HMRC: A general anti-abuse rule

Executive summary

ACCA welcomes the opportunity to comment on the proposals for a General Anti Abuse Rule for the UK tax system. While the aim of the project, to reduce the scope for abusive and fundamentally unproductive efforts to extract unintended advantages from the system, is one with which ACCA is in full agreement, we must nevertheless have reservations at both the scope and the proposed timetable of the developments set out in the current consultation.

We note in this context that proposals for a GAAR in India, frequently cited as a rival to the UK for the ‘most complex tax regime’ title, have been subject to review, and the Prime Ministerial panel has advised that the implementation of the rule be deferred for three years. The less pressured timetable will allow for a more measured approach to drafting the legislation and establishing the true impact of its provisions on both taxpayers and the tax authority.

ACCA supports the principle of certainty in the tax system. By removing the incentive to investigate wholly artificial schemes a targeted anti-abuse rule would encourage redistribution of efforts into productive activities, rather than the sterile process of seeking a pure tax advantage based on controversial or unexpected features of the increasingly complex tax system.

The initial panel proposals were tightly focused and incorporated a number of specific safeguards designed to create a regime which gave double certainty - not only to abusers of the system that they would be challenged, but equal certainty and comfort to the compliant majority of taxpayers and their advisers that tax planning in general would not be affected by the proposed rule. ACCA is concerned to note that the proposed draft legislation includes no explicit safeguards whatsoever.

By using broader existing concepts to define tax arrangements, and even lowering the test for establishing their existence, the draft legislation does not adequately restrict the scope of its potential application, and it runs the risk of increasing the administrative burden on taxpayers and in particular their advisers as they are forced to consider an extra layer of legislation when seeking to undertake any transaction which presents opportunities for alternative tax treatments.

The UK has an unusually complex system of tax incentives and reliefs built into the legislative structure, and tax is accordingly a motive for a higher proportion of transactions than might be the case in other jurisdictions. The UK needs a tightly focused and clearly delineated test to distinguish those tax motivated transactions which should be subject to further scrutiny as potentially ‘abusive’, and so reliance on the proposed broad and simple test (which is similar to those in most other jurisdictions) would not be appropriate in the UK.

The complexity of the UK system already imposes a disproportionate burden on the small and medium sized enterprises which make up the majority of UK businesses. ACCA is strongly opposed to the imposition of any further uncertainty or complexity on them and their advisers. In the absence of specific terms limiting the application of the legislation to businesses exceeding the limits for the generally accepted SME definition, there must be clarity around the application of the rule to remove the threat of the GAAR from the vast majority of everyday transactions undertaken by a typical business.

The only significant new element to the tests, the ‘double reasonableness’ formulation, is insufficiently precise to be assured of targeting only those structures which would be generally recognised as abusive. The current draft legislation would not only tackle the abuses which it has been created to tackle, it would also have the prima facie capacity to tackle a far wider range of situations and circumstances. Without the explicit safeguards and exclusions originally proposed, taxpayers cannot have certainty that the broader range of existing accepted tax planning strategies would not in future be susceptible to challenge under the new rule.

ACCA has further concerns about the proposed timetable for introduction of the rule. While ACCA supports the existence of the panel, and its role in the drafting of guidance around the rule, we are concerned that the influence of the panel on individual cases may be unduly restricted, and could with time be significantly diminished. In particular, the proposal that every panel include HMRC representation could in practice give rise to a perception of a lack of independence.

The proposed timetable for production of guidance and implementation of the rule also causes concern. Given that the makeup of the panel is not yet known, and the language of the legislation has not yet been finalised, nor would be within the proposed time frame, ACCA does not believe that this is a realistic suggestion to create comprehensive and relevant guidance.

Opening comments

In November 2011, a panel headed by Graeme Aaronson QC put forward its suggestions for a limited scope rule which would have a net effect of increasing, rather than reducing, certainty for taxpayers by ensuring that schemes which both their promoters and other taxpayers would expect to be challenged through the Courts would actually be subject to a robust challenge, and at an earlier stage and on broader grounds than is currently the case. ACCA supports the principle of certainty in the tax system. By removing the incentive to investigate wholly artificial schemes such a rule would encourage redistribution of efforts into productive activities, rather than the sterile process of seeking a pure tax advantage based on controversial or unexpected features of the increasingly complex tax system.

The initial panel proposals were tightly focused and incorporated a number of specific safeguards designed to create a regime which not only gave certainty to abusers of the system that they would be challenged, but also provided equal certainty and comfort to the compliant majority of taxpayers and their advisers that tax planning in general would not be affected by the proposed rule. ACCA is concerned to note that the proposed draft legislation includes none of these explicit safeguards whatsoever. By using broader existing concepts to define tax arrangements, and even lowering the test for establishing their existence, the draft legislation does not adequately restrict the scope of its potential application, and it runs the risk of increasing the administrative burden on taxpayers and in particular their advisers as they are forced to consider an extra layer of legislation when seeking to undertake any transaction which presents opportunities for alternative tax treatments.

The need for an overriding ‘abuse of law’ type mechanism to counteract potentially abusive or unintended tax consequences has been long debated in the UK, and we now have the benefit of being able to consider the equivalent provisions already in place in the majority of significant and sophisticated tax regimes around the world. However, when considering the principle features of such measures and their potential applicability to the UK it is important to consider not only the similarities between the UK and other regimes but also the differences.

A key distinction between systems is the extent their own level of internal complexity, and the extent to which differentiation between taxpayers is used as a mechanism for governmental policy implementation. Where jurisdictions display a significant degree of reliance on so called ‘tax expenditures’ to influence taxpayer behaviour in the furtherance of government policies, it is to be expected that there will be a large volume of tax legislation which changes on a comparatively regular basis. This will in turn inevitably lead to a greater number of potential interactions between tax provisions, and a correspondingly greater likelihood of debate over the underlying intentions of individual measures and of the combined effects of those measures.

OECD research indicates that the UK has a comparatively high level of tax expenditures within its system. This is simply a feature of the UK model of government, and is an entrenched feature of the structure of the UK economy. What it means however is that the UK exhibits a comparatively large number of opportunities for taxpayers to modify their behaviour in response to tax stimuli. The scope for individual taxpayers to influence their overall liability by pursuing slightly, or significantly, different courses of action with similar economic outcomes is considerable, and indeed almost universal. Starting from the most fundamental aspects of business structures, and the implications of different funding models, the interaction of tax with economic activity is a factor which has the scope to affect almost every decision that is made.

Where tax considerations are so fundamentally intertwined with business decision making, certainty of the tax position is key for business confidence. For any transaction which is, or could reasonably be concluded to have been, influenced by tax to be potentially open to challenge under a GAAR is reasonable only if the vast majority of transactions are not realistically susceptible to tax differentiation. If taxpayers undertaking a given transaction will be subject to the same level of taxation however they structure it, tax cannot be an influence. The greater the number of deviations from homogenous taxation a system displays, the greater will be the scope for tax to be a purpose of the transaction.

Accordingly, systems which display a low reliance on tax expenditures can reasonably tolerate a relatively broad formulation of the ‘tax purpose’ test, as relatively few transactions will be susceptible to analysis under the test. Conversely, systems which incorporate a large number of potentially tax motivated transactions will need a more tightly formulated test to identify which of those transactions might be susceptible to challenge if undesirable uncertainty is to be avoided. An overly broad anti-abuse rule might ‘work’ by its own terms, but there is a significant risk that the rest of the tax system would not continue to function effectively if taxpayers were deterred from pursuing strategies and opportunities whose ends the government might be concluded to have favoured, but which would rely for their efficacy upon exploitation of tax consequences whose validity might now be called into question. 

The UK has an unusually complex system of tax incentives and reliefs built into the legislative structure, and tax is accordingly a motive for a higher proportion of transactions than might be the case in other jurisdictions. The UK needs a tightly focused and clearly delineated test to distinguish those tax motivated transactions which should be subject to further scrutiny as potentially ‘abusive’, and so reliance on the proposed broad and simple test (which is similar to those in most other jurisdictions) would not be appropriate in the UK.

One of the few major economies not to have an existing GAAR is India. The Indian system shares many features with the UK system, not least because of their shared heritage. It has been observed that the volume of Indian tax legislation rivals that in the UK, and the sheer scale of the Indian economy encourages further complexity. Against this similar background to the UK, the Indian government has recently deferred plans to introduce a GAAR.

The only significant new element to the government’s proposed UK tests, the ‘double reasonableness’ formulation, is insufficiently precise to be assured of targeting only those structures which would be generally recognised as abusive. The current draft legislation would not only tackle the abuses which it has been created to tackle, it would also have the prima facie capacity to tackle a far wider range of situations and circumstances. Without the explicit safeguards and exclusions originally proposed, taxpayers cannot have certainty that the broader range of existing accepted tax planning strategies would not in future be susceptible to challenge under the new rule.

ACCA has further concerns about the proposed timetable for introduction of the rule and related administrative procedures and guidance. The proposals include a panel of independent advisers to consider the application of the rule, both in individual cases and more broadly in the creation of the guidance for HMRC, taxpayers and the Tribunals. While ACCA supports the existence of the panel, and its role in the drafting of guidance around the rule, we are concerned that the influence of the panel on individual cases may be unduly restricted, and could with time be significantly diminished. In particular, the proposal that every panel include HMRC representation could in practice give rise to a perception of a lack of independence.

ACCA understands that in order for taxpayers and HMRC officials to familiarise themselves with the guidance and the various new concepts introduced by this rule it is proposed that the panel review HMRC drafted guidance for approval before the end of 2012, prior to introduction of a rule from April 2013. Given that the makeup of the panel is not yet known, and the language of the legislation has not yet been finalised, nor would be within this time frame, ACCA does not believe that this is a realistic suggestion to create comprehensive and relevant guidance.

It would always be important for the guidance to be clear, measured and authoritative, and this need increases as the potential scope of application of the rule increases. The greater the number of taxpayers who might be affected by the rule, the greater will be the number of instances in which taxpayers, their advisers and the HMRC officials dealing with their affairs might need to refer to the guidance. If the potential application of the rule is to be as wide as currently appears to be the case, then it will be essential that the guidance is drafted as carefully and completely as possible, and ACCA does not believe that the current timetable would realistically allow HMRC drafters sufficient time to consider all the potential implications and interactions of the rule, let alone permit scrutiny at the appropriate level by a great enough number of panel members of sufficient seniority.

When considering the long term impact of any principles based tax rule of this kind, regard must be had of the wider legal system and its operation. Here again the UK has the benefit of being able to observe the long term impacts of rules implemented in other jurisdictions, and attempt to establish those systems whose similarities might permit a useful comparison. While it is easy to distinguish those jurisdictions which are founded in a civil or Roman law tradition, the relevant distinctions in common law systems may be harder to discern. It is however more important than ever to understand the relevant features of common law systems as it is these which are most likely to be incorporated into any UK rule. The principle feature of any tax system which will influence the effectiveness of a GAAR is not in fact one intrinsic to the system in normal operation, but one which is usually external to the operation of the tax system, namely the judiciary.

Academic research undertaken to assess the potential effectiveness of the US economic substance doctrine has identified a correlation between the broader judicial approach to statutory interpretation and the success of the relevant tax authority in using the GAAR to defeat taxpayer planning. Recent UK jurisprudence seems to indicate a general trend for purposive interpretation of taxing statutes, but a longer view shows trends in judicial thinking which can ebb and flow. This freedom of the judiciary to develop jurisprudence is of course an essential feature of the common law system. It is nevertheless a feature which can lead to a degree of uncertainty for businesses, and from an economic perspective it would be desirable to confirm in statute the extent to which tax legislation should be construed with reference to factors external to the legislation itself. Clarity as to the construction of legislation would also aid the parliamentary draftsman in preparing future legislation, leading in time to a more consistent and coherent body of tax laws.  

The issue of whether the GAAR should include a clearance system has generated much discussion, and ACCA agrees that a properly formulated rule with clearly limited application should not need any formal clearance mechanism. However, this is dependent upon the rule itself being sufficiently clear and targeted. If taxpayers and their advisers cannot be certain that the rule may not be brought to bear on their proposed transaction, then there will be an associated cost with that uncertainty. However ACCA believes that where the concerns raised by uncertainty and the need for a clearance mechanism are principally the financial cost of either uncertainty or a clearance process, the answer is not the introduction of an expensive and burdensome clearance mechanism, but rather the restriction of the anti-abuse rule to potentially abusive transactions.

The consultation document declares that HMRC will give no ‘formal or informal’ clearances as to the application of the GAAR. While ACCA fully supports the customer relationship management model based on openness and transparency (and would welcome further extension of this model and principles to all taxpayers’ affairs) there is a danger that the CRM model may lead to the perception of a ‘two tier’ model for GAAR purposes, where taxpayers would be able to discuss their concerns in advance and effectively receive clearance (or perhaps more likely, confirmation of failure) for potentially abusive transactions.

More worrying that the risk of this perception is of course the concern that increasing the incentive for taxpayers to attempt to influence the activities of their CRM. In an extreme case, the motivation of the taxpayer is clear – if they can point to evidence that their CRM considered a particular transaction ‘reasonable’ it will be impossible for HMRC to assert that the transaction cannot ‘reasonably be regarded as a reasonable course of action’ without the risk of embarrassment. The risk of ‘regulatory capture’ increases as the incentives for taxpayers to abuse the underlying principles of transparency, openness and honesty are increased. The need for clear objective signposts to which transactions might be caught and which could not is emphasised, along with guidance clarifying the scope for CRMs to discuss the potential application of the GAAR to particular circumstances.

A potential solution to the problem of larger taxpayers having exclusive access to HMRC advice would be to introduce a system for publishing relevant statements from HMRC officers, similar to the US Private Letter Ruling model. Publication of suitably anonymised details of proposed transactions, together with HMRC’s stance on whether these would prompt a challenge under the GAAR, would remove the perception of privilege accorded to certain taxpayers.

The availability of detailed guidance as to HMRC’s position in respect of particular transactions would assist taxpayers in understanding the boundaries of HMRC’s interpretation of the legislation without the need to complete a transaction and await a challenge, a clearly inefficient state of affairs. Consideration of the wider body of rulings would in time enable taxpayers and advisers to establish whether the practical rulings handed down by HMRC accord with the declared spirit and principles of the GAAR itself. Being an exercise of their statutory powers, the creation of such rulings would of course be subject to judicial review, providing a broader and more effective check on HMRC than reliance upon appeals against litigated cases.

Consultation questions

1. Do you agree that the GAAR should be limited to the taxes and duties set out in clause 1(3) of the Draft GAAR initially?  Are there any particular issues relating to how the GAAR would function in relation to the taxes (including NICs) that are proposed to be included?

The inclusion of IHT and the trust regime will significantly expand the impact of the measures and extend their reach to a far wider range of individuals than would otherwise have been the case. While the arguments in favour of controlling tax abuse in these cases is of course as strong as in corporate or other income taxes, the application of the proposed GAAR in situations where structures will have been put into place years or even decades ago will introduce significant uncertainties in the application of the rule, and could cause worry in particular for individuals who have set up their affairs for reasons other than tax in a manner which might nevertheless be susceptible to a tax interpretation.

The timescales over which inheritance tax in particular operates are such that public, and Parliamentary, attitudes can change significantly even though the relevant legislation may not actually have changed. The absence from the current proposals of any significant objective factors which might effectively restrict the application of the rule would create a situation where historic transactions could be subjected to analysis based on purely subjective criteria.

Without a specific exclusion for situations where tax can be shown not to have been the purpose of the transactions, inclusion of trusts and inheritance tax law could result in significant uncertainty for a large number of taxpayers.

3. Do you agree that (1) the proposed “main purpose” rule serves as a useful filter, when coupled with the concept that arrangements must also be “abusive” and (2) a specific exclusion for arrangements without tax intent is not required? If you think a specific exclusion is required, please explain why.

The proposed main purpose rule is too broad to act as a useful filter giving certainty to the vast majority of those engaged in legitimate tax planning. The proposals of the Aaronson report concerned themselves with safeguards and restrictions on the rule, with a view to creating a clearly delineated provision which created ‘double certainty’ – certainty for the vast majority of taxpayers that they need not concern themselves with the workings of the GAAR, and certainty for those determined to abuse the legislation that their efforts would be in vain.

It is an acknowledged weakness of the existing test that it is broad in its application, but this is justified in the instance of targeted anti-avoidance rules by the detailed and restrictive nature of the transactions to which each such rule can relate, and the tightly defined situations where the provisions can be invoked.

Moreover, it is usually the case that the main purpose test is applied factually. The requirement is for actual tax intent on the part of the taxpayer; the purpose must be shown to have actually existed. However, in the new proposals the Revenue need not show that the purpose of obtaining a tax advantage was an actual purpose in the mind of the taxpayer, merely that it would be ‘reasonable to conclude’ that obtaining a tax advantage would be a purpose of the transaction(s). The effect is almost to bring the anti-abuse rule almost to the status of strict liability which is directly opposed to the original stated intention of the proposals, that they were to counter the activities of tax planners who set out to derive an unfair advantage from the complexities and uncertainties of the tax system.

4. Do you agree that the proposed “double reasonableness” test operates as intended to counteract only artificial and abusive schemes (such as those described in Annex B)?

No. ACCA is concerned that the proposed double reasonableness test is not explicitly capable of applying only to artificial and abusive schemes (such as those described in Annex B). For a General Anti Abuse Rule to be proportionate, targeted and effective, it must apply only to those transactions which it is designed to tackle, and it must not be capable of broader general application to any tax planning. However, the term ‘artificial’ does not appear anywhere in the current draft legislation, nor any explicit reference to ‘abnormal’ transactions, or other features which clearly distinguish the schemes caught by the test from ‘run of the mill’ tax planning.

In the absence of specific exclusions and safeguards designed to restrict its application, any double reasonableness test can be applied to any planning. While it may be less likely that it will operate to catch more mainstream planning opportunities, the uncertainty created by its potential application will be harmful to business and taxpayer confidence generally.

Based on international surveys of the application of GAAR provisions, the dividing line between countries that have significantly restricted scope for tax planning, and those where freedom to arrange affairs is broader, can be traced to differences in judicial interpretation rather than in the (typically broad) underlying legislation which is being applied. The test incorporated into the UK legislation is not significantly different to those in many other common law general anti-avoidance rules, which are designed to operate in a broad manner.

Given the scope for judicial interpretation to expand or restrict the scope of such a broad rule, it is impossible to agree that, in the absence of specific statutory restriction, the current draft would only be able to apply to artificial and abusive transactions. While it could  be interpreted narrowly by the judiciary, taxpayers would have no certainty on this point, and a survey of judicial approaches to tax avoidance cases over the last 25 years would reinforce taxpayers’ concerns and uncertainty.

5. Do you agree that the counteraction provision in the draft GAAR is appropriate?

Yes, but we look forward to the advisory panel being given robust powers to advise on whether a proposed solution could be considered ‘just and reasonable’ by anyone other than HMRC. There is the potential for a scheme to be put together which cannot (within current frameworks) be counteracted through the existing provisions. To be fully effective the GAAR must be able not only to apply in theory to strike down a scheme, but also to apply in practice to reverse its effects. However, it is also important to ensure that damage done by the uncertainty around the application of the rule is not compounded by further apprehension of excessive counteraction measures being taken.

Even if there were scope for a punitive element to the regime it needs to be a separate, clear and explicit regime and not part of the counteraction and consequential amendment process. The proposal at draft section 4 (4)(b) that consequential amendments may be applied to any person is understandable, given the risk that the party receiving the initial tax advantage may, as part of the scheme, cease to exist prior to any assessment being made. The further suggestion in the consultation document that the power could be used to prevent excessive taxation is welcome, as there could clearly be situations where the reversal of the initial steps taken by the abusive party could negate transactions which might otherwise have afforded deductions to taxpayers not in any way a party to the abuse.

However, it must be clear from the legislation that a consequential adjustment  applied to a taxpayer not directly party to the arrangements would only be used to avoid double or excessive taxation. The only possible exception to this might be where the benefit of the tax arrangements has been clearly and demonstrably transferred to taxpayers who were not directly party to the arrangements, but are the successors to those beneficiaries – for example a corporate structure where the original taxpayer has been liquidated, but the effects of their tax advantage have been transferred as cash distributions or other value enhancements to the owners of the taxpayer.

6. The Government is continuing to develop its analysis regarding the appeals processes in relation to counteraction and consequential adjustments under the GAAR, and welcomes views which may inform detailed proposals to be published later in the year.

There is a degree of tension between the desire for the GAAR to operate as far as possible within the existing self-assessment regime and the proposal that HMRC alone be responsible for determining and counteractions or consequential adjustments. On this basis, it would perhaps seem more appropriate for taxpayers to have the opportunity self-assess their revised liability, with a backstop provision for HMRC to issue a determination where no such return is forthcoming.

There are of course clear administrative benefits for HMRC in operating the adjustment process, as the sole party who will have an overview of all the transactions involved and all the counteracting and consequential adjustments which may be required, in particular for taxpayers who were not directly involved in the original transactions.

It is a term of the legislation that such adjustments be just and reasonable, and in order to ensure this HMRC should consult closely with the taxpayer to ensure that the adjustment does not have unintended consequences far beyond the simple reversal of the original tax advantage. If it is impossible to reverse the tax advantage without also significantly affecting the economic reality of the transactions then this would of course suggest that the transaction had not in fact been undertaken for an abusive purpose but was in fact the consequence of non-tax considerations.

7. The Government would welcome views on the options set out regarding commencement, how transitional arrangements should be dealt with, and whether there should be different rules for different taxes where appropriate.

ACCA remains concerned that the timetable for implementation of a rule of such potentially broad scope is under any analysis too tight and consideration should be given to deferral of implementation, as has been the case in India.

Given the supposedly deterrent, rather than revenue raising, nature of the provisions, we consider it would not be appropriate for the GAAR to operate in respect of transactions predating its enactment. Such transactions will in any event be subject to legal challenge and the application or otherwise of the GAAR will make no difference to that risk of challenge. The general rule against retrospectivity in tax legislation should be respected, and the rule applied only to arrangements which begin after the date of implementation. This is particularly the case given the current uncertainty around the potential application of the rule; applying it to pre-commencement transactions would simply increase the impact of the uncertainty.

The same considerations apply to any transitional rules seeking to apply the GAAR to transactions which have started but are not completed by the date of commencement. Given the lack of currently available guidance, taxpayers would be required to predict the precise scope of the rule, and the manner in which the Panel will approach its duties. This would in turn increase the uncertainty facing taxpayers, who will not therefore be able to make an informed choice between the options available to them. This uncertainty would of course be reduced if it were clear from the legislation that the GAAR is only to apply to artificial and abusive transactions, but as discussed above we do not believe that this is the case.

8. The Government welcomes views on clause 5(1) of the Draft GAAR.

ACCA is concerned that, having removed all the preliminary filters and objective factors delimiting the application of the rule, HMRC is leaving too much for the ‘double reasonableness’ test to do. HMRC bears the burden of proving that the arrangements are ‘abusive’ (demonstrating that there are ‘tax arrangements’ is as discussed above trivial given the broad definition at draft section 2 (1)), and under the revised test everything hinges upon the definition of the term ‘reasonable’

9. Do you agree that it is appropriate for particular weight to be given in the legislation to the GAAR guidance and the opinion(s) of the Advisory Panel on the arrangements?

Yes. The Advisory Panel will be in a unique position in the tax system, as befits the unique status of the GAAR legislation.

10. The Government welcomes comments on whether particular issues arise in relation to Self-Assessment (where the relevant taxes operate within a Self-Assessment regime) or within the existing administrative rules for those taxes that do not operate within a Self-Assessment regime.

Notwithstanding the fundamental absurdity of the concept of a taxpayer self-assessing that their transaction falls within the scope of the GAAR, there are good reasons not to treat the GAAR as an integral element of the existing self-assessment regime.

Although there is no formal adoption of the tripartite separation of powers within the UK constitution (such as it is) it is nevertheless broadly accepted that adherence to the principles of independent legislature, executive and judiciary will aid transparency and enhance the perception of fairness within the tax system. Typically, this will be evidenced by Parliament enacting legislation which HMRC is charged with enforcing; only in the event of a dispute over a particular item of legislation will there be any need to refer to the judiciary. In a perfect tax system the legislation would be clear and the judiciary would have no role at all. As such, the majority of substantive (as opposed to administrative) tax laws have no need to refer directly to the existence of the judiciary, or the intention of the legislature. Even where administrative provisions do set out appeal processes and the mechanism for referring a disagreement between the taxpayer and the executive to the judiciary, it is implicit in the process that ultimate discretion as to the application of the law rests entirely with the judiciary.

However, the proposed rule conflates the operation of all three arms of the state. The assumption of a quasi-judicial role by HMRC is fundamental to the operation of the rule; it does not deal with the consequences of a disagreement over the intention of the legislature, but rather with the very existence of that disagreement. Unusually for tax legislation, the draft clauses include specific direction to the Tribunal as to the materials it must consider in reaching its conclusion, a further indication of the unique nature of a general, rather than specific or targeted, anti-abuse or avoidance rule.

As such, it assigns a specific and substantive role to HMRC in the underlying assessment of the taxes due, and operates contrary to the general principle of self-assessment. Attempting to incorporate it into the administrative machinery of self-assessment would blur the distinction between abusive transactions meriting counteraction under the GAAR and simple difficulties in interpretation arising during self-assessment. In designating a transaction as abusive, the legislation is in effect denying the taxpayer the opportunity to (ab)use the system; such abuse would by definition not be a valid self-assessment, and it is accordingly difficult to see why it should be dealt with through self-assessment.

11. The Government invites comments on the general proposal that the GAAR should as far as possible operate within existing administration rules for the taxes involved; and on what adaptations may be necessary to existing administrative rules to ensure that the GAAR operates with as little as possible additional administration cost and burden for taxpayers, advisers and HMRC. Is there a case for having a new type of assessment given the cross-regime range of the GAAR?

There are clear benefits to using the existing tax administration system to implement any adjustments arising from application of the GAAR. Even though the GAAR is fundamentally a different type of taxing instrument to the rest of the Taxes Acts, its operation is inextricably intertwined with that of those acts, and every transaction and tax effect affected by the application of the GAAR will be a part of the existing administrative structure.

However, there is a risk that the existing administrative structures may not be able to give full effect to the required amendments. While specific examples may not be easy to identify or even apparent from existing legislation, a party which has been subjected in the first instance to investigation under the GAAR will almost by definition be one who is prepared to explore the finest and most arcane aspects of tax law. While disputes over the application of administrative provisions would themselves be subject to appeal and litigation in the usual way, and given a suitably purposive approach from the Tribunals might be unlikely to succeed, the wasted energy and resource consumed by such disputes could be considerable.

This being the case, there is a good argument for the introduction of a separate power, applicable only in pursuance of the GAAR, allowing HMRC (subject to supervision of the Tribunal) to make such adjustments as may be just and reasonable to counteract the abuse without the need to operate through the existing administrative mechanisms.

However, such arguments could only be sustained if the application of the GAAR were to be a genuinely rare occurrence. Firstly, the extension of such discretion to HMRC could give them considerable powers to ‘short cut’ the properly laid out administrative procedures, and there must always be concerns that powers at this level be restricted only to those occasions which truly merit their use, and only to exercise by suitably senior officers and subject to the supervision of the Tribunal. Secondly, there should be no possibility of exercise of such powers becoming widespread in any way, nor should there applicability be of concern to the overwhelming majority of taxpayers or their advisers. The existing regulatory regime is complex and burdensome enough without imposing upon advisers the need to consider or understand any further legislation.

ACCA could only support the existence of a ‘GAAR Assessment’ mechanism if suitable safeguards were in place to ensure that it could only ever be applied in the most extreme of circumstances. However, the current low level of filter for cases which may be subject to challenge leaves open the possibility that a significant number of ‘ordinary’ taxpayers could in future become subject to the application of the rule, in which case the burden of compliance must be kept to a minimum, and adjustments must be made in accordance with the systems and processes with which they have some degree of familiarity.

12. The Government invites comments on whether time limits should be set for each of stages two, three and four and if so what those time limits should be.

While these are going to be matters of considerable complexity, it is almost inconceivable that a taxpayer might enter into any arrangement which is subject to challenge under the GAAR without having already considered in detail their position and the risk of HMRC challenge. Equally, while there may be relatively few HMRC specialists with the experience to properly handle GAAR cases, the cases themselves will be if considerable importance and a high priority to HMRC.

At each of these first two stages, a time limit of 6 weeks would appear reasonable. This will allow for other external factors such as year-end group reporting for the taxpayer, or annual leave of the relevant HMRC specialist, to be accommodated without the need for further relaxation of the limits.

The panel will of course be less familiar with the structures and the tax and commercial aspects of the particular transactions than HMRC and the taxpayers, and will be senior figures with many other commitments. A longer timeframe of say 8 weeks would incorporate scope for some ‘reading time’ in which the Panel members could consider the specific details of the transactions subject to challenge.

13. The Government welcomes comments on the proposals relating to the Advisory Panel.

The Advisory Panel should be composed of suitably experienced and impartial experts, who are able to give an opinion based both on detailed knowledge of the legislation in question and also of the wider commercial pressures and realities facing the parties to the relevant transactions. They should in particular have the necessary expertise to recognise and evaluate the impact of international considerations which may result in otherwise apparently unnecessary or unusual transactions or structures in the UK.

The membership of the Panel pool should be drawn from experts in industry, advisory practice and where possible from HMRC. ACCA believes there is significant value in the balance which can be gained from all interested parties in the administration of the tax system. There is no question that the Tribunal could benefit from the perspective of HMRC’s experience of a broad range of transaction which advisers and industry experts operating in an environment of commercial confidentiality might not be able to replicate.

ACCA does however share the concerns of some other commentators that the automatic inclusion of an HMRC representative on the Panel may lead to actual or perceived conflicts of interest which could compromise the independence and authority of the Panel. There would of course be no question of the taxpayer being directly represented on the panel, and it is assumed that representatives of the taxpayer’s advisers would equally be excluded from consideration for the Panel in any particular case.

It is often the case in large advisory firms that ‘Chinese Walls’ exist in order to ensure the independence of separate teams which might otherwise be subject to a conflict of interest. It is not clear whether such mechanisms exist within HMRC, and given the small and specialised nature of the teams dealing with many of the areas which are likely to provoke GAAR disputes it does not seem practical for HMRC to maintain independent teams which would deal exclusively with certain taxpayers’ affairs, leaving them free to act as independent experts in the case of other taxpayers. In a specialised case where HMRC has only a limited capacity to maintain a small department to deal with the area there is a risk that every case will have to be referred at an early stage to one individual, which could compromise in reality or in perception their independence on a subsequent challenge.

14. The Government would welcome views on the proposals for producing and updating the guidance.

While it seems sensible for HMRC to undertake the initial drafting, and ACCA would welcome independent review of the guidance by the Panel, we have concerns that the current proposed timetable is too tight to give time for considered drafting and measured review.

ACCA supports the suggestion that the GAAR guidance should have a special status, separate from the non-binding status usually given to HMRC’s guidance materials. If the guidance is to be of particular relevance for the Tribunal, and taxpayers are to have a legitimate expectation that HMRC and the Tribunal will operate in accordance with the guidance, then there needs to be a mechanism to ensure that it cannot be changed or reinterpreted.