On Targeted Anti-Avoidance Rules

[Readers' Note: I wouldn't normally publish a piece (reproduced from the Tax Journal, 15 October 2015) written for the technical reader on this blog. But it touches on an important and broader policy issue. If it assists the non-technical reader, a "Targeted Anti-Avoidance Rule" is a (now very common) statutory rule that (relevantly) says a transaction will attract a favourable tax consequence (or fail to attract an unfavourable one) only where achieving a tax saving is not one of the main object or one of the main objects of that transaction.]

We might, rather loosely, divide reliefs in our tax system into two types: those designed to shape our tax system – by improving progressivity, ensuring the ‘correct’ calculation of profits, and so on – and those through which the Government of the day seeks to encourage certain types of behaviour in order to advance its policy objectives. Come to think of it, there’s a third type too: those through which the Government of yesterday sought to deliver its policy objectives, and which no-one has bothered to remove. But I’ll park that whine for another day. What I want to focus on here is that second type. Government spends – perhaps more accurately it relinquishes in taxable receipts but the result is the same – north of £100bn on tax expenditures.  And it does this because it hopes, though the provision of those reliefs, to change the way in which people behave. It must do, mustn’t it, because otherwise it’s wasting that money. Now let’s add another element to the mix: Targeted Anti-Avoidance Rules. What function do they perform in relation to tax expenditures? The answer, clearly, is that they switch the relief on or off according to whether the relief is being accessed for good or bad reasons. So far so good. But what happens if the TAAR switch gets stuck at ‘off’? That’s no rhetorical question, as the Lloyds Leasing saga (see [2015] UKFTT 0401 for the latest episode) shows. I’m not going to set out here the backstory. Stripped to its essentials, Lloyds Leasing arose in the context of capital allowances, a form of tax expenditure designed to encourage capital investment, and involved the question whether the main, or one of the main, objects of capital investment undertaken by the taxpayer was to obtain a capital allowance. Now, even outside a tax expenditure context, main object tests are not amenable of fine analysis. I can understand how a draftsman might sensibly ask a court to identify the main object of a transaction. But “the main, or one of the main, objects”? How do you make sense of a formulation predicated on a three-tiered causative hierarchy (main object, one of the main objects, and not one of the main objects) and how is a court sensibly to distinguish between the second and third tiers? That’s a rhetorical question, by the way, and a good thing for you, Dear Reader, because although Judges, having no alternative, strive hard for a semblance of sensible analysis what we end up with, is cakes, icing on cakes, cherries on the icing, and now (thank you, Llloyds Leasing) “headroom… above the icing on the cake”. But profound though these difficulties are when TAARs are used to regulate liabilities to tax they are as nothing compared to the difficulties when TAARS are used to regulate the availability of tax expenditures. We have tax expenditures type reliefs because we want to alter the way in which taxpayers behave. This must be so; there is no other reason why Government would incur the expenditure. But if a taxpayer responds to the incentive, and changes her behaviour, how does she then contend that obtaining that relief was not one of her main objects? This was the question in Lloyds Leasing. Before the Court of Appeal, Jonathan Peacock QC contended, unsurprisingly, that you can’t construe the main object test in such a way as to “emasculate” the availability of the relief. You had to read it in light of the fact that the relief was designed to incentivise action. Speaking, obiter, Rimer LJ doubted this: he said it amounted to “an unwarranted suggestion that the ordinary interpretation and application of the inquiry mandated by [the main object test in] s 123(4) must in some manner be diluted.” Even approached as a matter of pure linguistic construction, I don’t think this can be right. To denude the question posed by subsection (4) (do you have a tax reduction main object?) of its statutory context (behave in this way and you’ll get a tax reduction) is to misapprehend the interpretative exercise facing a judge. But the point emerges even more forcefully if you judge the quality of that reasoning by the outcome it produces. And, fortunately for us (although unfortunately for Lloyds Leasing), we are able to do that. Because the matter then went back to the First-tier Tribunal, Mr Peacock QC ran the point again, the FTT roundly rejected it, and we were able to see the outcome it produced. At [84]-[85] the FTT appeared to regard the fact that the transaction would not have been entered into without capital allowances as somehow antithetical to the availability of those allowances. But the very purpose – the only purpose – of having a tax expenditure relief is to cause people to alter their behaviour. Yet applying the main object test in such a way denies them the relief if they do. And if, you might well ask, they would behave in such a manner absent the tax expenditure, what on earth is the general body of taxpayers doing funding the relief? (Don’t worry. That’s a rhetorical question too.) The point is made again at [87]-[88]. The FTT found that the fact that Lloyds Leasing sought tax advice to ensure that the capital allowances would be available and “structur[ed] the transactions in such a way that (as it thought) they would indeed be available” was somehow inimical to its entitlement to those allowances. But Government has defined the types of behaviour it wishes to encourage and has encouraged people to bring their behaviour within that definition by providing tax reliefs to those that do. Having so behaved, why should they not get the relief? I could go on. The problem with this approach is, of course, and even leaving aside the fact that it strips intellectual and moral dignity from witnesses asked to affect before the FTT an indifference to the very tax relief the draftsman used to encourage them, is that it fixes the tax relief switch permanently to the off position. One consequence of this will be a reduction in the cost attached to tax expenditures – but if they are to switched off, that’s properly a decision for the legislator not for the courts. And that’s the rub: by reasoning thus the courts strip from Parliament a critical tool for shaping behaviour.