Future of tax administration

Jeremy Hirschhorn, Second Commissioner, Client Engagement
Paper delivered to the PricewaterhouseCoopers Global Tax Symposium
Paris, 14 November 2019

Introduction

Good afternoon. Thank you for the opportunity to be a part of this year’s global tax symposium, to listen to and be a part of robust discussions around the new tax landscape and what opportunities that brings us as policy makers, administrators, advisors, and taxpayers to interact in a more effective way.

Today I am here to provide an administrator’s view of what is happening in the world of taxation of multinational businesses. Much attention is rightly placed on policy change, and the BEPS initiatives, whether 1.0 or 2.0, but much less attention is given to developments in tax administration, which can have as big, or even bigger, an effect on the taxation landscape.

Over the last five years, there have been significant changes in Australia as to how we administer multinational business taxation, and I have been asked to share Australia’s experience with you today, both to provide insight as to where taxation administration might be heading more broadly internationally, but also to help those companies investing or planning to invest in Australia understand how we are striving to make Australia a more “tax certain” and attractive place to invest.

I will start with some context of the Australian environment to set the scene; then talk about how administrators see their world and their own incentives/performance measures; how these incentives or performance measures are changing and what that then implies for administrative practice.

I will then make some observations on developments in tax transparency (both of the ATO and of large companies), and the potential uses of data and analytics and automation across the board but particularly in the large market.

I will finish with some reflections on what this all might mean for multinationals and their advisers, leaving some time for questions.

The Australian context

Australia has a high reliance on corporate tax, a highly tax conscious population and one that has a low appetite for budget deficits.

Importantly, the number of large corporate groups in the Australian tax system is comparatively small, yet the impact they have on revenue is significant. Beyond their tax contribution, they also play a critical role in the tax system as community perceptions of their compliance underpin willing participation in other taxpayer segments.

To put some figures to this, Australia’s 1,600 or so large corporate groups contribute around 60% of all corporate income tax reported, about $50 billion or so per annum, give or take depending on commodity prices; and more than 10% of total ATO tax collections each year.

As a percentage of GDP, corporate tax collections in Australia have historically been significantly higher than other OECD countries, bouncing between 4% and 5%, versus 3% or so for the OECD average. A table which demonstrates this is attached as Appendix A to the paper. Given general comparability of tax rates over the period, this implies one or more of the following: a higher share of GDP going to corporate profits, a broader tax base (or fewer implicit concessions), or greater tax compliance (voluntary or involuntary).

At the same time, Australia is highly exposed to base erosion and profit shifting:

• A now relatively high corporate tax rate at 30% (and a low VAT at 10%) as other countries reduce their corporate tax rates and increase their VATs;
• An open economy in terms of both capital and trade, exposing the tax system to transfer mis-pricing and debt loading (with minimal “drag” from customs duties or tariffs);
• A nearby low tax rate jurisdiction which is a trading hub for the region.

A table which demonstrates Australia’s physical trade, high levels of related party transactions, and misalignments between the two is attached at Appendix B.

Overlaying this is a very fiscally aware but sceptical community. Delivery of a surplus (or the relative size of projected surpluses) is a consistent theme of Australian politics, and considered a sign of that party being a good economic manager. Australians are therefore acutely aware of the relationship between tax collections and government expenditure on social programs. As such, any sense that one group is not paying its way is directly linked in the community’s mind to a rationing of government expenditure.

To give what might seem a slightly frivolous example of the Australian community’s tax awareness, it is generally accepted that an opposition leader lost an “unlosable” election in 1993 by failing to adequately explain on a current affairs show the effect on the cost of a birthday cake of moving from a wholesale sales tax system to a VAT.

In parallel, a general sense of unease or lack of trust had been building in Australia (as elsewhere) that large companies were not participating fairly in the Australian economy. Five years ago, in the context of taxation, this was put fairly and squarely into the public domain by a report by the Tax Justice Network and the formation of a special Corporate Tax Avoidance Inquiry of the Australian Senate.

The Australian Government acted forcefully to address this community unease through all measures at their disposal, including:

• Increasing the funding of the Australian Taxation Office to form a Tax Avoidance Taskforce directed at public and multinational businesses and the largest private companies;
• Rapid and pragmatic adoption of a range of BEPS 1.0 measures, including anti-hybrids, Country by Country (CbC) reporting, best practice anti-transfer-mispricing rules.

Five years on, it is generally accepted that this government initiative has been very successful, and the taskforce has been extended and expanded. Later in the paper, I will touch on our estimates of “tax performance” of large corporates, which support this position. A year or so ago, the government ran an advertising campaign across all media “Earned here, taxed here” informing the community as to the success of the initiative.

However, it is worth noting that although community attitudes have shifted, still significantly fewer than half of Australians believe that large companies are paying the right amount of tax. For example, in the last couple of weeks there have been headlines about the minimal level of Australian corporate tax paid by Netflix as a prominent B2C business with minimal physical presence in Australia (and hence expected to have minimal corporate tax under the established international framework).

There also remains scepticism of the role of large law firms and the Big 4 professional service firms in relation to the provision of taxation services, with the following themes often recurring:

• As large providers of consulting services to government, are they earning fees from the government at the same time as they are reducing its ability to offer services through aggressive tax planning?
• Are they somehow capturing the tax legislative/policy agenda?
• Are the efforts of the ATO being subverted by a “revolving door” between the ATO and large firms (but at the same time recognising the increase in capability of the ATO driven by external recruitment)?
• Are taxation services tainting the Big 4’s independence as external auditors?

Understanding administrator incentives/metrics

Traditionally, the reporting of tax system health was relatively limited, with a focus on revenue collections and audit yield, and also efficiency measures. The attractions of these measures are that they are easy to robustly measure and require minimal judgement.

Unfortunately, they do not necessarily provide much information as to the health of the system – a high audit yield could be a signal of a very healthy system (high voluntary compliance, most non-compliance being caught), but also the opposite (low voluntary compliance, some non-compliance being caught).

On the other hand, total revenue collections are also unlikely to be a good proxy for revenue authority performance, as they reflect the volatility of the economy. For example, in an economy with a strong commodity exporting sector like Australia, volatility in total revenue collections in a given year will be more strongly linked to commodity prices than to revenue authority performance.

Where this can easily tend to lead is for a large market tax administrator to focus on:

• A target dollar level of audit adjustments based on a “pay-off ratio” compared with auditor costs (often linked to new/additional funding from government)
• A focus on audits of the largest companies (because that is where the largest nominal audit dollars are likely to be) and minimal focus on the next tier of companies
• A focus on “strike rates”, what percentage of audits give rise to adjustments (and failing to make an adjustment being a “fail”) (avoiding these “fails” can be a motivation behind lengthy audits)
• A risk based approach that maximises “strike rates” with as few false positives as possible
• Minimal incentive to actively prevent future non-compliance (not reflected in metrics)
• A focus primarily on what has gone wrong, rather than how to make the system healthier in future

Often community confidence is also measured, but there are significant challenges in communicating the actual health of the system with the community: the only “facts” are about how much that has been found to be wrong – there is limited information as to how much is “right”, other than total tax collections, and no information as to how much is wrong and has not been found. As such, higher audit liabilities can be persuasive both that the problem is under control, but also that it is out of control. This means that community confidence will be based primarily on perceptions rather than evidence.

An analogy I sometimes draw is with drug testing at the Olympics. If the World Anti-Doping Agency (WADA) can only speak to the number of athletes that it catches, increasing positive tests will, for a while, increase community (and peer athlete) confidence. However, after a while, and if levels of positives keep increasing, this may actually decrease confidence, and potentially be self-reinforcing in a vicious circle – other athletes may feel that they need to take drugs to compete, and spectators will stop watching. The number of positive tests is a dubious proxy for the true goal – minimising the number of “uncaught” drug takers.

Another way of drawing out the limitations of this type of metric is to pose the question: what would happen if we achieved ultimate success and there were no drug cheats/all taxpayers paid the right amount of tax? An audit liability based metric suite would give the appearance of failure, not great success!

How are administrator incentives changing?

The increased focus on large company taxation from the Australian Government and the community forced us to think deeply about our performance as administrators, but also how we communicated with the community.

The ATO had of course always recognised that a focus on audit liabilities was not consistent with longer term success, reflected in some internal rules of thumb “prevention over correction” and “you can’t audit your way to success”. However, we did not have the measurement tools to support this thinking, and were perhaps held back by our success under traditional metrics (and noting that, in practice, quantitative metrics will often be taken much more seriously by staff than qualitative metrics).

The step change was to move to concepts of “tax gap” (and its flip side, “tax performance”), with a target of sustainable reductions in the tax gap.

In short, tax gap is a measure of the shortfall in actual tax collections as compared with the tax theoretically payable. It has two levels, the “gross tax gap”, which measures the level of non-compliance at lodgment; and the “net tax gap”, which measures the residual level of non-compliance after compliance activity by the administrator. Importantly, it does not measure or make any comment on policy settings as reflected in law. (It is beyond the scope of this paper to go into the methodologies by which different “gaps” are measured, but for those interested they are set out in detail on the ATO website.) Tax performance is simply the reverse of this – a 5% tax gap is equivalent to 95% tax performance.

Once an administrator moves to a tax gap or tax performance “philosophy”, the next step is to measure and, in my view, critically, to publish the gap results. This demonstrates to internal, cross government and external stakeholders that the revenue authority is serious about this as a longer term performance measure.

In terms of large market corporate tax in Australia, when the taskforce commenced our estimate is that large corporates paid about 91% of their tax due at lodgment (9% gross gap), and 94% after compliance activity. Our most recent estimate is that this has shifted to 92% at lodgment and 96% after compliance activity. Our ambition is, by the end of the taskforce, to have moved this to 96% at lodgment and 98% after compliance activity.

This is already a very high performance level, but benchmarking this against the tax performance of individuals (not in business), their performance is 94% at lodgment (voluntary compliance). Unless and until the voluntary tax compliance of large companies is (and is perceived to be) significantly higher than the voluntary tax compliance of individuals, individuals will rightly challenge large companies to do better as role models in the taxation system.

However, “tax gap” is a whole of system measure, and its linkage to the work of underlying teams is too indirect to provide performance feedback. This then requires the development of internal metrics to provide strong signals to teams as to their best work.

In the ATO, we developed the following “balanced scorecard” performance suite for our large market teams (reflected at a site level, teams of about 50–100):

• Active compliance (audit results)
• Active prevention one to one
• Active prevention one to many
• Tax / economic activity assured.

I will discuss these in more detail later on.

The next phase of “tax gap” thinking that we are working on is to develop the concepts of “addressable gap” and “gap at risk” and robust methodologies for calculating these.

In more detail, addressable gap looks to how much of the existing gap could pragmatically be eliminated given existing policy settings etc. (much the same way as theoretical full employment is not 100% employment, but will usually embody, say, 4% unemployment). Referring to our large company aspirations above, our current estimate is that a 96% to 98% target is effectively full achievable compliance in the real world scenario of a country with significant cross border trade. On the flip side, “gap at risk” focuses on how much (and how quickly) the gap would increase if resources were redeployed away from that area.

It is important to emphasise that, even under a “gap” focus, audits will always be with us and of great importance. In the “full achievable compliance” scenario above, we estimate that audit liabilities will still be 2%, recognising that a credible audit program supports high levels of voluntary compliance both in terms of deterrence, but also in terms of perceived fairness of the system for those inclined to comply.

Another key metric that the ATO has recently adopted is the OECD concept of “tax assured”. It is an estimate of the proportion of tax that we have the highest level of confidence is being correctly reported, and requires positive evidence: in the large company market this is based on our one-on-one “justified trust” reviews, discussed later on. In other markets this is based on third party data which can be relied upon with high degrees of confidence. We currently do not split this out in our annual report between markets, but we are now approaching 50% of total tax paid being assured to this level. (Note that this should not be confused with the levels of assurance we provide to large companies under the “justified trust” initiative – refer later.)

One limitation of “tax assured” as applicable to the large market (under the OECD methodology) is that it is based on the amount of tax rather than the level of economic activity. This “rewards” reviews of large profit making entities with commensurate large tax payments rather than large entities with more mixed economic and tax performance (arguably the companies that should be more closely reviewed). As such, at a team level, we have developed a metric with its focus on economic activity assured.

With tax gap measuring the amount that we predict is incorrect, and tax assured measuring the amount with the highest level of confidence, we are currently working on metrics which measure interim degrees of confidence to provide the community with an even more nuanced understanding of large company tax compliance.

What this means for administrator strategies

Of course, having new internal philosophies and over-arching metrics is all very well, the key question is how it informs strategies and actions.

As a critical pre-condition supporting our general change in strategies, the Australian Government was an early and pragmatic adopter of initiatives to address BEPS concerns, outside the scope of this paper, but at a high level including a mix of both G20 and OECD led “BEPS 1.0” reforms, and domestic Australian initiatives such as:

• Moving to “best practice” OECD transfer pricing provisions
• Implementation of the foreign hybrid mismatch rules, bolstered by a targeted integrity rule to stop simple conversions to “next generation” or “synthetic” hybrids
• Introducing the “Multinational Anti-Avoidance Law” (“MAAL”) which addresses permanent establishment avoidance schemes
• Introducing the “Diverted profit tax” (“DPT”) which addresses combined transfer mis-pricing/general avoidance schemes
• Tightening of thin capitalisation rules
• Country by country reporting
• Enhanced information sharing with other jurisdictions including exchange of ruling information
• Adoption of the “Multilateral Instrument” (“MLI”) to modify Australia’s double tax treaties to implement BEPS treaty related measures such as incorporation of the Principal Purpose Test (“PPT”) to address treaty abuse, and mandatory binding arbitration for unresolved Mutual Agreement Procedure (“MAP”) cases.
• First adopter of the VAT cross border measures relating to low value imported goods primarily through the major market place platforms as well as joining other countries in introducing VAT on imported services and digital products.

Resourcing

Five years ago, like most revenue agencies, we had a primarily risk based approach to compliance activities.

Due to the concentrated nature of the corporate tax population, the “top 100” (responsible for about half of all corporate tax) had annual “pre-compliance reviews” every year, so full coverage of population, but the reviews were still focused on risk events. The remainder of the large corporate population were covered through risk analytics or engines, with the outliers being referred to further compliance activity in the form of risk reviews or audits.

It became clear that this level of population coverage was not meeting community or Government expectations, and so the ATO was funded to dramatically increase its work in relation to the largest companies under the Tax Avoidance Taskforce, with additional funding of almost $700 million over four years.

Following this funding, under the “justified trust” initiative, coverage has been extended to:

• Annual “justified trust” (comprehensive) reviews of the top 100 companies;
• “Justified trust” reviews of the next 1,000 largest companies on a four year cycle;
• Supported by enhanced risk analytic engines applying to the entire population (including those subject to “justified trust” reviews.

The (then) Treasurer proudly announced this as “embedding” ATO officers into taxpayers.

The success of the taskforce has been reflected in its recently announced extension and expansion for a further three years, with additional funding of about $1 billion. At a high level, this supports the extension of our work in relation to the largest public companies and multinationals, and as a significant expansion of our work with the largest private groups.

'Justified trust' reviews

Justified trust is a concept from the OECD. Justified trust builds and maintains community confidence that taxpayers are paying the right amount of tax. It also allows us to focus our resources in the right areas.

To achieve justified trust, we seek objective evidence that would lead a reasonable person to conclude a particular taxpayer paid the right amount of tax. This is a higher level of assurance than confirming certain risks do not arise. We tailor our assurance approach based on the unique business profile of a taxpayer. It is somewhat similar to a “tax due diligence” process.

When engaging with a taxpayer, we review the following four key areas:

• Understanding a taxpayer's tax governance framework

We confirm the existence, application and testing of a tax risk management and governance framework. We recognise entities use different governance practices based on a range of factors.

• Identifying tax risks flagged to the market

We review risks or concerns we communicated to the market (for example, through Taxpayer alerts, Practical compliance guidelines or Public rulings). We then determine whether these may be present.

• Understanding significant and new transactions

We seek to understand current business activities, particularly significant or new transactions, and the tax outcomes.

• Understanding why the accounting and tax results vary

We analyse the various streams of economic activity and how they are treated for taxation and excise purposes.

This requires a holistic understanding of the taxpayer's business operations and financial performance. We compare this to its tax performance.

For example, we analyse:

• the Effective Tax Borne (ETB) and global value chain to understand why accounting and income tax results vary
• sales, acquisitions and other data and compare this to net goods and services tax (GST) paid.

Our view is that by looking at a taxpayer’s tax structure and performance holistically, rather than individual transactions or issues in isolation, we are in a better position to understand the overall tax risk profile.

Effective tax borne

“Effective tax borne” is a concept that is worth expanding on, in part because it has strong parallels to some of the work that will be required to flesh out the BEPS 2.0 Pillar 1 and Pillar 2 proposals.

Understanding the effective tax borne (ETB) of the Australian channel is critical to the ATO understanding the tax risk profile. This is particularly important for transfer (mis) pricing matters. We often see a price set based on an optimistic application of one methodology. To truly judge the risk of a transfer pricing position, it is important to triangulate a range of methodologies, as well as to understand the big picture of where channel profit is being landed around the world (especially untaxed profits).

The methodology identifies an economic group’s worldwide profit from Australian-linked business activities and the Australian and offshore tax paid on that profit. As a Head of Tax or adviser, ETB is another tool which will provide you with an early insight into how you or your client may be profiled and risk assessed by the ATO, in an effort to establish justified trust. As above, it is also a useful sense check on any transfer prices and whether they are giving plausible, common sense outcomes. Further, it may flush out any upstream hybridity or non-taxation relevant to the anti-hybrid rules or the MAAL or DPT.

We are currently expanding our methodology to develop a VAT or GST equivalent and looking to bring it within both the Top 100 and Top 1,000 programs. This is to address the following concerns that we observe in practice, even at the largest companies:

• reliance on bottom up systems
• relatively frequent system failure (not technical interpretation failure)
• there is no natural systems check to determine if the results are plausible.

Ultimately, both an income tax ETB and GST or VAT/ETB should be part of the client’s governance systems, not an ATO review system.

For a copy of the ETB Guide and workbook we use to complete our ETB calculations, email Top100@ato.gov.au

For a copy of the GST analytical tool (GAT) Guide and method statement we use to complete our GAT calculations, email GATSupport@ato.gov.au

Focus on the important, not the interesting

An (internal or external) adviser’s role (looking deeply at an individual case) is often at odds with the role of an administrator who is looking at the system in practice and thinking about the ‘others’. What will others do?

As tax professionals (including those within an administrator), it is all too easy to place a disproportionate focus on the most interesting technical or theoretical tax issues, rather than those that are important at a system level.

Instead, we will prioritise or triage actions or risks based on:

• potential risk to revenue
• likely proliferation in the market, including the likelihood to flow through to smaller players
• wilful subversion or confrontation of policy intent.

Particularly in such a concentrated system as Australia’s large corporate tax environment, the ATO will be acutely focused on the risk of a “race to the bottom”. Where a race to the bottom gets out of control, this is also prone to trigger a legislative response. One way of thinking about this from an adviser’s perspective is “what would happen to tax collections if everybody did this?”

In the context of the Australian economy, most significant tax issues relate to cross border dealings and their pricing, whether it be physical trade or the borrowing of money. These must be “transfer priced” on an arm’s length basis, not “transfer mis-priced”. For the Australian economy, the key issues are around in-bound loans and applicable interest rates, exports of commodities and in-bound distribution businesses.

We recognised that companies had to determine transfer prices and, in the absence of practical ATO guidance, were often (inadvertently or deliberately) entering into positions that we viewed as high risk. As such, the ATO has been much more deliberate in its focus on these areas, and on exposing its risk analysis and frameworks to the taxpaying community. These are often in the form of Practical compliance guidelines (PCGs), discussed further later in the paper.

This transparency allows companies to make informed decisions as to the risk profile that they wish to adopt, rather than potentially inadvertently taking on tax risk.

New metrics for team-level performance

As above, to support the new approach, we developed the following “balanced scorecard” performance suite (reflected at a site level, teams of about 50–100):

• Active compliance (audit results)
• Active prevention one to one
• Active prevention one to many
• Tax/economic activity assured.

Active compliance is traditional audit yield. As noted above, even under a “gap” focus, audits will always be with us and remain of great importance. In the “full achievable compliance” scenario for the large company market, we estimate that audit liabilities will still be 2%, recognising that a credible audit program supports high levels of voluntary compliance both in terms of deterrence, but also in terms of perceived fairness of the system for those inclined to comply.

Active prevention “one to one” measures those activities where it can be demonstrated that a particular taxpayer changed their position (either prior to lodgment or by self amendment) through the influence of the case team. An example may be where a company came in asking for an APA at a particular price, and then changed that price following discussions with the team. Another key example is where a settlement “locks in” future compliance at a different level (e.g. a lower related party interest rate). The estimated increased tax payable attributable to the lower deductions will be attributed to team efforts (and will ultimately be reflected on a cash basis over time in the Annual Report as “wider revenue effects”).

Active prevention “one to many” measures those activities where it can be reasonably estimated that activities changed the behaviour of multiple taxpayers. A typical example is where a scheme has been detected with a likelihood of spreading through the system, a warning is sent out and the scheme does not spread. Of course, the calculation of these effects is less reliable (and would rarely be externally reportable), but is arguably the most effective contributor to system health and therefore must be encouraged.

Tax/economic activity assured is a measure which applies where the team has positively assured that the tax paid was correct (very different from not finding anything wrong). This provides recognition to a team which has looked into a taxpayer and proves that its tax is correct (to be contrasted with a “miss” under traditional metrics). Building up “tax assured” case by case can also support a “whole of system” metric of “tax assured”, as well as that positive finding of health being used in methodologies to estimate “tax gap”.

Influencing corporate governance of tax

A good, lived, tax governance framework will ensure fewer inadvertent positions are taken (both system type errors, but also risk positions inconsistent with the company’s risk appetite). As such, moving to a system health perspective means that the administrator has a strong incentive to encourage better tax governance frameworks.

In addition, being able to demonstrate how your good tax governance is embedded in positions taken, disclosures in returns and tax calculations provides us with evidence we can rely upon which can reduce the intensity of our enquiries, to the benefit of both the company and the administrator.

When working with our largest corporate taxpayers, we first look for evidence that a tax control framework exists, like a board-endorsed tax policy, documented procedures for preparing returns, including income tax returns and business activity statements (BAS), or a testing program to validate the operating effectiveness of the tax control framework.

Once we’ve established a tax control framework exists, we then look for objective evidence that the framework is designed effectively.

To reach our highest rating for tax governance (stage three) taxpayers must be able to demonstrate that their tax control framework has not only been designed effectively, but is also operating as intended. This stage can be evidenced by a periodic tax controls testing program as well as reports describing the outcomes of that testing. We seek an independent review and testing of tax controls, for example by internal or external auditors that provide an independent level of assurance to the audit committee and the Board.

At the other end of the spectrum, we may assign a “red flag” where taxpayers cannot provide evidence to demonstrate a tax control framework exists or if we have significant concerns with their tax risk management and governance. These concerns may include their approach to tax compliance, for example, where there are significant errors their tax control framework is not detecting.

To help large corporates understand our practices, we have a guide on our website which sets out what we believe better tax corporate governance practices look like so they can develop or improve their own tax governance and internal control framework, test the strength of the design of the framework against our best practice benchmarks, and understand how to demonstrate the operational effectiveness of key internal controls to their stakeholders, including the ATO.

As an aside, I note that in all too many cases, we see taxpayers with stated conservative risk frameworks enter into aggressive tax structures. We can only assume that insufficient consideration has been given to the risk framework and/or contradictions with the risk framework have not been brought to the board’s attention.

Increasing the sophistication of systemically important firms

From an administrator’s perspective, there are fundamental differences between boutique and systemically important firms.

Boutique firms operate within a system and can, in a sense, view the system as a constant environment in which they operate. However, once a firm becomes systemically important, any actions of the firm will change the system, and other participants in the system (including the ATO and government) will respond.

As such, a systemically important firm cannot simply view itself as a club of boutiques. If you describe a partnership as like a franchisor/franchisee model, where each partner runs their own franchise, this means that a systemically important firm must have much tighter controls over each of the franchise businesses, to ensure that they are operating consistently with the core brand.

To make this more tangible, where a partner in a boutique firm comes up with a “cute” piece of tax planning, which is then implemented by one taxpayer, the ATO will see this as a “one-off”. If, by contrast, a partner in a systemically important firm comes up with the same idea, the potential (or reality) of that idea being rolled out across Australia’s highly concentrated corporate tax base will mean that the ATO must respond more forcefully.

This means that an administrator has a strong interest in the corporate governance and approaches of the “systemically important” firms and ensuring that they are at a level of maturity which matches their significance to the system. I will touch on some of the questions we have posed (and clients should pose) to their advisers:

Too focussed on technical interpretation risk?

Traditionally, when judging the riskiness of a piece of advice, the focus has been on whether the advice meets a bar such as “reasonably arguable”, sometimes implicitly “non-negligent”. This analysis was often through an academic technical lens.

I would posit that this approach is insufficient for several reasons, including:

• it is very hard to self-judge your own ideas (“I consider that I am right: but I will now objectively quantify how likely am I to be wrong”), and so any meta-analysis of advice runs the risk of being too optimistic – this implies you need to leave a “buffer” – maybe a “strong should” is the true “more likely than not”?
• failing to appreciate the context of the advice in the real world of the tax system means that you are not fully judging the risk of your position being disputed
• this type of analysis is very difficult for trickier questions: the position may ultimately not be judged based on today’s perspectives, but on the perspectives of five or more years into the future.

Different standards for “tax infrastructure” and minor issues?

In any risk analysis, it is important to judge not just the risk of something going wrong, but the consequence if it goes wrong.

In my experience, advisers and their clients have not necessarily sufficiently distinguished between:

• tax matters which go to the very heart of their structure
• matters which potentially have material exposures if they fail (I would refer to these two categories as “tax infrastructure”), and
• day-to-day matters which might be very interesting, but are not really that important (and failure can be tolerated by the client).

If an analogy is drawn with physical infrastructure, the desired confidence might be 99% or more: should tax infrastructure be at “more likely than not” or a “weak should”?

Does the advice reference the likely ATO response or relevant ATO views?

As mentioned earlier, an internal or external adviser’s role (looking deeply at an individual case) is often at odds with the role of an ATO officer who is looking at the system in practice and thinking about the ‘others’. What will others do? When you have a good idea, be aware it will not be considered by the ATO as a “one-off”.

Instead, we will prioritise or triage actions or risks based on:

• potential risk to revenue
• likely proliferation in the market, including the likelihood to flow through to smaller players
• wilful subversion or confrontation of policy intent.

As an administrator we deal with risks through prevention – where possible – through guidance and alerts – and correction through assessments and litigation where necessary. This helps to prevent the "race to the bottom" I mentioned above.

Consistent with the taxpayer’s stated tax risk appetite or governance framework?

Of course the taxpayer’s perspective on what ‘good advice’ looks like should largely be determined by their tax risk appetite or governance framework.

As stated earlier, in all too many cases, we see taxpayers with stated conservative risk frameworks enter into aggressive tax structures. We can only assume that insufficient consideration has been given to the risk framework and/or contradictions with the risk framework have not been brought to the board’s attention.

Taking into account the Australian (local) environment?

When investing in Australia, many multinationals assume that the tax environment in Australia is broadly equivalent to that in their home country (and vice versa – some advisers assume that other countries are broadly equivalent to Australia).

It is therefore important to understand how the Australian environment differs to ensure that they are making their tax risk appetite decisions on an informed basis. Some key areas to flag might include:

• higher levels of transparency to the public, combined with community attitudes to and awareness of the levels of tax paid by multinationals
• high political interest given the relatively high reliance on corporate tax collections, coupled with a tax base that is highly concentrated across relatively few large corporations – this also leads to rapid action to protect the tax base when threats emerge [refer for example the Multinational Anti Avoidance Law (MAAL), the Diverted Profits Tax (DPT), pragmatic and rapid implementation of other BEPS measures, e.g. hybrids etc.]
• higher levels of transparency to the administrator (ATO), including reporting like the Reportable Tax Position Schedule
• higher levels of coverage of the large market (with full annual review coverage of the top 100, full annual risk model coverage of the top 1,000 and full review coverage on a rolling four-year cycle) means that detection risk is extremely high compared with other countries, and
• (I would like to think) a well-staffed, global best practice ATO Public Groups & International team focused on the important tax issues and with extensive transfer pricing or mis-pricing expertise.

Promisingly, we are seeing the systemically important firms in the Australian environment engaging positively with these themes, providing more sophisticated, wiser advice to their clients, and putting additional controls and safeguards in place to stop the delivery of overly “clever” advice with systemic impacts.

Transparency to the market of our risk settings and areas of concern

In terms of specific issues, we are doing our bit by also trying to be more transparent to the market.

As a developed, open economy, with significant inbound and outbound investment and trade, Australia relies heavily on foreign investment to support our economy, infrastructure and standard of living. It is therefore important that the practical application of the tax policy settings are as clear and predictable as possible, so that foreign investors can confidently invest knowing what the tax settings are (and making deliberate choices around accepting tax risk).

Many of the most significant tax issues relate to cross border dealings and their pricing, whether it be physical trade or the borrowing of money. These must be “transfer priced” on an arm’s length basis not “transfer mis-priced”. For the Australian economy, the key issues are around in-bound loans and applicable interest rates, exports of commodities and in-bound distribution businesses.

The ATO has been much more deliberate in exposing its risk analysis and frameworks to the taxpaying community. These are often in the form of PCGs, which set out rules of thumb for determining whether the ATO is likely to accept the price at face value, or will more deeply probe whether the price makes sense in the particular circumstances.

We are using PCGs more and more to allow companies to make informed decisions as to the risk profile that they wish to adopt, rather than potentially inadvertently taking on tax risk.

Through this approach we are seeing taxpayers looking to engage with us earlier to resolve issues and 'lock in' future arrangements in line with the guidance – this provides certainty to both the ATO and the taxpayer moving forward.

To just expand on a few of our recent focus areas, where we have been quite overt about what is our view of compliance risk:

• Related party loans continue to be used by Australian taxpayers, and we have made great inroads in addressing taxpayers who have sought to achieve artificial transfer pricing benefits through these arrangements. For example, we have brought about $80 billion in previously high risk related party loans into low risk or ‘green zone’ arrangements, and expect this to increase as we resolve existing issues. Our unofficial estimates are that, already, about $25 billion of interest deductions will not be claimed over the next decade as a result of this activity.
• Marketing hubs continue to be a key focus area. But again, by working actively with taxpayers to resolve these matters we are seeing a positive impact of our hubs strategy on reducing the use of these arrangements to reduce tax in Australia. It’s a matter of public record that BHP will now be paying their full Australian tax on their Singapore hub, and that their ongoing arrangements are in line with our ‘green zone’ set out in our guidance materials. This was a significant development in our marketing hubs strategy; as one of Australia’s major taxpayers this sends a strong signal to taxpayers with similar arrangements in that industry, as well as to the emerging oil and gas industry.
• More recently we have also signalled to the market our focus on inbound supply chains. We observed a race to the bottom in terms of the profit landed in Australia and realised we had to intervene with expected returns. Not only was this inappropriately reducing Australian tax, but also clogging up our APA processes with ambit claims, adversely affecting taxpayers with genuine prices seeking certainty.

We have also been much more deliberate in communicating to the market (in as near to real time as possible) when we come across arrangements which concern us, and have a risk of proliferating (recognising that there may previously have been a misapprehension that ATO silence was equivalent to permission). This is usually through the issue of a “Taxpayer Alert”, which sets out the broad features of the arrangement of concern.

Tax disclosures

Australia has always had an environment of high transparency to the ATO.

Income tax returns and related schedules like the International Dealing Schedule mean that the ATO has had significant insight into the tax affairs of large companies, often more than other countries.

Like other areas of transparency, this has increased over the last five years, in particular the Reportable Tax Position Schedule has been re-energised and expanded, asking very pointed questions as to whether particular Taxpayer Alert arrangements have been entered into, and the “risk zone” of transfer pricing positions under PCGs.

These questions are very important from a risk perspective, but are often difficult to discern from normal data sources such as standard disclosures in a tax return.

Settlements

Negotiated settlements will continue to play a role in the ATO’s approach to resolving disputes in the large market. That said, the ATO can only settle on a principled basis, and will only agree to a settlement after careful consideration of the risk to revenue, precedential value of the dispute, and likelihood of success in litigation.

We will continue to use litigation where we believe a point of law requires clarification or we need to call out unacceptable behaviour. For this reason, settlements are also less attractive to the ATO where we see systemic issues that need to be addressed.

We are being firm with taxpayers looking to settle with us and making it clear we will only do so where we can also lock in future compliance to achieve certainty of appropriate tax outcomes into the future, which is particularly relevant for transfer pricing matters.

Settlements will generally need to not only address the back years, but also forward years. Forward years will (almost always) need to be in the ‘safe zones’ set out in our guidance products to secure revenue and create certainty for the future. We have no interest in settling back years only to immediately kick off the next audit.

This ensures all arrangements meet what we have publicly stated as acceptable behaviour, and achieves consistency in our approach between like taxpayers and like issues (consistency being a critical element of tax administration). Of course, access to the independent view of the courts is always available if a mutually acceptable settlement cannot be reached.

Particularly where a matter is already in the public domain, we are increasingly requiring taxpayers to include a public disclosure of the broad strokes of a settlement to assist in the community deriving confidence from the settlement (rather than a further detraction from confidence because of a perceived or alleged “sweetheart deal”).

Supporting the Foreign Investment Review Board (FIRB)

In Australia, particular types of acquisitions by foreign entities are subject to the FIRB process. In short, an application must be made to the FIRB, which then considers whether the transaction is in Australia’s national interest and makes a recommendation to the Treasurer, who then formally approves or disallows the transaction. The Treasurer may also allow the transaction, but subject to conditions.

As noted earlier, as a developed, open economy, with significant inbound and outbound investment and trade, Australia relies heavily on foreign investment to support Australian’s economy, infrastructure and standard of living. It is therefore important that the practical application of the tax policy settings are as clear and predictable as possible, so that foreign investors can confidently invest knowing what the tax settings are (and making deliberate choices around accepting tax risk).

One element of the FIRB process is that FIRB asks for the ATO’s advice as to the taxation implications of the transaction as proposed (as one input into the national interest test).

In assessing our view of the effect of a transaction, we will flag where we consider a transaction has been aggressively tax structured (for example if it incorporates a “Taxpayer Alert” type scheme or whether transfer pricing is in a high risk zone under a “Practical Compliance Guide”). We may also suggest that some tax related conditions be imposed on the transaction.

In certain cases, the FIRB may communicate our tentative recommendations to the applicant and facilitate a meeting directly between the applicant and the ATO.

Some stakeholders have suggested that this is somehow pre-empting or usurping the applicant’s full access to law, and the ATO should effectively rate every transaction as low risk (and audit afterwards). Apart from the fact that the ATO is required to make this assessment to assist the Treasurer in making an informed decision, there are significant benefits of the current approach, including for applicants, including:

• Firstly, basing our advice on public guidance like TAs and PCGs means that the process is predictable and consistent across applicants (and applicants are not “out-bid” for assets by more tax aggressive applicants).
• Secondly, as a country that is seeking foreign investment, it would be somewhat perverse for the Treasurer to effectively invite an investor in, only for the ATO to be waiting on the other side of the welcome mat with compliance action.

One difficulty that we are facing in practice is that some applicants have advised of their structuring in only vague terms at the time of the FIRB process, and then have implemented aggressive tax structures when ultimately successful. This is viewed particularly dimly (in relation to all participants in the application process, not just the foreign applicant), as it fundamentally distorts the ability of the Treasurer to properly apply a national interest test.

This unfortunately means that, where the structuring of a transaction is vague, we now have to almost assume the worst or suggest strict tax related conditions.

By contrast, where an applicant fully discloses their structure, and it does not include Taxpayer Alert type features and has low risk pricing (or an undertaking is made to act on this basis even if the exact structure has not yet been fully determined), the transaction will generally be rated “low risk” by the ATO, and FIRB will concentrate on other aspects of the transaction.

Of course, once an acquisition has occurred, as part of our on-going compliance activities we will examine whether the transaction was implemented as advised, and whether all tax conditions are being met.

Trends in transparency

Transparency to the public around system health

Every year, the ATO, like other government agencies, produces an Annual Report. It sets out how the ATO has performed in the year, including its own financial statements, its management and accountability, its performance against a range of non-revenue based measures and its revenue performance.

As mentioned earlier, traditionally the reporting of tax system health was relatively limited, with a focus on revenue collections and audit yield. Unfortunately, this does not provide much information as to the health of the system – a high audit yield could be a signal of a very healthy system (high voluntary compliance, most non-compliance being caught), but also the opposite (low voluntary compliance, some non-compliance being caught).

Recognising that Australians deserved more information as to the health of their tax system, the ATO now annually publishes system level information such as:

• Total revenue effects: the full effects of ATO activity, not just money raised from audits, so including amounts from actions like “active prevention one on one” as those benefits emerge over time
• Tax gap/tax performance estimates across market segments and taxes: these estimate how much of the tax payable according to law is paid voluntarily, how much as a result of compliance action, and how much goes unpaid
• Tax assured: the amount of tax over which we have the highest confidence or assurance that it has been reported correctly
• Tax and Corporate Australia: a detailed analysis by sector and ownership of the tax performance of large companies.

As stated earlier, in terms of the large market, there is some good news: our most recent estimate (for the 2016–17 year) is that large companies paid about 92% of the tax due at lodgment, increasing to 96% after compliance activity. This is improving from prior years, which were already at globally high levels. (However, flipping it the other way, we estimate that large companies did not pay $2 billion of tax due, even after compliance activity of $2 billion.)

Our aspiration is to increase this level of performance to 96% correct at lodgment and 98% after compliance activity.

I note that calculation and publication of this type of data is very sensitive or even uncomfortable for a revenue authority, as you are putting effort into publishing information which shows that you are not perfect and have more to do. However, in my view it is necessary if you are to provide your citizens with a realistic picture of the health of their tax system.

I am sometimes asked in forums like this why there is such focus on large company tax performance. Apart from being role models for the entire tax system and whether it is fair for the “little guy”, it is more useful to compare the voluntary compliance of individuals (94%) and the largest companies (92% after recent improvements). I would posit that, until the voluntary (not total) compliance of large companies significantly exceeds that of individuals, individuals will quite rightly challenge large companies and the ATO to do better.

Transparency around settlements

Another area where we have worked to increase transparency is around settlements with large public and private businesses.

Of course, due to taxpayer privacy, we cannot publish details of particular settlements. However, we acknowledge that Australians have a legitimate interest in assuring themselves that the ATO is not doing “sweetheart deals” with the “big end of town”, so we have taken the following steps:

• Every year, we publish the aggregated settlements by market segment (number and value) in our Annual Report (noting that large market settlements are “lumpy”, over the last four years, the average “discount” on settlement for public and multinational businesses was about 35% versus 52% across all other markets);
• For every large settlement, we have a retired Federal Court judge (from a panel of judges) review the settlement to independently determine whether the outcome reached was a fair and reasonable outcome for the Australian people, and report the results in our Annual Report.
• On appropriate matters, we encourage the taxpayer to publish broad details of the settlement (both in relation to the past and the future) so that the public can have additional insight to (and hopefully confidence in) the settlement.

Transparency of the ATO view of a taxpayer’s own affairs

Under the “justified trust” program (enabled through increased funding), the ATO is performing what might loosely be described as “tax due diligence” on Australia’s largest companies, and has a better, more comprehensive understanding of these companies’ affairs than ever before.

As part of this, we are more transparent with companies as to the areas of assurance and the (hopefully limited) areas of concern, providing a detailed review report to the company at the end of the review, including our overall level of assurance (on a scale from low to high). We are also periodically publishing information on both the “Top 100” and “Top 1,000” programs as a whole, which allows companies to understand how they stand relative to their peers. We are told that many boards and directors find this information extremely useful.

Findings report - Top 1,000 income tax and GST assurance programs

As discussed above, we are also much more transparent to the market generally as to our risk settings around things like key transfer pricing challenges (through the issue of PCGs) and transactions of concern (through the issue of Taxpayer Alerts).

Transparency within a taxpayer

I have discussed the importance of corporate governance earlier in this paper. Corporate governance can usefully be seen as internal transparency:

• A strong framework sends a strong signal to the rest of the company as to what is acceptable;
• A strong compliance framework to make sure the governance is “lived” ensures that the board knows that the company is acting in accordance with its intent.

While on the theme of internal transparency, I would respectfully suggest that a board member (and possibly even significant shareholders) should understand, at the very least:

• The tax governance framework;
• The risk stance and structural tax settings of the company;
• The current (and historic) relationship with the ATO;
• Where profits are not fully taxed, and whether that relates to explicit policy concessions or “grey zone” tax planning (from both accounting profit and cash paid perspectives), or even that the profits are not considered high enough quality to be taxable (yet?).

And if a specific “tax infrastructure” question comes up to the board, I would suggest some potentially useful questions:

• Is the position consistent with the risk appetite of the organisation?
• Is the ATO likely to dispute this position? Have we sought certainty from the ATO in the form of a ruling?
• What would happen to revenue collections if everyone did this?
• What is our level of confidence as compared with that required on our physical infrastructure (with like levels of economic exposure)?
• (To the adviser) have you been given a full scope, or are there areas that have been scoped out that are relevant?
• Are the facts and assumptions underpinning the advice supportable and could be evidenced in court proceedings? What happens if they are wrong or disproved?

Transparency to the public of a company’s affairs

In Australia there have been a range of measures aimed at increasing tax transparency to the public:

• The corporate tax transparency measure, under which the ATO reports the gross accounting income, taxable income (really taxable profit) and tax paid of the largest companies;
• The voluntary tax transparency code, currently 161 companies have published at least one report;
• The requirement for some significant global entities to lodge general purpose financial statements (even where they may not trigger ASIC requirements).

I do not propose to walk through each of the measures in detail, but will reflect on the community response and some personal observations as to best practice.

The corporate tax transparency measure is produced annually and, in fact, the fifth release is scheduled for early to mid-December. The information provided in relation to each company is limited. At the time of release, the ATO also prepares significant information and analysis at a population level to assist the community position the raw company by company information. Many companies will also prepare some form of contextual information to flesh out the limited data.

The voluntary tax transparency code is, as the name suggests, voluntary. The best reports provide excellent insight, but it is fair to say that the quality of reports is quite variable, and I will make some further comments below.

For certain Significant Global Entities (broadly more than $1 billion gross income), there is a requirement to lodge general purpose financial statements (if they do not already do so). In practice, this can technically often be satisfied by producing whole of group GPFS of the ultimate offshore parent, which provides little insight into the Australian operations.

Taking a step back, I would put forward the proposition that “limping in” to these requirements is not the right strategy for a large company operating in Australia (or globally) today. Rather, I would suggest that a forward thinking large company would reflect on the underlying driver, namely the community appetite to have an understanding of the largest participants in the economy. This implies using official measures as a floor to be built on, not a ceiling to aspire to.

Best practice may look something like the following:

• Aggregated financial statements of all your operations in a jurisdiction (whether or not they have to be consolidated under strict accounting rules)
• A level of information consistent with General Purpose Financial Statements
• Clear explanations of both accounting effective tax rates and cash effective tax rates, with explanations as to the structural reasons where they differ from the headline rate
• Clear explanations of where channel profit is booked around the world, and the tax rates applicable (and tax paid)
• Disclosure of disputes with revenue authorities.

I strongly recommend that every CFO pull out the BHP tax performance report to see what is possible.

Conversely, I strongly recommend that tax transparency not be seen as some form of marketing response. This may work in the short term, but increases medium to longer term risks. Warning signs include:

• Primary responsibility for tax transparency is in the marketing or public relations area
• A focus on economic contribution and taxes withheld on behalf of others before it discusses corporate taxes
• Multiple sub-groups with a variety of accounting approaches (whether GPFS, SPFS or less) with no over-arching bridge to explain the overall operations, then “limping in” to tax transparency requirements.

General observations on transparency

Although in no way professing to be an expert on transparency or that the ATO is perfect, I would like to make some general observations on being a large organisation striving to be transparent (derived from both observing other companies but also as a leader in the ATO):

• Firstly, transparency is hard, and by providing additional information, it is of course possible that others will deliberately or accidentally misuse that information. It is easy to use this as an excuse for inaction;
• However, with current views as to access to information, a lack of transparency will also be used against you: in my view, the risk of saying something is less than the risk of saying nothing;
• If you are concerned about making your (tax) actions transparent, you do not have a transparency problem: you have an action problem;
• It is dangerous to have a “media” strategy, as this encourages “nuancing” (spin?). One should aim for a “doing the right thing transparently” strategy.
• Being transparent publicly provides strong signalling and discipline across an organisation: it shows your people that you mean what you say;
• Transparency benefits those doing the right thing: it makes it harder for others to hide in the shadows, and focuses the attention of scrutineers.

Use of data and analytics in large corporate tax administration

The ATO is collecting more data than we ever have before and it’s becoming more and more important as to how we manage and use this data to ensure the trust and confidence of the community.

We are looking at the advances being made around data, analytics, artificial intelligence and automation as well as the tools providing insights based on data to give us a better understanding of our clients and where we need to focus our attention.

All organisations are grappling with these opportunities and challenges, but I set out below some broad thinking as to how this might change tax administration.

Pre-data world, to data testing, to data driving

If one reflects on the evolution of tax systems, they arose in a “pre-data” world.

In other words, the relevant information was in the possession of the taxpayer, and it was difficult or impossible to verify that information against other data points (even on a one-on-one basis, let alone at scale). The mechanisms adopted to cope with this problem included:

• Limiting the tax base to small parts of the population (so there was less to check), for example through high thresholds;
• Requiring disclosure of detailed information through a “bespoke data set” (which we called an “income tax return”); and
• Having significant penalties for incorrect disclosures.

The next phase in the data evolution was when other information became practically accessible to the revenue authorities at scale. It was then used to test the “bespoke data set” and identify those tax filings which were most likely to be most incorrect. An example of this was where financial institution information was accessed after filing season and retrospectively back-tested against filed returns.

The third phase is when the data drives the system, or indeed at its ultimate where the system is defined around the available, verifiable data. In the Australian system, more and more of the typical taxpayer’s income is effectively at this third stage and driven by data, with employers and banks (amongst others) required to transmit details of salary and interest income, which is then used to prefill tax returns. In a further evolution, data is increasingly being provided in real time, in a form which can be directly linked to the client, rather than at the end of the year: this allows greater insights early, and allowing us to identify issues or concerns requiring early attention, and nudge accordingly.

System features of large market are limitations

In terms of large corporates, and corporate income tax in particular, I cannot foresee moving to the third phase in the short to medium term, for reasons including:

• Taxable profits are not intuitively derived from source data, rather being tertiary data (accounting profits being an interim step);
• Even if it were possible for there to be a derivation from source data, each company has its own bespoke system (or set of systems), and so it would not be possible to review efficiently across the population;
• If the key concerns are transfer (mis) pricing, there is no natural data “in system” to validate or disprove the transfer price.

These concerns mean that in relation to the large market, the focus of administrators will be on using data and analytics to have a more sophisticated “phase two” approach.

Even there, the nature of large companies and their issues will mean that the administrator will tend to “push” more complex risk analysis to the company. For example, in the Reportable Tax Position Schedule, the ATO requires companies to disclose:

• Their own assessment as to where they stand on PCG risk settings
• Whether they have entered into any schemes covered by Taxpayer Alerts.

Risk engines development

Developments in data and analytics will allow for a much more sophisticated set of risk engines. With simpler models, a challenge for administrators is the rate of “false positives” (and indeed “false negatives” when an audit team does not follow through on a “true positive”).

Using more modern tools, we are able to develop a more connected risk model architecture that enables both horizontal and vertical aggregation of risk models. It will be much easier to meld single issue risk models together to identify “clusters” of flags that are highly predictive and identify patterns and trends not previously readily identified.

In conjunction with high coverage “justified trust” reviews, it is also possible to “reverse test” risk models against populations (both of proven positives and proven negatives) to develop risk models that minimise both false positives and false negatives.

Dealing with unstructured data

Due to developments in data and analytics, it is also much easier to deal with unstructured data.

This is perhaps most important in the large private group market for example, where the level of public data is lower and less reliable. This capability has been critical, for example, in analysing the Panama and Paradise document leaks.

In the large corporate group market, it will allow better population based risk engines, for example trawling financial statements, but also at a specific taxpayer level, for example documents produced in the course of an audit to more quickly identify key documents.

'Scale' markets – individuals

In the “scale” markets (for examples individuals and small businesses), the ATO already has very sophisticated risk engines to identify anomalies. The challenge has been how to act on anomalies identified given limitations in human resources.

We are increasingly using real time (or near to real time) “nudges” and “data transparency” rather than traditional compliance interactions. For example, where we know from third party sources what a taxpayer’s salary or interest income is, we pre-fill the return where possible. If the return is filed before pre-fill information is available, we may subsequently propose a consequential amendment to the taxpayer, who can simply accept the amendment.

A more sophisticated example is in relation to “nearest neighbour” analytics around tax deductions. Where a taxpayer is preparing a return, immediately prior to processing the ATO systems will compare the level and type of deductions with an estimated peer group (the “nearest neighbours”), grouped by factors like profession and location. Where there is a significant variance, the taxpayer will be notified of this variance and have the opportunity to reconsider their claim prior to filing.

The next challenges to work on include:

• Whether the responses to simple calls and interactions with the ATO can be automated for example through increasing use of chatbots;
• Whether there are opportunities for automation of elements of more traditional compliance activities (particularly the aspects where people are currently “data sherpas”, simply moving information from one place to another).

All of this will free up our people so that more of their time can be spent on work requiring human judgement and application of empathy – neither of which is able to be automated or readily coded in an artificial intelligence context.

'Scale' markets – small businesses

There are some features of small businesses which potentially provide data and analytics opportunities not present in the large market:

• In the Australian market, small businesses are increasingly using one of a small range of sophisticated accounting packages.
• At the same time, their tax affairs are generally simpler and more linked to accounting or cash information.
• Our gap analysis shows Australian small business that are “in system” will generally return their income correctly.

These features mean that there are significant opportunities for “straight through processing” of tax compliance (potentially doing away with the need for separate tax return filing, rather system based pre-fills from accounting records supported by “in system” reviews by tax professionals), as well as the ATO exposing more of its risk engines and filters to software providers to allow them to be embedded into commercial accounting software (rather than only live in the ATO systems).

We already expose to the market our benchmarks for particular industries to “nudge” compliance and identify outliers.

Separately, at a “phase 2” level, we are accessing new data sources to check against small business income tax and GST returns, such as merchant data, transactional accounting information and platform data (such as sharing economy organisations).

For completeness, I note that Australia is not moving down the trend of requiring all invoices to be fed through the tax system. Putting aside broader political or social appetite for increased data provision to government, ATO analysis shows that once income “hits the till”, Australian taxpayers are highly compliant. As such, the “compliance dividend” of a more tightly controlled system may not be as high as in other countries.

VAT and top down analytics

Much taxpayer focus on data and analytics in the VAT arena has been around “bottom up” data mining for miscoded items.

At least in the Australian context, the key administrator concern in relation to large companies is that VAT or GST systems are “bottom up” with no natural sense check. In practice, we see even the largest companies making simple data/system errors (for taxable businesses, double counts or omissions). For input taxed businesses, the challenges of misallocations are akin to transfer pricing or mis-pricing challenges in income tax – there is no natural data source to check against.

As such, our data and analytics focus is in developing top down reasonableness checks for taxable businesses, and “nearest neighbour” type benchmarks for input taxed businesses. Ultimately, these checks should be part of a company’s corporate governance systems and reviewed by the ATO.

Data and automation ethics

Like many Government and non-Government organisations, we are in the course of significantly bolstering and formalising our data and automation ethics framework.

Our current thinking is shaped around six principles, supporting the reason for our data activities, namely that we conduct data activities for the purpose of administering and ensuring the integrity of the tax and superannuation systems for the Australian community. This includes making interactions with us easier, and identifying and engaging with those who are not doing the right thing.

Throughout our data activities (source, ingest, manage, integrate, use and share) the ATO will abide by these data ethics principles (when finalised) to improve government service delivery and to build and maintain public trust and confidence.

Our (working draft) principles are as follows:

• Act in the public interest, be mindful of the individual

We understand our actions will impact the community as a whole as well as you, the individual. We will be clear about the intent of our activities.

• Transparency and Explainability

We are open and will communicate with candour to you and the community about our data activities.

• Ensure privacy, security and legality

We respect your privacy, we ensure that the individual and community information we hold is kept safe, protected and shared securely as authorised by law.

• Purposeful data activities

We engage in data activities that are necessary to perform our functions, while balancing the impacts on you, the community and the ATO.

• Maintain human oversight

We will maintain human accountability for, and oversight of, our data activities.

• Maintain data stewardship

When we share your data, we will maintain ongoing stewardship and integrity in the use. We engage with others to uphold these principles.

What does all this mean for taxpayers and their advisers

Ultimately a tax system is not an end in itself, but about a country collectively sharing our resources for the benefit of all. Australia is fortunate to have a robust tax system and an effective tax administration which supports the social benefits Australians all enjoy.

Crucially, beyond their tax contribution, large companies play a critical role in the tax system as community perceptions of their compliance underpin willing participation in other markets – especially to small business and individual taxpayers. We know the community’s interaction with the tax system is directly impacted by their perception of compliance of the large market: tax compliance by large corporates sends a clear signal to other taxpayer groups about fairness, and ultimately leading to social licence to operate.

I hope my reflections today have been useful in helping you gain a better understanding of an administrator’s perspective, and might prompt a few reflections of your own:

• Understanding a country’s reliance on corporate tax and the broader perspectives of the community
• Understanding changing administrator aspirations and performance measures, including new measures like “tax gap” and “tax assured”
• How administrator philosophy, practices and internal metrics may change over time in response to these more sophisticated performance measures, including an increased emphasis on sustainable improvements in system health and a more proactive impact on the system and its participants
• The role of tax transparency and its different aspects, not simply transparency to the public of company by company information
• Some thoughts on where developments in data and analytics and automation might take administrators dealing with the large market (and where it is less likely to go).

As a final comment, as an administrator we are seeking to provide greater levels of certainty in the Australian tax system in general and to large companies in particular (importantly past, present and future certainty). This benefits the system as a whole, but also assists taxpayers that are trying to do the right thing, as it:

• provides confidence that expected tax outcomes will ultimately be achieved
• helps our largest corporate taxpayers avoid taking inadvertently risky positions
• reduces compliance costs for compliant taxpayers
• provides a level of comfort that there is a level playing field, and (perhaps most importantly)
• makes Australia a more “tax certain” and attractive place in which to invest.

Appendices

Appendix A

Appendix B