HMRC’s anti-avoidance strategy: the next squeeze

There has been previous comment in Tax Journal on the two significant generic anti-avoidance measures in Finance Act 2014, with strong objections being raised by some commentators to the executive giving itself the power to act as judge, jury and executioner. To understand why this is, we need to be familiar with the challenges HMRC has faced in responding to tax avoidance. This also points the way to future developments in anti-avoidance.

History

The history of tax avoidance goes back as far as the history of taxation, but this is not an academic lecture and there is no need to go back that far: two years is sufficient for present purposes. In July 2012, HMRC published Lifting the lid on tax avoidance schemes. At one level, this consultation was no more than a continuation of HMRC’s longstanding battle against avoidance, but at another it marked some very significant steps.

Firstly, the name of the document itself was significant – gone were the days of staid titles to consultation documents that described in technical terms what was proposed (c.f. A general anti-abuse rule, issued only a month earlier); this was HMRC deliberately becoming much more ‘in your face’ on tax avoidance, albeit not quite ‘gotcha!’

Moreover, the tone made it clear that HMRC saw a clear distinction between, on the one hand, reputable tax advisers and taxpayers who had noted the change in the tax environment and were no longer engaged in tax avoidance (if they ever had been) and, on the other hand, a minority of tax advisers and taxpayers who had yet to see the light. HMRC therefore sought to build consensus and a sense of community with the reputable to help it respond to the minority. A significant part of this was a move away from the traditional reliance on amending legislation to remove avoidance opportunities (with operational activity to investigate and collect back tax, and litigation where necessary). Instead, HMRC would use subtler approaches to discourage avoidance and encourage good behaviour. The main tool for this, as set out in Lifting the lid, was communication.

Shortly after Lifting the lid, HMRC reinvigorated its Spotlights series in which it gives early warning of avoidance that it believes is not legally sound. Clearly, where HMRC finds schemes that it thinks may work legally, it has always been willing to invite ministers to announce prospective (or rarely, retrospective) changes in the law to close the loophole. Where the scheme doesn’t work, this avenue may still be available to put the matter beyond doubt, but HMRC was seeing schemes where there was no reasonable doubt about the law and yet promoters were still selling them. This was, at least potentially, a form of mis-selling in which the taxpayer risked being left with a large tax bill on top of the fee for the ineffective scheme.

HMRC was even more concerned about the impression created by the publicity about tax avoidance. The risk was that taxpayers would believe everyone was avoiding tax. If they got this impression, their willingness to pay tax voluntarily would be damaged, and this would be a major problem in a country where the norm is that 90% of the tax due under the law is paid without any need for HMRC intervention.

HMRC’s response to this was to begin issuing press releases in the late summer of 2012, seeking to generate publicity for its tribunals and court successes against tax avoidance. While HMRC is not always right and has had some setbacks in cases such as Rangers EBT (Murray Group Holdings and others [2014] UKUT 0292 (TCC)) and the UBS and Deutsche Bank cases (HMRC v UBS AG; DB Group Services (UK) Ltd v HMRC [2014] EWCA Civ 452), it has succeeded in the large majority of loss generation schemes, such as Tower MCashback [2011] UKSC 19, Eclipse Film Partners No. 35 LLP [2014] EWCA Civ 184 and Icebreaker (Icebreaker 1 LLP) [2011] STC 1078 and Acornwood LLP and others [2014] UKFTT 416). A drip feed of publicity for these wins was intended to bolster voluntary compliance.

By the autumn of 2012, the combination of success in court and various legislative changes, plus the public awareness and rejection of arrangements entered into by multinationals and rich individuals that were perceived as tax avoidance, meant that reputable advisers and their clients had largely, if not completely, stepped away from new tax avoidance.

The National Audit Office

But this left a legacy of past avoidance. HMRC had been effective in turning down the flow of new avoidance, but it also needed to turn up the flow of settlements of enquiries into past avoidance. The NAO published its review, Tax avoidance: tackling marketed avoidance schemes, in November 2012. The key number disclosed in this report – flagged as a key fact on page 4 ahead of the summary of findings – is that there were 41,000 open cases. Inevitably, this was seen as a ‘backlog’ of work that needed tackling and it resulted in pressure on HMRC and government ministers to do something to reduce it.

HMRC’s analysis was that the large majority of these open cases were not due to a backlog of unworked post, as the critical commentary seemed to suggest, but as a consequence of the way disputes were litigated and the incentives this gave. A large proportion of the 41,000 cases were accounted for by a small number of scheme types, e.g. those who had participated in loss generation schemes, SDLT schemes, EBT-type arrangements to avoid tax/NIC on bonuses, etc. HMRC had around 20 operational anti-avoidance projects with at least 1,000 participants and its practice had been to litigate a small number of lead cases in the hope that this would resolve the legal issues and allow all the cases to be resolved. But its experience in practice was that the litigation settled one case, leaving 999 asserting that they were different. So litigation was a slow and unwieldy tool and the pool of disputes was growing as the number being settled was outpaced by the number of new enquiries being opened. This problem was only mitigated by the fact that the subsequent 999 cases tended to concede before a tribunal heard their arguments, but only after lengthy preparation for such a hearing. HMRC needed something to change the incentives.

Raising the stakes

The answer was set out in Raising the stakes on tax avoidance, a further consultation in August 2013. As well as measures to turn up the heat on so-called high-risk promoters, it set out a new penalty aimed at followers, i.e. the cases that HMRC believed ought to have conceded immediately following the decision in the lead case. The use of penalties was an obvious mechanism to achieve HMRC’s objectives. There was already a penalty for submitting an incorrect income tax return (under what used to be, and for relevant periods often still was, TMA 1970 s 95 and its equivalents for other taxes). This penalty would in principle be triggered if tax avoidance was unsuccessful, since the taxpayer would in most cases have submitted a return that turned out to be incorrect. But these penalties did not apply in the large majority of these cases because they focused on the taxpayer’s behaviour at the time the tax return was submitted. Taxpayers would normally have taken sufficient professional advice that they could not be said to have failed to take reasonable care and hence TMA 1970 s 118(5) protected them. What was needed was a mechanism to require taxpayers to reconsider their filing position following the lead litigation and to take reasonable care in doing so. A penalty could then be imposed on those who had no real arguments, without having any effect on taxpayers who either genuinely believed that their circumstances were different or accepted they were the same but wanted to challenge the decision in the original lead case through further litigation.

But the mechanism set out in Raising the stakes on tax avoidance was different. Under this alternative mechanism, HMRC issues a notice which requires the taxpayer to amend their return to give effect to the court or tribunal decision. So taxpayers will have a choice. If they accept that the lead case determines their dispute, they can concede. But if they do not, they must put themselves in a position of having failed to comply with a tax compliance obligation; many reputable taxpayers will be very reluctant to do this. However, many large corporates have tax strategies that commit them to complying with tax obligations, so a proposition that they should fail to comply creates major, unintended difficulties for them. Those who argued HMRC would become both judge and jury had a point, because failing to comply with a follower notice leads to one of several possible outcomes:

1. The taxpayer eventually prevails, whether or not litigation is needed. Clearly no (tax-geared) penalty can be levied as no additional tax is due.
2. The taxpayer eventually loses, whether or not litigation is needed, on a basis that is different from the reasoning in the original litigation. For example, a court may overturn a tribunal decision on the interpretation of a section of the Taxes Acts, but still find against the taxpayer because a separate anti-avoidance rule applies. Clearly, no penalty should be imposed in these circumstances because the conditions for the issue of the follower notice were not met and HMRC will withdraw it. Nevertheless, many responsible taxpayers will find it uncomfortable that they have technically breached a tax compliance obligation, and constitutional lawyers will continue to point out that this scenario is not consistent with the rule of law.
3. The taxpayer eventually loses in litigation on the same basis as the reasoning in the original litigation. This could either be because they advanced legal arguments on the same basis as category (2) above but were unsuccessful in getting the original decision overturned, or merely that they argued their facts were different and the tribunal disagreed. On the clauses in the Finance Bill as introduced, taxpayers would have had to rely on HMRC not pressing for a penalty, since there was nothing in the legislation that would rescue them.
4. The taxpayer pursues their dispute but subsequently concedes, without further litigation, on the same basis as the reasoning in the original litigation. This scenario will look to HMRC like precisely the behaviour the follower penalty was designed to stamp out. But there will clearly be a difference between setting out from the start with no arguments and no intention to litigate (i.e. an intention to delay settlement in the hope that something may turn up) and setting out with an intention to litigate if necessary but changing that intention in the light of events elsewhere. If this is a re-appraisal following the appointment of new advisers or new management or a simple change of heart, HMRC is unlikely to have much sympathy. But if it reflects a new legal analysis, perhaps following developments in non-tax case law (e.g. a decision in an intellectual property case that undermines a claim to tax relief), there may be more scope to put together a persuasive argument that HMRC should not press for a penalty.

The good news is that amendments introduced at Report Stage helpfully enable the taxpayer in all these scenarios to appeal against the penalty on the grounds that ‘it was reasonable in all the circumstances ... not to have taken the necessary corrective action’. Clearly, it is reasonable not to concede (i.e. not to ‘take the corrective action’) if a reasonably advised taxpayer would believe there was a litigable argument despite HMRC’s success in the case that led to the follower notice. HMRC will achieve its objectives because those who are merely spinning things out will be unable to meet this test of reasonableness.

Accelerated payments

The objective of this follower notice measure is not to raise revenue through penalties: it is to change behaviour and drive a reduction in the 41,000 cases – since reassessed at 65,000. But this measure will be limited in its impact because it cannot apply until a lead case has become final. There are large numbers of disputes where the criteria for HMRC to issue a follower notice will not apply. One way to respond to this might have been to widen the criteria for follower notices, but the criteria are already wider than respondents to the consultation were comfortable with. These include extending follower notices to cases which are final at the First-tier Tribunal, despite its not being a court of precedent; limited appeal rights; and a wide ambit that goes beyond mass-market avoidance. So HMRC believed it needed a further measure to create leverage more quickly on the rest of the legacy avoidance cases.

The government announced at Autumn Statement 2013 that it would introduce a measure requiring the accelerated payment of the disputed tax in certain avoidance cases. There is a simple cash flow element to this: the chancellor would like to further reduce the deficit by bringing in disputed liabilities as early as possible. There is also a simple behavioural element: the government would like to discourage future avoidance and it believes that some schemes have been sold on the basis of the cash-flow advantage from an ineffective avoidance scheme (pay in 10 years’ time and have the benefit of the cash at an interest cost that is cheaper than alternative methods of borrowing). Whether any avoidance was ever sold on quite such an explicit basis is beside the point; this measure is intended to alter the calculations of anyone still thinking of undertaking avoidance.

There is a more subtle behavioural element too. One of the reasons it is hard for HMRC to settle legacy avoidance is that anyone who is faced with signing a cheque at the end of a dispute will be inclined to hold out and keep hoping; it is easier to settle with those hoping for a repayment that has been withheld. This is basic human psychology: we are all more willing to give up the prospect of a future gain than to accept a loss. In forcing taxpayers to hand over the disputed money, HMRC is therefore trying to move them from the mental state that causes them to cling onto their cash to a different mental state where it will be easier for them in future to give up the hope of a repayment.

The initial proposal was linked to the follower penalty, since this was the least contentious definition of avoidance. As long as HMRC is trusted to identify the right cases (not everyone would have that level of trust!), these are disputes where the taxpayer has relatively little chance of prevailing since a court or tribunal has already agreed with HMRC’s arguments in another case. The government is saying, in effect, that these taxpayers have the right to argue about their tax liability, but that the normal rules of self-assessment should be disapplied and they should pay now and receive a repayment later if their arguments prevail. In effect, this puts the followers in the same position as a litigant who has lost at a tribunal but wishes to take his or her appeal to the higher courts.

This does not expand the pool of cases that the new tools can apply to, though. The real prize which HMRC is seeking is only available if a wider definition of avoidance can be found, taking accelerated payments beyond cases with follower notices. The Autumn Statement therefore announced consultation on extending accelerated payments to schemes that are disclosable under the disclosure of tax avoidance scheme (DOTAS) regime, as well as those which HMRC is counteracting under the general anti-abuse rule (GAAR). The second category is self-evidently sensible but irrelevant in practice, since none of the legacy cases can be subject to GAAR counteraction – we have yet to see the first GAAR cases.

The extension to DOTAS has been criticised because, in some respects, DOTAS is more a definition of innovation than of avoidance. Certain schemes may have been disclosed through caution, even though arguably they do not fall within DOTAS. Therefore, arrangements that cannot be regarded as avoidance may still have a DOTAS number and hence be caught. HMRC will no doubt respond by pointing out that:

• it has cleansed its DOTAS list and removed from scope those disclosures it accepts are not avoidance;
• for any non-avoidance that slips through this net, it is unlikely there will be a current dispute so nothing much is likely to rest on the point; and
• this is about who holds the money while a dispute over liability to tax is resolved, so it is less important to be absolutely precise with the boundaries of accelerated payments than it is with a taxing provision.

The more significant criticisms, including from the Treasury Select Committee (see its Budget 2014 report), concern the retrospective operation of the accelerated payments measure. There seems little chance of persuading the government to alter its stance: the whole raison d’être of the measure is to help drive legacy avoidance cases to resolution. It is at least as much directed at the past as the future. And HMRC will justify this by pointing out, as noted above, that it is about who holds the money while a dispute over liability to tax is resolved. The liability – if it exists – is not being imposed retrospectively, since any resolution of a dispute that leads to a liability will, by definition, establish that the liability was already there. Of course, there will be a new payment obligation that did not exist before royal assent to the Finance Bill, but that is imposed prospectively. Taxpayers will challenge this analysis and it will be interesting to see whether the courts find that the new law exceeds the discretion that they allow to governments to raise taxation. Moreover, to the extent that they allow these changes, the balance will shift: frustration at the lack of progress towards a resolution of disputes, currently felt by HMRC, will be felt by taxpayers, unless HMRC is able to meet their expectations.

Where next?

So what happens next in anti-avoidance strategy? It is already clear that the response to avoidance is multi-layered, including:

• a communication strand that is aimed at discouraging tax avoidance and underpinning the high levels of tax compliance in the UK;
• a legislative strand that aims at making tax law more robust against avoidance than it has been in the past and seeks to ensure that HMRC has the powers it needs to both deter new avoidance and bring legacy avoidance cases to resolution;
• an intelligence strand to ensure that HMRC has the information it needs, in real time, to react quickly however the tax avoidance market develops; and
• heavy artillery through the high-risk promoters strand to make life very difficult for any adviser that wants to continue promoting tax avoidance in spite of everything.

Reputable advisers will welcome this last strand as long as it is well targeted, as it is very much a minority sport. It is unlikely HMRC will have hit the target at the first attempt and there will therefore be more work to do to refine the regime.

But there will continue to be stories in the press that concern both past behaviour – probably triggered by litigation, bearing in mind how long it takes cases to come to court – and current arrangements that the public perceive as unfair. So HMRC will continue with its communications in the hope of reinforcing its messages. Those involved in litigation should take account of and, where appropriate, prepare themselves for the possibility of publicity in the event of the tribunals or courts finding against them, even if they subsequently succeed on appeal. And we can expect ministers to continue to respond to the pressure with new announcements at party conferences, Pre-Budget Reports and Budget Statements. Perhaps a summer 2014 consultation on Tax avoidance: the next squeeze will offer some ideas – it will undoubtedly include a discussion of updated powers to enable HMRC to ensure compliance with DOTAS.

Finally, what is the future for follower and accelerated payment notices?

HMRC will take some time to implement follower notices. It needs to do so carefully, using a central coordination group, to ensure that it does not misuse and tarnish its powerful new tool. There is no obvious scope for developing this; a broader penalty for failed tax avoidance has always eluded HMRC because of its incompatibility with the rule of law. HMRC may, though, start to look for other ways to expand accelerated payments, by adding to the list of circumstances that trigger payment.

Having crossed the Rubicon of inconsistency with self-assessment (the fundamental reason HMRC’s 2011 consultation on an earlier cashflow proposal foundered), could a much broader definition of avoidance be introduced? Perhaps it could draw on the building blocks of the GAAR to sweep into accelerated payments any arrangement HMRC considers to be contrary to the spirit of the law? There is no obvious technical reason why this could not be done.

The question is whether ministers have the will to go further, and this in turn will depend on how they balance the pressures from commentators who still believe the UK is too soft on avoidance with that from businesses, who need certainty to invest in the UK and create jobs.