Special focus: The proposed GAAR (1)

As previously reported by Tax Journal, the government is consulting on proposals to introduce a general anti-abuse rule (GAAR) targeted at artificial and abusive tax avoidance. This follows publication of a report by the GAAR study group, led by Graham Aaronson QC, in November 2011. The consultation document, which sets out draft legislation, is available to view via www.lexisurl.com/fmNLJ. The consultation will run until 14 September 2012, and written responses should be submitted by email to study.gaar@hmrc.gsi.gov.uk. A summary of responses will be published in autumn 2012, with a view to introducing legislation in Finance Bill 2013.

A selection of views from tax professionals on the draft measures is reproduced below.

Paul Stainforth, Editor

► Update 3 July: Read more at  Special focus: The proposed GAAR (2)

A general guide to the proposed GAAR

Bradley Phillips, Partner, Herbert Smith LLP

For an overview of the proposed rules download the flowchart (pdf).

The Consultation Document adopts the Aaronson Study Group Report recommendation that there should not be a formal statutory clearance process. The Government rejected the Report's suggestion that the existing statutory clearance procedures could be expanded to also enable confirmation regarding the GAAR.

In accepting the certainty of taxes (alongside death!), taxpayers generally do want real certainty and not to be left in a position of not knowing how HMRC may tax a proposed transaction, particularly in the current environment where any suggestion of tax avoidance can become front page news.

The proposed GAAR poses two key questions: is there a ‘tax arrangement’ and is the tax arrangement ‘abusive’? The first question involves applying the familiar ‘main purpose or one of the main purposes’ test and it is clear from reading the July 2009 HMRC discussion document on simplying unallowable purpose tests and the relevant case law that there is real scope for taxpayers to be in doubt as to how HMRC and the Courts may apply this test.

It is arguably easier to determine whether a transaction is ‘abusive’ as defined in the draft GAAR legislation – it is clear that the most ‘egregious’ will be caught and that transactions where there is no tax planning at all will not be caught. It is the transactions around the so called ‘centre ground’ where uncertainty will arise and in the absence of detailed and clear guidance, there is likely to be considerable disquiet around applying the ‘double reasonableness’ test.

However, even with comprehensive guidance (including input from the advisory panel) and even with having a tax adviser willing to give firm advice (not always the case!), taxpayers will obviously yearn the absolute certainty of HMRC sign-off in advance of significant commercial transactions. It is therefore going to be absolutely critical that in practice HMRC customer relationship managers do, as is suggested in the Consultation Document, provide confirmations that it does not regard particular arrangements as tax avoidance. If this happens in reality and in real time, only then may taxpayers agree that there was no need for a statutory clearance system.

Certainty or uncertainty?

Judith Freedman
Professor of Taxation Law, University of Oxford

The proposed GAAR offers greater safeguards and more predictability than exist at present under fluctuating case law.

Readers of UK newspapers over the past few weeks could be forgiven for thinking that tax avoidance is a new problem, discovered by journalists. It will be annoying to those of us who have been proposing action for many years if the proposed GAAR comes to be seen purely as a response to media comment on celebrity behaviour, but perhaps this will not matter if we end up with a sensible provision.

We have been here many times before and yet failed to produce a principle that categorically modifies the tax minimisation licence given by the Duke of Westminster’s case. A proposed GAAR was rejected in 1998 by the government, tax authorities and tax community alike and this set back the chances of changing the tax avoidance climate in the UK yet again. At around the same time, the judiciary made it clear that it was unwilling to develop a coherent anti-avoidance principle through case law, preferring to rely on statutory interpretation, albeit statutory interpretation which refers to the facts ‘viewed realistically’, which in the hands of some judges seems to give considerable scope. So now not only is the law now uncertain, but an environment has been created where, despite all the provisions about disclosure and the many specific anti-avoidance provisions, artificial tax avoidance schemes are still seen as worth trying by some.

In 2003 I argued, in Tax Journal (dated 2 June 2003) that what was needed was an overarching legislative principle to guide taxpayers, revenue authorities and the courts; a general anti-avoidance principle (GANTIP). This could then be weighed against the principle in the Duke of Westminster's case and provide a framework to guide the actions of taxpayers and revenue authorities. It would give the courts authority to carry on developing this legitimate principle through case law as necessary. The draft clauses in the government’s consultation document create such a principle and create the necessary signals to taxpayers and the courts. They do not spell out the consequences of every possible transaction with precision; to argue that they should misses the point of a principle. They do offer greater safeguards and more predictability than exist at present under fluctuating case law.

Had such a principle been enacted in 1998 we could have saved ourselves pages of legislation and much complexity. Now the process will take longer. The sooner we start the better.

Judith Freedman was a member of the GAAR study group.

Comparing the draft GAAR with Aaronson’s illustrative GAAR

Chris Bates
Partner, Norton Rose

The draft GAAR has lost some of the focus of Aaronson’s illustrative GAAR.

The Aaronson report was crystal clear that an appropriate GAAR for the UK economy was one targeted at ‘highly abusive’ schemes, and not a ‘broad spectrum’ GAAR which would potentially undermine the attractiveness of the UK as a location for business investment. It is welcome that the consultation document’s stated recommendation is the wholesale acceptance of this conclusion. The problem is how to define what constitutes an abusive transaction. The illustrative GAAR proposed in the report was a delicate balancing act. Although much of the draft GAAR is borrowed from the report, when conflated together, it has arguably produced a very different result. Whether it has been achieved by the draft GAAR requires some consideration of the detail.

In separating reasonable tax planning from the egregious, both Aaronson’s illustrative GAAR and the draft GAAR take a two step approach, the first step being to identify a ‘short list’ for consideration as a potential GAAR target. In both versions, an objective approach is taken. However, the draft GAAR relies on a familiar formulation of identifying arrangements which have as the or a main purpose the obtaining of a tax advantage, an approach expressly rejected by Aaronson. Aaronson’s illustrative GAAR started from the premise that it only applied to abnormal transactions which either had no significant purpose other than tax abuse, or contained features which would not have been included but for the main purpose of obtaining an abusive result.

What is the difference? The problem with a main purpose test is that it is a poor filter. Many areas of the tax code offer the taxpayer a tax advantage if he orders his affairs in a particular way. He may for example obtain a deduction for pension contributions and it is not unreasonable to suppose that that obtaining that deduction is a main purpose of investing in that way. Making pension contributions is not however egregious tax avoidance. Using a main purpose test risks bringing many vanilla transactions onto the short list.

The next step is to separate the ‘reasonable’ from the ‘egregious’, by identifying ‘abusive’ transactions. In the HMRC draft GAAR, one is asked to postulate whether the arrangements can reasonably be regarded as a reasonable course of action, having regard to widely drawn factors, including policy objectives and the ‘implied principles’ behind the legislation. Although the draft GAAR does include some indicia of abusive transactions such as transactions which result in a lower profit than the economic result, these may well be features of arrangements which seek to access favourable reliefs which were intended by the legislation.

The illustrative Aaronson GAAR instead identified abusive transactions by carving out from the ‘short list’ any transactions which fall into the acceptable middle ground. Presentationally, it reminds one what this is all about, separating the reasonable from the egregious. Substantively, what the illustrative GAAR carved out from the ‘short list’ were transactions where, objectively, it can reasonably be regarded as a reasonable choice of conduct afforded by the legislation. The illustrative GAAR had much repetitive drafting but this served the purpose of clearly demonstrating that it was abnormal and abusive transactions which were the target of the legislation. The draft GAAR has lost this focus and can be read as a broad spectrum general anti-avoidance provision, which is inconsistent with the policy conclusions of the original report.

The ‘double reasonableness’ test

David Harkness
Partner, Clifford Chance

The ‘double reasonableness’ test alone does not protect the centre ground.

In my view this test is the weakest point of the draft legislation. Arrangements are abusive if they cannot ‘reasonably be regarded as a reasonable course of action’. The policy intent to target only the ‘egregious’ end of the avoidance spectrum is laudable, but this test leaves great uncertainty.

The test is nebulous and highly subjective. There is inevitably a wide spectrum of what is a reasonable course of action (a spectrum that runs at least from prime minister to stand up comedian). The problem for taxpayers and their advisers is that it will often be difficult, sometimes impossible, to come to a certain view and that cannot be a good thing for our tax system.

Take the example of the issue of loan notes on a takeover. A gain may be deferred into the loan notes, but if the issuer goes bust then some rather technical tax rules mean the holder ends up having paid tax on a gain that was not received. To avoid this a foreign currency redemption option is commonly inserted to ensure non-QCB status. There is no intention to exercise this option – it is inserted solely to ensure that if the issuer defaults the holder's tax charge accords with the economic reality. Ignoring the fact that there may be evidence of an established practice, is using a 'trick' to avoid a tax disadvantage a reasonable course of action? Aaronson thinks it can be, but only in exceptional circumstances.

If taper relief were still around, what if the option is inserted to clock up more taper relief? Now you are using a ‘trick’ to reduce your tax. Would that be caught? It is not clear.

If the GAAR were interpreted in the light of the thrust of the Aaronson QC report and the HMRC consultation document that would help. Instead of the circular and meaningless statement of purpose in the current draft legislation I would prefer an acknowledgement of Aaronson's overarching principle from his report that the GAAR: ‘should target those highly abusive and artificial schemes which are widely regarded as intolerable, but should not affect the large centre ground of responsible tax planning.’

Geoff Lloyd
Director, Ernst & Young

What is ‘reasonable’?

One of the key potential benefits of a UK GAAR would be greater clarity over what is, and is not, legitimate tax planning. HMRC's draft GAAR follows the study group recommendations in treating ‘abusive’ planning as not legitimate, with the test of what is ‘abusive’ being whether or not a particular course of action is ‘reasonable’. But it makes no attempt to define (as opposed to illustrate) ‘reasonable’.

The illustrative indicators of what might be ‘abusive’ are not much help. For example there are plenty of instances in the tax code where securing a tax result significantly different from the pre-tax ‘economic’ result is precisely the intention. The difficulties in defining ‘reasonable’ solely by Parliament's intentions are highlighted in the study group report. Yet some linkage to what a Court might regard as the likely intention of Parliament – had it been aware of the arrangements in question – would seem to be a minimum level of definition needed to prevent the 'reasonableness' test being largely subjective.

As has been highlighted by the National Audit Office’s recent report on Settling large tax disputes, disputes over tax liabilities can be complex, with a range of possible ‘right’ answers. A complex tax dispute can legitimately be resolved either through a Court process or (potentially more cost effectively) through taxpayers and HMRC engaging constructively and collaboratively to test the strength of opposing arguments and reach an outcome which both sides see as robust in law. But, either way, there needs to be a clear framework of law against which opposing arguments are assessed, rather than a wishful thinking approach where each party takes a position (as HMRC officers had clearly done in some of the case studies reviewed by Sir Andrew Park) based on what they would like the law to mean.

Some objective definition of what is meant by a ‘reasonable course of action’ will be needed if the GAAR is indeed to give more, rather than less, certainty about the tax treatment of transactions. This is surely too central to the legislation to be left to HMRC guidance.

The advisory panel

Bill Dodwell
Partner, Deloitte

Why we should be concerned about the role of the panel.

Graham Aaronson was excited about his recommendation for an advisory panel to accompany his proposed general anti-abuse rule. He saw it as the most important safeguard, so that taxpayers could be assured that HMRC would not be able to apply the GAAR unreasonably.

The panel that emerges in the HMRC version seems rather less important and HMRC suggested at last week's Open Day that they do not expect many cases to be referred to the panel. The advisory panel will not have a judicial function, which must be a good thing. However, it does make it rather more obvious that the panel will simply advise. If HMRC, taxpayers and ultimately the Tribunal don’t like the advice, it may simply be ignored. This is probably the right answer, since the panel is supposed to provide a business and advisory perspective on whether or not a taxpayer’s choice is reasonable. The draft law provides that the Tribunal must take account of the panel’s views, which is a slightly odd expression for something that it may then choose to ignore – but perhaps is simply intended to make sure that the panel’s views are at least considered.

The panel will be requested to approve guidance on the GAAR prepared by HMRC, which is a better solution than asking the panel to write the guidance itself. Approved guidance has a higher level of importance than regular HMRC guidance.

The reason we should be concerned about the role of the panel is because the operation of the GAAR is based on a subjective assessment of whether or not a particular piece of tax planning is reasonable. As currently set out, almost any form of planning is potentially within the ambit of the GAAR – so a subjective assessment of whether or not it is a reasonable exercise of the choice posed by the legislation must be made. It would be much better to focus the GAAR simply on transactions where saving tax is the sole or primary purpose. That would greatly ease the role of the panel – and make it clear to taxpayers what transactions will no longer work once the GAAR arrives.

Eloise Walker
Tax Partner, Pinsent Masons

How would the panel work in practice?

Given the width of the proposed GAAR, and the absence of a proper clearance mechanism, a lot will be riding on the advisory panel – and it is being presented as a panacea that will provide impartial opinions, digests of principles and guidance to light our way.

That sounds great, but how is it actually going to work in practice? HMRC will have their work cut out surmounting the practical problems:

Leaving aside the important questions of – are they going to get paid, who's going to pay them and how much, more fundamentally – who is going to be on this panel? A pool of individuals is suggested, comprising independent members with expertise and HMRC personnel. Conflicts seem unavoidable here, and I'll be keen to hear whether HMRC panel members are proposing to overcome inevitable pressure from their colleagues. And as for individuals with industry expertise, they'll need to avoid either tax professionals with vested interests in the outcome or a taxpayer's competitor sitting on the advisory panel. How are they to be chosen to ensure impartiality – are we to have a stringent selection process along the lines of appointments to the tribunals?

The ability of HMRC to re-refer an unhelpful (to them) advisory panel decision if ‘further relevant information comes to light’ is also of concern. HMT state they want to avoid an open-ended process without resolution, but the ability to re-refer, coupled with HMRC's information powers, encourages that eventuality. It is also worth noting that whilst HMRC can re-refer an opinion they don't like, there seems to be no suggestion that a taxpayer should be able to do likewise – one hopes this is an inadvertent rather than intentional omission.

When the draft 2013 Finance Bill is published it is to be hoped we will see some further and better detail.

John Watson
Partner, Ashurst

The advisory panel and the courts.

Having taken commendable trouble to simplify many of the proposals in the Aaronson report, the authors of the consultation document lose their heads when it comes to the advisory panel. The panel will comprise a mix of HMRC officers and outsider appointees, at least one of whom has expertise in the relevant area. It will have two key roles, the provision of non-binding opinions on particular cases and the updating of guidance.

If it was just a mechanism to bring in outside experts to help HMRC understand commercial practice, the arrangement could be applauded for its good sense. Unfortunately, it goes further than this. Under clause 5(2) of the draft legislation, a court must take into account any guidance approved by the advisory panel and also the panel's opinion on the arrangements being considered.

Pity the poor judge. He has to determine whether the taxpayer's course of action is reasonable but, instead of using his own perception and abilities, he has to have a notional lobotomy and use those of the reasonable man (in those rare cases where the perception of the judge is less than that of the reasonable man he presumably cannot apply the test at all). It is not clear whether it is in his ordinary or lobotomised state that the judge has to weigh up the advisory panel advice but, either way, it is unclear what status it has. Plainly, the Supreme Court is not going to be swayed by the panel's views as to the law and for consistency that should go for tribunal chairmen too. It must be, then, the panel's view as to fact which is relevant so presumably its report is taken as evidence as to commercial practice. If that is the case the members of the panel should be available for cross-examination. We really cannot have a system where a court has to take into account the untested assertions by HMRC officers and their ‘friends’, whatever the fine qualities of the individuals concerned.

How do we escape from this mess? The involvement of outsiders is sensible but the panel should be an internal committee within HMRC. Its views should certainly be made available to the taxpayer but if the court are to rely on it its members must be subject to cross-examination.

And, while we are at it, if the purpose of the GAAR is to counteract abuse, why doesn't it simply leave it to Her Majesty's judges to interpret and apply the term. After all, that is what they are paid for.

Self-assessment

Simon Wilks
Tax Partner, PwC

The government's proposal is that the GAAR operates within the self-assessment regime. This is surprising and likely to be costly to taxpayers in practice.

Why do HMRC want the GAAR in the self-assessment regime? The government intends that the GAAR will operate in a familiar way reducing the administrative and cost burden but it follows penalties can be levied for a failure to self assess counteraction under the GAAR.

The first and compelling reason for not having the GAAR self assessed is that this undermines taxpayer confidence that it will be a narrow rule of limited impact. Under self-assessment the GAAR will need to be considered in relation to many transactions even if eventually ruled out. The additional administrative burden placed on mainstream taxpayers – not just ‘extreme avoiders’ – is self evident.

Secondly, the way the GAAR is intended to work, particularly the referral process and the advisory panel, is unfamiliar and does not sit easily with the fundamental need to be able to ascertain one’s own filing position. Not only would a person have to consider substantive law (including TAARs) but they would have to consider external factors such as the probable view of the advisory panel and whether HMRC could ‘show’ to a Tribunal that all the conditions are met for counteraction. These are not easy considerations for a self assessment regime.

Any taxpayer seeking good quality advice before undertaking planning is less likely to be liable to penalties for carelessly self-assessing that they are not caught by the GAAR. So any penalties are more likely to fall on the unwary or badly advised.

The GAAR is better dealt with by HMRC-initiated direction as was implicit in the Aaronson proposals. That would also assist with governance at HMRC. The GAAR would operate more like the transactions in securities regime (but without the benefit of a clearance). Objections could be raised that if transfer pricing can be in self assessment why not a GAAR? It should be remembered that transfer pricing was introduced in 1951 and only joined self assessment in 1999 once established and understood. That seems about right.

The role of HMRC

Peter Cussons
Tax Partner, PwC

Does HMRC's proposal that there will be no formal or informal clearances on the GAAR sit comfortably with their duty of management of the tax system?

The consultation document states that ‘The government's intention is that the GAAR should, as far as possible, operate within the self-assessment regimes.’

Para 5.16 then says ‘HMRC discusses [with large businesses and wealthy individuals] commercial arrangements and confirms where appropriate that it does not regard particular arrangements as tax avoidance. However, HMRC will not give formal or informal clearances that the GAAR does not apply’. Consequently the Aaronson proposal for GAAR clearances to be combined with existing statutory clearances where available has been dropped. This should be reinstated.

Even where a targeted anti-avoidance rule (TAAR), such as anti-arbitrage, operates on the basis of HMRC issuing a notice, HMRC operate an informal clearance system. Where however the anti-avoidance is not a TAAR but a GAAR, it is even more important for taxpayers to be able to at least informally discuss a commercial arrangement that they and their advisers think is not caught by the GAAR, but which HMRC have not covered in guidance. All the examples given in the condoc are arrangements which HMRC consider would be caught by the GAAR – there is as yet no guidance on what won't. To refuse to engage at all will leave many taxpayers (not just those whom the GAAR is trying to target) in the dark as to whether they could at some point face counteraction.

What if a UK group has a carried forward non-trading deficit (NTD) in one company, and arranges to utilise these by generating non-trade credits in the NTD carry forward company, with corresponding non-trading debits in another UK group company with sufficient current year taxable profits to offset those non-trade debits? Most groups and advisers would say this is a reasonable course of action envisaged by the legislation, and can reasonably be viewed as such. Why cannot HMRC then say so in, at least, an informal discussion? Better still guidance that is available to all, but that will still leave situations that are not covered. To offer no route for checking at least informally that a transaction is not within the target of the GAAR (for all taxpayers, not only those with CRM relationships) does not sit comfortably with HMRC's role in ensuring a fair tax system.

Yash Rupal
Head of UK Tax Practice, Linklaters

The back door clearance system.

The GAAR consultation document states that HMRC will not give formal or informal clearances on whether the GAAR applies (although the legislation is silent on this point). However, para 5.16 states that large businesses and wealthy individuals can obtain confirmation on whether HMRC regard ‘particular arrangements as tax avoidance’. What’s the difference between the two, and why are ‘confirmations’ only available to certain taxpayers?

Given the inevitable uncertainty on the ambit of the GAAR, it can be expected that most large businesses and wealthy individuals will take up the opportunity of seeking HMRC’s confirmation on whether the arrangements contemplated will be viewed as tax avoidance – a back door clearance procedure on the GAAR. There is no obligation on HMRC to publish any details of any such confirmations to the GAAR Panel or otherwise, even in anonymous form. The stated excuse for this is taxpayer confidentiality (a point which does not seem to be an issue in the US – see the Private Letter Ruling system).

It would seem that HMRC’s objective is to use the GAAR to move the tax system away from one based on the rule of law and towards a system of taxation based on HMRC’s discretion. If the current form of the GAAR is enacted, most taxpayers will approach HMRC and seek HMRC’s blessing to their proposed actions. If HMRC consent then taxpayers will proceed. If HMRC regard the arrangement as tax avoidance, taxpayers will realise that they can only proceed with the threat of the GAAR hanging over their heads – an obvious practical deterrent. The GAAR will lead to a lack of transparency and accountability in the tax system and in many cases taxpayers will have no idea of whether they have received the same tax treatment as others in similar situations due to the cloak of confidentiality.

Clause 5(3)(b) of the draft GAAR provides that a court or tribunal may take into account ‘evidence of established practice’ at that time. Given that there will be no requirement for HMRC to publicise any back door clearances, how can the taxpayer or the courts ever be sure what established practice is?

The commencement rules

Heather Self
Director, Pinsent Masons

One notable omission from the draft legislation is the transitional rule on commencement.

It is clear that the GAAR will apply fully to arrangements entered into on or after 1 April 2013 (not 6 April, for individual taxpayers?) and will not apply to arrangements ‘fully completed’ by that date.

That leaves two key questions – what does ‘fully completed’ mean, and what happens to arrangements which are not fully completed before 1 April next year?

The government clearly does not want a ‘buy now while stocks last’ rush in the first quarter of 2013 – even if it might give a temporary boost to GDP! And they seem to have missed the opportunity to make a ‘Rees rules’ announcement to apply the GAAR from an earlier date – such as the date of this year's Budget – although there must be a risk that it will be brought forward to the date of this year's Autumn statement. But getting a transitional rule which is both proportionate and effective will be difficult.

The worst situation (from the government's point of view) would be promoters selling schemes where the first step happens just before the commencement date, in the hope of pre-empting the GAAR – rather like turning ‘one spadeful of soil’ to start construction works. But at the other extreme, much IHT planning is not ‘completed’ until the original donor dies – it would be a severe case of retrospection for the GAAR to apply to IHT planning undertaken many years ago.

Perhaps the trickiest question is what will happen if an arrangement is in place, but something happens – or doesn't happen – after the commencement date. For example, will a minor change to partnership sharing ratios trigger the GAAR? I would expect the rules to be tough on this aspect, with HMRC wanting to bring prior planning into the net whenever they can find an excuse to do so.

Views from business

John Bartlett
Group Head of Tax, BP

Why I am persuaded that a GAAR is appropriate for the UK.

For the last few years I, like many Business Tax Directors, have thought that a UK GAAR was inevitable. We know that abusive tax arrangements and evasion are unacceptable and we have no time for them; but we have struggled to explain how the responsible management of taxes is a positive part of the way in which business and government operates.

Graham Aaronson’s GAAR study last year had two objectives that were critical for business:

• First, the UK’s competitive position: Ensuring that any GAAR would work fairly and would not erode the UK tax regime’s attractiveness to business;
• Secondly, how burdensome a GAAR would be – ensuring that the tax treatment of transactions could be decided without undue compliance costs.

The government’s consultation document is commendably concise and straightforward and I think that its proposals meet these objectives and form a good framework for a GAAR that would be appropriate for the UK. Why am I persuaded?

First, the government has accepted that a broad spectrum general anti-avoidance rule would not be beneficial for the UK tax system. A broadly based GAAR would I believe carry a real risk of undermining the ability of business and individuals to comply with today’s enormously complex tax regime. Graham Aaronson recommended that the GAAR would need to have a very targeted scope – focused on the abusive, contrived and artificial tax schemes, which most in business regard as intolerable. It is good that the proposals have adopted this recommendation.

Secondly, the proposals do appear to contain appropriate safeguards for the business taxpayer, limiting the potential for uncertainty. The definition of abusive arrangements that have a tax advantage as their main purpose, or one of their main purposes, is clear enough to provide adequate certainty around most business transactions. With the onus of proof being clearly with HMRC and the oversight of an advisory panel, this GAAR seems unlikely to create a climate of concern for the business taxpayer.

There are of course some important points to be explored in the consultation process. For example, the fact that the proposals do not provide for any clearances. On the basis that the application of the GAAR is restricted to abusive arrangements this seems reasonable and avoids putting a significant burden on taxpayers and HMRC. However, without a clearance process, we must ensure that the application of the GAAR is even handed across all classes of taxpayer – large and small.

The day to day application of the GAAR will be strongly influenced by guidance notes and the consultation document indicates that these will be drafted by HMRC. The content and tone of those notes will be important to how the GAAR is both interpreted and applied. It would therefore be valuable for drafts of the guidance notes to be shared and views of independent experts and the GAAR panel taken into account. Demonstrating impartiality and objectiveness will be key to the GAAR’s success.

Finally, complexity is probably the worst enemy of the UK’s tax regime. So it is key that, once a GAAR is in place, we test the government’s resolve to simplify and reduce the existing anti-avoidance rules.

John Bartlett was a member of the GAAR study group.　

Paul Morton
Head of Group Tax, Reed Elsevier Group

The main purpose test may be difficult to apply in practice.

From a practical point of view one of the more difficult tests in the proposed GAAR may be that of determining whether the obtaining of a tax advantage was the main purpose or one of the main purposes of the arrangements as provided in clause 2(1).

As indicated on page 13 of the consultation document at para 3.8 ‘Whether a purpose is a main purpose of an arrangement is a question of fact’. In practice any material transaction undertaken by a large corporate is likely to reflect appropriate tax planning. There will be internal papers and e-mails which indicate that tax planning was undertaken. In the majority of cases this will simply reflect the fact that any properly managed large corporate will ensure that every transaction is undertaken in a tax efficient manner. If the taxpayer and HMRC need to look back to the inception of the transaction in order to determine the initial purpose it may be, as is so often the case, that there is less documentation because early discussions are commonly exploratory, oral and not well or even at all documented. Memories will fade and people will move on to new roles and it may be more difficult to precisely determine the original purpose while the subsequent tax advice will weigh heavily in the minds of the HMRC tax specialists.

The Aaronson report proposed that transactions wholly without tax intent should be excluded. A complete absence of tax intent might be difficult to prove and in any case there will be a ‘middle ground’ of transactions where the documentary evidence is simply inconclusive and both sides will struggle to arrive at a conclusion. Of course this is true of all ‘main purpose’ tests and this is an area where a further review of the practical difficulties might be worthwhile.

The GAAR and TAARs

Chris Sanger
Head of Tax Policy, Ernst & Young

The government’s response to the Aaronson report is not quite so bullish regarding the ability of a GAAR to remove the need for targeted anti-avoidance rules (TAAR).

This is not least because this GAAR is intended to act only where there is real abuse. However, it remains questionable whether any GAAR would, in reality, be able to really bring simplification to an existing tax system.

Some advocates for a GAAR have argued that this could allow the government to ‘at a stroke’ remove whole swathes of tax legislation, which would no longer be needed. However, if this did indeed happen, consider the work of the humble tax adviser. Faced with a question about whether the GAAR applied, she or he would first need to refer to the removed legislation to see if it would have applied. If so, the government clearly intends the GAAR to apply. Hence, even the removal of existing tax law from the statute book in favour of a GAAR does not allow such law to be forgotten. Instead, it is effectively immortalised in stone and becomes reference material for ever more.

Another reason why tax practitioners might prefer the retention of existing targeted anti-avoidance is the lack of a clearance mechanism within the GAAR. Many of the existing anti-abuse rules cater for giving certainty in this way and this would be a clear downside of the removal of the existing legislation.

Finally, care needs to be taken to ensure that the GAAR does not undermine the incentives built into the tax system. Whilst many of these incentives would be expected to be clear and fall outside the GAAR provisions, GAAR will now need to be considered when responding to any tax-based initiative.

So, the GAAR seems to offer little in relation to the current state of complexity of the tax system. The greater hope, as expressed by the government, is that the GAAR impacts on the attitude of taxpayers, so that they do not enter into abusive schemes and therefore the need for TAARs diminishes. This might be possible, but we could be waiting for a long time.

The litigation perspective

Heather Gething
Head of Tax Planning and Disputes, Herbert Smith

The following features of the operation of the GAAR are likely to give rise to significant ancillary litigation and increase the cost of litigation generally.

Nature of advisory panel decisions: the decisions of the panel are not binding on the parties or the Tribunal but a decision is clearly intended to influence the parties and the Tribunal. In this respect the proposed GAAR accords with the structure recommended by the study group which seemed to be of the opinion that this formulation was not susceptible to challenge by way of judicial review.

The only point of a panel review is to provide an independent appraisal of the nature of the transaction – independent, that is, of the taxpayer and HMRC and the only purpose of the decision being made known to the Tribunal must be to influence the Tribunal in an appeal. In that context, the decision of the panel is likely to be susceptible to challenge in judicial review. (The decision of the Designated (HMRC) Officer to refer the matter to the panel is also susceptible of challenge in judicial review.)

As the Upper Tribunal has, so far, shown itself resistant to hearing judicial review and tax appeals simultaneously, the judicial review will need to be conducted separately. Further, following the Supreme Court decision in Gaines-Cooper [2011] UKSC 47 (para 6) where the decision under review affects the liability to tax – which a GAAR decision would, the judicial review ought to precede the statutory appeal.

It would be preferable for the panel's decision to be determinative and to give a right of appeal as in CTA 2010 s 745 and s 750 in relation to the application of Part 15 of that Act.

The weight to be attached to a decision: The Condoc gives no indication of the weight to be attached to a decision of the panel. At an open meeting conducted by the study group it was suggested by a panellist that the opinion of the panel should be given the weight of an expert's opinion. This would be an unfair outcome unless the panel or the chairman of the panel was available to be cross examined. To counter this unfairness taxpayers are likely to seek multiple expert opinions on whether the GAAR applies for consideration by the Tribunal. This will escalate the costs, length of hearing, complexity and unpredictability of the litigation. Further, as the issue on which these experts will be giving opinion – whether the transactions can ‘reasonably be regarded as a reasonable course of action’ are mixed questions of law and fact which are strictly matters for the Tribunal to decide, there is ample scope for confusion and the possibility that the opinions of experts on this topic would be inadmissible.

The GAAR and DTAs

James Maclachlan
Partner, Baker & McKenzie

Should the GAAR be capable of counteracting UK tax advantages under DTAs?

One of the more troubling aspects of the illustrative GAAR proposed by the Aaronson study group was a recommendation that, in appropriate circumstances, it should override any UK double taxation arrangement (DTA) except where the provisions of TIOPA 2010 s 2 would prevent its application.

This would generally only be where a particular DTA article includes a limitation on benefits or other anti-abuse provision which, on a proper reading of the precise terms of the DTA and applicable OECD commentaries, excludes the application of any domestic anti-abuse provision such as a GAAR. Jonathan Schwarz in his comment on this aspect of the illustrative GAAR (see Tax Journal dated 20 January 2012) provides a trenchant critique of the recommendation from the standpoint of international law.

The government has duly adopted the Aaronson recommendation in its proposed GAAR, but goes further by concluding that it should apply to any DTA (or article in a DTA) negotiated and signed before or after 2003, whether or not TIOPA 2010 s 2 would prevent its application. The consultation document justifies the proposed DTA override by reference to OECD commentary on the Model Tax Convention and the existence of broader GAARs in other jurisdictions (eg, Australia, Canada and South Africa) which also override DTAs. It therefore concludes that the proposed GAAR accords with international law. This is, and will surely continue to be, a matter for significant debate to say the least.

An equally important question, however, is whether DTA override by the proposed GAAR is good policy if the government's objective is to achieve the most competitive tax system in the G20. The proposed GAAR, even with its safeguards, will be inherently unpredictable and uncertain in its application in the absence of any HMRC clearance facility. This will be a matter of concern for overseas investors in, and lenders to, businesses in the UK, not least those from non-OECD jurisdictions who may reasonably wonder why OECD commentary should influence when a domestic GAAR may trump the UK's explicit DTA obligations.

Another question is whether in practice the UK really needs a GAAR to prevent DTA abuse. How much of a real problem is this? The consultation document cites a couple of reported UK tax cases (one going back to 1989) as examples of such abuse, but no examples of DTA abuse are given in Annex B.

International comparisons

Jon Richardson
Tax Partner, PwC

The proposed GAAR is intended to have a narrower application than those in other jurisdictions. How does the current draft compare when we look at the detail?

Nearly all of the GAARs around the world have similar design features which require the identification of an arrangement, quantification of a tax advantage and some form of purpose test. On this count, the UK proposal of ‘main, or one of the main purposes’ looks broader than most other countries. For instance in Australia the tax advantage has to be a ‘dominant’ purpose, in South Africa it is the ‘main’ purpose and in Brazil it is the ‘sole’ purpose.

However in the proposed UK GAAR we have the additional requirement that the tax arrangement has to be abusive. When we look at other countries we find that this additional safeguard is not unique. Several overseas GAARs include a misuse or abuse provision which requires an arrangement to produce tax results that are not consistent with, or otherwise defeat, the object, spirit or purpose of a particular tax provision. The Canadian, South African and the proposed Indian GAARs have such a provision. In Canada it is this requirement that has given rise to the most difficulty in the interpretation and application of the GAAR, notwithstanding that it is tempered by the word ‘reasonably’, suggesting some judicial leeway in determining abuse.

So despite the stated intentions that the proposed UK GAAR is to be more narrowly drawn, this does not appear to be the case based on the legislation alone. It looks very similar to other GAARs around the world in terms of its ambit. It is worth noting that when the Australian GAAR was first introduced the policy intent was expressed as to ‘strike down blatant, artificial or contrived arrangements, but not cast unnecessary inhibitions on normal commercial transactions by which taxpayers legitimately take advantage of opportunities available for the arrangement of their tax affairs’. I suspect there are many companies, including some well known UK plc’s, that would strongly argue that this has not been their experience of the Australian GAAR in practice.

For further detail on the key features of GAARs in other jurisdictions, download the table (pdf).

Back to top