convention).

Since then, over 3600 bilateral treaties have been entered into based off the model tax conventions drafted, which would seem to indicate a level of success despite the numerous compromises required to develop them. Although there were significant interpretations of the arm’s length principle in the 1960s and 1990s, which resulted in the drafting of the OECD guidelines in 1979 and various revisions thereafter, overall the framework of the international tax system has not changed much in almost one hundred years.

For most of this period the separate accounting approach has probably seemed unremarkable to those who have looked at it given the structure of the global economy – particularly as for much of that period there were fewer opportunities for profit shifting. Most cross-border transactions involved trade in goods, global capital mobility was limited, there was less functional integration, intangible property was less prominent and there was generally less risk allocation to affiliates.

Of course, you can find the odd example of taxing foreign residents that don’t have a physical presence if you go looking. For example, in 1930 the Australian government inserted a provision into the Income Tax Assessment Act to tax foreign resident insurance companies that were not carrying on a business in Australia. The provision came about following dissatisfaction amongst insurers operating in Australia that non-resident insurers, such as Lloyd’s of London, had been obtaining insurance business through resident insurance agents or brokers without incurring any expenses or paying tax in Australia.

Lloyds was not a corporate body so special action was considered necessary. A provision was enacted that deemed a premium payable to a non-resident, with no physical presence in Australia, to be assessable income from Australian sources of the non-resident. The tax was based on the actual premiums paid, for which an arbitrary 10 per cent profit margin was specified. Alternatively, the non-resident could be taxed on their actual profits if this could be established to the Commissioner’s satisfaction. The government understood the difficulties of collecting the tax and so included in the legislation that the insured was not eligible for deduction for the premium until the tax was paid.

Today, some 400 income tax returns are lodged annually under the successor provisions (incorporated into Division 15 of Part III of the Income Tax Assessment Act). Persons deemed agents of the non-resident insurers lodge the returns, as well as pay the income tax liability that exceeded $350 million for the 2020 income year.

But such examples are the exception rather than the rule. For much of the last one hundred years, most business models were decentralised – functions, assets and risks were onshore where customers were located. This started to change during the 1990s, when to achieve economies of scale, specialisation and cost efficiencies there was a noticeable shift toward more centralised business models – which for instance, centralised funding, intangible ownership, R&D, manufacturing and risk management functions in one entity. This increased the need for cross border activities and increased functional integration between separate legal entities. In turn this created more opportunities to shift profits to low tax jurisdictions.

While the international tax system may not have changed much over the last one hundred years, the business environment most certainly has. Of most significance, capital has become much more mobile. According to the IMF gross capital flows comprised less than 5 per cent of global GDP between 1980-1999 and then peaked at around 20 per cent in 2007. This has created the risk that intra-firm funding decisions and structuring is influenced by tax outcomes, rather than solely on commercial outcomes. Examples could include Australian companies entering long-term (inbound) loan arrangements with related parties, despite only having short term working capital needs (and no other commercial reasons to enter a long-term arrangement). Alternatively, Australian companies could enter short-term (outbound) funding arrangements with related parties which in substance should be characterised as a long-term arrangement (and therefore should attract a higher interest rate).

Today intangible assets are extraordinarily valuable – according to the Forbes 2020 most valuable brands list – the value of the 10 most powerful global brands is almost $1 trillion USD. Moreover, ownership of intangible assets seems to be concentrated in investment hubs. The OECD’s 2020 Corporate Tax Statistics publication highlights that investment hubs account for 25 per cent of global profits of multinationals, but only 4 per cent of employees and 11 per cent of tangible assets.

While such observations about capital flows and intangible assets are probably of interest to a select few, digital disruption is of much broader interest to the community. The last decade has seen dramatic changes in information technology which have facilitated the disruption of many long-standing business models. Here I will draw on an insightful observation from The Simpsons. One of my favourite exchanges is when Mr Burns is checking on his investment portfolio with his advisers and asking how his portfolio is tracking. They assure him he is making excellent decisions when in fact his investment portfolio is clearly somewhat dated. So, I suppose there are two observations actually. But the one I am interested in today is that business and industry structures change, sometimes quickly.

In 2010, the top 5 US companies by market capitalisation came from a variety of different industries – Exon Mobil, Microsoft, Walmart, Procter and Gamble, and Apple with an aggregate market capitalisation value of approximately US$1 trillion. Today the top 5 exclusively comprise of technology companies – Apple, Microsoft, Amazon, Google’s parent Alphabet Inc. and Facebook. Facebook, the smallest of the five, has a market capitalisation more than US$1 trillion.

For much of the last one hundred years large multinational companies predominantly produced physical goods. Today, digital or highly digitalised companies can disrupt long-established industries including in the services sector (which is far less familiar with foreign competition than the goods sector). These new disrupters display a capability to substantively scale without mass (that is, without a physical presence in the customer market) and dominate their targeted market mainly through their unique technologies and extensive user base, importantly at an accelerated pace. It is worth spending a few minutes reflecting on the most prominent examples – each of which have had books written about them which I will draw on.

The Google Story chronicles how Google started out as a free web-based search engine that stumbled on how to monetise the data collected from their users, specifically, through providing targeted advertising services. They were eventually able to substantively scale their business and achieve high profitability, because of their free and unrestricted access to user data and highly automated, low labour-intensive organisational structure. Effectively Google pioneered a new means of production based on (free) digitised data and knowledge, which if contrasted with traditional labour or capital inputs, generated economic activity that was harder to trace, at least in a physical sense. That is, Google can extract data from users around the world to provide advertising services without the need for significant physical business operations in those jurisdictions.

In the case of Amazon as told in The Everything Store, a company that started out as an online book retailer transformed into a global business by leveraging its e-commerce experiences and in-house cloud infrastructure to consult and provide services to other businesses. Amazon Web Services (Amazon’s cloud business) was launched in 2006, well in advance of other competing cloud service providers such as Google/Microsoft that launched several years later. Amazon Web Services is now one of Amazon’s most profitable business segments constituting more than 50 per cent of Amazon’s total global operating income in 2020 and Amazon has approximately one third of that market segment. Amazon has been able to effectively maintain a dominant market position, at least in the context of cloud and e-commerce, because of their first-mover advantage, which gave them strong brand recognition and customer loyalty.

The Upstarts tells the story of Uber, which has similarly achieved a dominant position in ride-sharing, accounting for more than two thirds of the market, through prioritising user base growth and investments in R&D over early-stage profitability. Founded more than 10 years ago, it now has over 101 million active users globally, generates revenues of US$11.1 billion, has a market capitalisation of US$77 billion, but isn’t profitable globally. Of course, if a firm is not making a profit, it won’t be paying corporate income taxes. Perhaps this seems an unremarkable statement in this forum. But it is worth reflecting that it is a new reality for many in the community to grapple with - for most of the last one hundred years once a firm achieved sufficient global scale to become a household name it would have been safe to assume it was profitable.

What happens when an international tax system that has been reasonably stable for one hundred years runs up against rapid changes in the global business landscape (including digital disruption) that provides increased opportunities for shifting profits around the world? Well, something has got to give. That is why governments around the world endorsed BEPS 1.0 and what continues to drive the momentum behind BEPS 2.0.

Australia was an early supporter of BEPS 1.0 – we are heavily reliant on corporate income tax and so are particularly at risk from global tax strategies that seek to shift profit to low- or no-tax jurisdictions. We also recognise that change cannot be achieved by countries acting alone and so multilateral action is essential. Initial estimates from BEPS 1.0 put global tax lost at between $100 billion and $240 billion per annum.

It’s important to note that even as BEPS 1.0 was emerging, Australia had already begun working to address such concerns by introducing new transfer pricing laws – Sub-division 815 A and 815 B of the Income Tax Assessment Act. These updates, referred to as the modernisation of Australia’s transfer pricing regime, looked to ensure arrangements between related parties were considered holistically and explicitly referenced the OECD transfer pricing guidelines. This represented the evolution of transfer pricing rules in Australia from one where the task was said to be to merely price what the ‘transactions’ that the related parties had entered into, to one where all the economically significant features of the related party arrangement needed to be evaluated to ensure they are what would be expected to exist between independent parties in comparable circumstances, acting at arm’s length – a principle that is also reflected in BEPS 1.0 action items 8 to 10.

Specific international tax risks for Australia include large amounts of debt and related interest deductions, hybrid arrangements, the rise of the digital economy and associated questions around residence, source and permanent establishments, the value of intangible property and its use and creation, and a general lack of visibility over the global operations of multinational enterprises. In short - transfer mispricing. All these risks were covered by BEPS 1.0 and Australia has moved quickly to implement its 15 actions. I don’t propose to go through them all today but will touch on the final three, all of which seem likely to have specific relevance to BEPS 2.0.

The first is Country by Country (CbC) reporting – Action 13. This goes to the ATO’s ability to detect and deal with the misalignment of value and tax outcomes. These rules were legislated late 2015 and require Significant Global Entities to lodge a CbC report, master file and local file - the full suite of recommendations of the OECD Action 13 Report. 

The local file was designed after consultation with advisors, software developers and corporates with the structure itself developed by a group of corporate taxpayers. The final design includes a short form for use by lower risk entities as well as the more comprehensive local file. ATO systems have accepted electronic statements since 2017 and the ATO was on-boarded by the OECD to the Common Transmission System, which has allowed Australia to meet automatic exchange obligations for CbC reports from 2018.

In September 2020, the OECD published its third annual peer review report. These peer reviews focused on each jurisdiction’s domestic legal and administrative framework, its exchange of information network, and its measures to ensure the confidentiality and appropriate use of CbC reports. The peer review concluded that Australia met all the requirements under action 13.

The second is around the Mutual Agreement Procedure (MAP) – Action 14. This action sought to improve access to MAP and remove some of the barriers which may deny taxpayers to ability to resolve double taxation through treaty processes. It also sought to improve the efficiency of the MAP process to improve overall case times in recognition that there was an increased risk of international tax disputes arising from the BEPS 1.0 process. While Australia’s MAP processes were already largely in line with those recommended by BEPS 1.0, the ATO undertook a comprehensive review and has amended our MAP guidance.

The third is the introduction of arbitration through the Multilateral Instrument (MLI) – Action 15. The motivation for the MLI was that updating the current international tax treaty network (developed over 70 years) was a large job that would require a coordinated effort. This was considered beyond the ability of the traditional bilateral treaty negotiation process – the MLI offered an innovative and swift mechanism to synchronise the treaty network and provide uniform international tax rules.

That mechanism has been embraced at a global level, with the MLI so far covering 95 jurisdictions with ratification by 65 jurisdictions. It is the first multilateral tax treaty of its kind, allowing jurisdictions to swiftly implement BEPS 1.0 agreements into their existing tax treaties, transforming the way tax treaties are likely to be modified in the future to address the constantly evolving business landscape.

As at 1 January 2021 around 650 treaties concluded among the jurisdictions that have ratified the MLI have been updated, with an additional 1200 treaties expected to be modified once the MLI has been ratified by all signatories. Australia signed the MLI on 7 June 2017 and it is expected the MLI will eventually modify 33 of Australia’s tax treaties – 18 of Australia’s existing tax treaties are expected to have arbitration provisions. Under those treaties, taxpayers will be able to request arbitration if an issue of their MAP case remains unresolved by the competent authorities within the time period specified in the relevant tax treaty (generally two years). Australia is in the process of agreeing and publishing memorandums of understanding with those treaty partners, which will set out the mode of application of arbitration.

Australia’s implementation of BEPS 1.0 was also complemented and strengthened by the introduction the Multinational Anti-Avoidance Law (MAAL), the Diverted Profits Tax (DPT), ATO Practical Compliance Guidelines as well as the Tax Avoidance Taskforce.

The MAAL, which amended Part IVA of the Income Tax Assessment Act 1936 took effect on 1 January 2016. The MAAL was enacted in response to the OECD’s Action item 7 and targets multinational enterprises that avoid a taxable presence in Australia by undertaking significant work in Australia to generate sales in Australia but book the revenue from those sales offshore.

As a result of the introduction of the MAAL, many multinational enterprises have restructured their operations to book income from Australian sales onshore, ensuring increased transparency around these arrangements. The MAAL has brought over $8 billion of sales annually into the Australian tax net and has resulted in more than $100 million per annum of additional income tax being paid. To complement the MAAL, recent law reforms have extended the operation of the GST to apply to cross-border supplies of imported services and digital products, as well as low value imported goods, sold to Australian consumers.

To further strengthen the ATO’s ability to tackle multinational tax avoidance and reinforce the integrity of Australia’s corporate tax base, the Government announced the introduction of the DPT within Part IVA in May 2016 to encourage greater compliance by large multinational enterprises with their tax obligations in Australia, including Australia’s transfer pricing rules in Division 815 of the Income Tax Assessment Act 1997. It applies to DPT tax benefits obtained in income years commencing on or after 1 July 2017, which may include schemes that were entered into before this time.

As an anti-avoidance provision, the DPT has a broad application in combatting tax avoidance by significant global entities across all sectors of the economy. Consistent with the OECD focus on aligning profitability with economic substance, one of the objects of the DPT is to ensure that the Australian tax paid by significant global entities properly reflects the economic substance of the activities that those entities carry on in Australia. The DPT also has an object of encouraging taxpayers to provide information to allow for the timely resolution of disputes about Australian tax.

The DPT contains unique administrative provisions which enable the Commissioner to act on limited information and is designed to address information asymmetries and encourage taxpayer transparency and co-operation. In December last year, the ATO issued its first DPT assessment to a large multinational taxpayer and we are progressing other cases where we believe the DPT may potentially apply.

Examples of the arrangements where we are actively exploring DPT risk include:

  • the use of offshore procurement hubs to bifurcate the procurement functions with related offshore services hubs
  • the migration of intellectual property and arrangements where intangible assets are not appropriately recognised for Australian tax purposes
  • restructures which purportedly transfer functions offshore to a centralised regional entity without any change in the underlying substance of the arrangements in Australia; and
  • arrangements which provide funds to Australian associates through interposed entities in treaty-partner countries.

The ATO has also developed a range of practical compliance guidelines to provide transparency, additional certainty and offer potential compliance savings. They assist us to direct our compliance resources to higher risk areas of the law and to better tailor our taxpayer engagement. Taxpayers who self-assess and have a low risk rating can expect the Commissioner will generally not apply compliance resources to review their taxation outcomes. These practical compliance guidelines cover related-party financing, marketing hubs and inbound supply chains.

On inbound supply chains we have released guidance that applies to subsidiaries of multinationals who purchase goods and services from related parties and on-sell to Australian customers. This remains a key area of focus for the ATO. In addition to general guidance for inbound distributers, we have also developed specific guidance for the information and communication technology, life sciences and pharmaceutical, and motor vehicle industries. When assessing risk, we include broader consideration of a taxpayer’s global supply chain, tax profile of any related parties as well as the amount of tax at risk. Obviously, entities with inbound distribution arrangements that consistently suffer losses pose a very high transfer pricing risk.

Our practical compliance guidelines with respect to marketing hubs cover issues related to centralised operating models that involve procurement, marketing, sales and distribution functions. These guidelines have been operating since 1 January 2017 and have helped to shift taxpayers that have marketing arrangements in the high-risk zones to the low risk zones. Under the guidance, taxpayers are also able to reach agreement with us on their transfer pricing outcomes and move to what is known as the safe zone.

Our related-party financing practical compliance guidelines have regard to a combination of quantitative and qualitative indicators and provide the ATO with a framework to assess risk and tailor engagement with taxpayers according to the features of their related party financing arrangement. Taxpayers with related-party financing arrangements rated as low risk, can expect the ATO will generally not apply compliance resources to review taxation outcomes of the arrangements, other than to fact check the appropriate risk rating. Taxpayers with related party financing arrangement that falls outside the low risk category, can expect the ATO will monitor, test and/or verify the taxation outcomes. The higher the risk rating, the more likely taxpayer arrangements will be reviewed.

The Tax Avoidance Taskforce has been particularly important. This week the ATO has released a summary of the key outcomes of the taskforce in 2020-21 - with the help of the taskforce since 1 July 2016, the ATO has raised $22.9 billion in liabilities against public groups, multinationals, wealthy individuals and associated private groups (including trusts and promoters) and collected over $15.9 billion in cash. Without the additional funding and support the ATO would not have had the resources to deliver on BEPS 1.0 so comprehensively. One simple example is that the ATO’s Economist Practice (which supports all areas of the ATO in transfer pricing matters) is around double the size that it would be in the absence of taskforce funding.

While our comprehensive compliance program remains a critical part of the taskforce, over time the work of the taskforce has expanded to include a range of assurance activities, meaning multinationals, large corporations and wealthy individuals are now more compliant than ever with their tax obligations. This is by far the most important contribution that the taskforce can make to the sustainability of Australia’s corporate tax system – prevention is better than the cure. We now have numerous examples of companies committing to long-term behavioural change, including restructuring, changing their business practices, and settling long-standing disputes with the ATO.

During 2021–22, we will continue to focus on specialist large market advisors that promote and implement tax avoidance schemes, and engage in uncooperative, misleading and obstructive behaviour, including the misuse of legal professional privilege (LPP) during our reviews and audits.

Through our Medium and Emerging and International programs we will continue to use our growing data holdings to identify and target tax avoidance behaviours. In addition, we also use this data to improve our system design to ensure leveraged approaches are applied across this population which will reduce risks that such behaviours will perpetuate.

The Top 1000 Combined Assurance Review program commenced in late September 2020 and builds on the Top 1000 tax performance program. Work will commence with the reviews and associated engagements with respect of those taxpayers that obtained red flag and low assurance ratings. Taxpayers who use complex trust structures and distribution flows designed to exploit the use of trusts also remain firmly in our sights.

Finally, the taskforce will continue to advance our data and analytics capabilities and use of cutting-edge technology to improve the way we use and analyse data. These improvements to data accessibility and risk detection services will enhance our ability to target our engagement and assurance work including helping us to manage, interrogate and provide insights from our expanding data resources.

Building on the increasing focus on the assurance work of the taskforce, the ATO continues to develop our program of work around the Tax Gap. Our Tax Gap program recognises that just focusing on audit liabilities is not consistent with longer term success. Tax gaps are a measure of the shortfall in actual tax collections as compared with the tax payable. It has two levels, the gross tax gap, which measures the level of non-compliance at lodgement; and the net tax gap, which measures the residual level of non-compliance after compliance activity by the ATO.

In terms of large corporate market in Australia, when the taskforce commenced our estimate is that large corporates paid about 91 per cent of their tax due at lodgement (that is a 9 per cent gross tax gap), and 94 per cent after compliance activity. Our most recent estimate is that this has shifted to 92 per cent at lodgement and 96 per cent after compliance activity. Our ambition over the next two years is to see this shift to 96 per cent at lodgement and 98 per cent after compliance activity. The key point here is that this would mean more voluntary compliance and less ATO initiated compliance activity relative to today.

So that brings us to BEPS 2.0. Since the program of work to develop a consensus-based solution was published by the OECD in May 2019, there have been several rounds of public consultation and negotiations between jurisdictions to iterate on key features of its two-pillar approach. Specifically, Pillar 1 includes Amount A, which is designed to allocate residual profits to market jurisdictions and Amount B, which seeks to define a fixed return for baseline marketing and distribution activities in some market jurisdictions. Pillar 2 relates to a global minimum tax that acts as a type of fail-safe to offset any shifting of profits to low tax jurisdictions - perhaps inspired by the US’ Global Intangible Low-Taxed Income (GILTI) regime implemented in 2017.

I won’t cover the details of the pillar work other than to touch on some administrative considerations – I can see that you have a session on Pillar 1 this afternoon with Michael Jenkins (who has extensive experience in the ATO’s Economist Practice) and Nick Marshall. Tomorrow I’m looking forward to hearing from Richard Vann and Mary Hu on Pillar 2.

But given where my talk started, I will make the obvious point - we seem to have gone back to the future and are re-exploring some lines of inquiry from the 1920s that perhaps some had assumed were settled. Most notably we are once again talking about when a taxing right should exist and how income should be allocated where multiple jurisdictions have a potential entitlement to tax income.

In terms of when a taxing right should exist there has already been agreement that market jurisdictions (i.e. where the end-user or customer is based) ought to be allocated more taxing rights, specifically ‘where value is created by business activity through (possibly remote) participation in the jurisdiction…’. That is, the conclusion reached in 1928 that a ‘physical presence’ was required to trigger a taxing right has been partially overturned. Perhaps this is unsurprising given the changes that we have observed in the global business landscape in the last 20 years. Certainly, our lived experience of the COVID-19 pandemic has reinforced how many aspects of our lives have entered the digital realm as we experience prolonged periods of working (including attending conferences) and consuming goods and services without leaving our homes.

How to allocate income and effectively determine which jurisdiction has priority taxing rights continues to be developed – there have been several iterations of the proposed Pillar 1 approach in response to consultations and negotiations. At a high level where jurisdictions have landed is to adopt a global formulaic approach to income allocation. Again, if we look back, it is fair to say that the preference amongst jurisdictions that Mitchell Caroll reported in the 1930s for a separate accounting approach has diminished somewhat.

I say ‘partially overturned’ and ‘diminished somewhat’, because Pillar 1 is very much a change at the margin – Amount A will only apply to multinational enterprises that earn more than 20 billion euros in global revenues and have attained at least 10 per cent profitability, while those in the extractives and regulated financial services industries are not covered. There is less clarity on other aspects of Pillar 1 like Amount B. But it is already clear that the administration of Pillar 1 will not require wholesale changes to the existing international tax system.

Although wholesale changes may not be required, the timeframes for the new administrative tools that will be required are very ambitious – a Multilateral Instrument for Amount A is set to be developed and signed by 2022, with implementation of Amount A from 2023. Unlike BEPS 1.0, this new package will require virtually simultaneous uptake from all countries to succeed. These are unprecedented challenges and there is much to be done. The determination of Amount A and how it should be allocated will be challenging – history has taught us this.

From the perspective of tax administrations, multilateral certainty around the application of the Pillars will be critical to ensure that double taxation (including from unilateral measures such as digital services taxes) is avoided. There will also need to be a strong focus on making sure that allocation mechanisms are as simple and transparent as possible and that there are strong mechanisms to avoid or resolve disputes in a timely manner. Trust in the system will be critical - not just for corporate taxpayers and tax administrations, but also for the broader community.

The infrastructure that has been built as part of BEPS 1.0 provides a solid foundation. The three action items touched on provide us with some useful building blocks – but our mechanisms for exchanging information, ensuring transparency and providing certainty will all need to be developed further to enable them to effectively support the administration of BEPS 2.0.

The other key lesson to take from BEPS 1.0 is the importance of consulting with advisors, data and accounting experts as well as corporate taxpayers in the design and operation of new mechanisms – this is very much front of mind for the ATO.

Thank you for your time today. I’d also like to thank the Tax Institute and the Organising Committee for inviting me to speak. The program for the conference looks fantastic and I’m looking forward to hearing from so many leading experts across the legal and accounting professions as well as corporate Australia.



Adams, TS (1917) ‘The Taxation of Business’ Nat'l Tax Ass'n Proc, vol 11, no. 185.

Caroll, MB, (1968) ‘International Tax Law - Benefits for American Investors and Enterprises Abroad’, International Lawyer, Vol. 2, No. 4 pp 692-728.

Graetz, MJ, O'Hear, MM (1997) ‘The "Original Intent" of U. S. International Taxation’ Duke Law Journal, vol. 46, No. 5, pp. 1020-1109.

Langbein, SI, Fuss, MR, (2018), ‘The OECD/G20-BEPS-Project and the Value Creation Paradigm: Economic Reality Disemboguing into the interpretation of the ‘Arm’s Length Standard’, The International Lawyer, vol. 51, no. 2, pp. 259-409.

McGaughey, SL, Raimondos P (2019) ‘Shifting MNE taxation from national to global profits: A radical reform long overdue’, Journal of International Business Studies, vol. 50, pp. 668-83.

Zuboff, S, (2015) ‘Big Other: surveillance capitalism and the prospects of an information civilization’, Journal of Information Technology, vol. 30 pp. 75-89.


Herndon, JG (1932) Relief from international income taxation: The Development of International Reciprocity for the Prevention of Double Income Taxation, Callaghan & Co, Chicago.

Stone, B (2014), The everything store: Jeff Bezos and the Age of Amazon, Random House UK, England.

Stone, B (2018), The Upstarts, Random House UK, England.

Vise, DA Malseed, M (2005), The Google Story, 1st edn, Delacorte Press, United States of America.

Reports / Publications Annual Report 2021This link will download a file

IBISWorld (2021) Cloud storage services in Australia.

IBISWorld (2021) Online Shopping in Australia.

League of Nations (1933) ‘Methods of Allocating Taxable Income’ Taxation of Foreign and National Enterprises Vol IV.' Submitted to the Fiscal Committee by Mitchell B. Caroll, Geneva: League of Nations, Document C. 425(b). M.217(b).

League of Nations (1923) ‘Report on Double Taxation’. Submitted to the Financial Committee by Professors Bruins, Einaudi, Seligman and Sir Josiah Stamp, Geneva: League of Nations, Document E.F.S.73.F.19; April 5.

OECD (2020) Corporate Tax Statistics – Second EditionThis link will download a file

OECD (2020), Country-by-Country Reporting – Compilation of Peer Review Reports (Phase 3) : Inclusive Framework on BEPS: Action 13, OECD/G20 Base Erosion and Profit Shifting Project, OECD Publishing, ParisExternal Link

OECD (2019), Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy, OECD/G20 Inclusive Framework on BEPSThis link will download a file

OECD (2018), Measurement and Identification of Capital Inflow Surges This link will download a file

OECD/G20 Base Erosion and Profit Shifting Project Countering Harmful Tax Practices More Effectively: ACTION 5: 2015 Final Report, at para 91External Link

United Nations (2001) United Nations Model Double Taxation Convention between Developed and Developing Countries, UN, New York.

Uber (2021) Uber Annual Report This link will download a file