Risk-mining the public exchequer: some follow-up points. Guest post by David Quentin
Introduction
A few months ago I published a discussion paper called Risk-mining the public exchequer, the thesis of which breaks down, in summary, into these two propositions:
The distinguishing feature of "tax avoidance", as a taxpayer behaviour to be contrasted with "legitimate" tax planning, is the deliberate creation or introduction of tax risk (e. the risk that you might owe more tax than you say you owe).
The opprobrium attaching to the label "tax avoidance" is applicable in the circumstances of deliberately-introduced tax risk, because introducing a tax risk factor in order to assess yourself to fall on the right side of it is to create the possibility of underpaying your tax.
I need to stress at the outset (to forestall a common misunderstanding) that merely adopting an uncertain filing position is not enough to constitute tax avoidance in my analysis. Risk-mining the public exchequer (which is the name I give to tax avoidance as defined in my paper) is a two-stage process: you deliberately introduce a tax risk factor in the way you arrange your affairs, and then you subsequently assess yourself to fall on the right side of it. Any adoption of an uncertain filing position could be said to create tax risk, but that tax risk is only deliberately created for the purposes of my analysis if the risk factor in question was introduced into the taxpayer's affairs pursuant to deliberate prior tax planning.
Jolyon has very kindly invited me to write further on the subject on his blog, since a guest blogger of his has addressed it below, and I thought I would also take the opportunity to pick up some points raised elsewhere.
Evidence
An assumption that my paper relies on is that tax planning (whether or not it amounts to "tax avoidance" by whatever definition) is a discrete and identifiable input into the eventual form that a transaction or a business structure takes. I did not expect this to be a controversial assumption, and I have never known it to be challenged in circumstances where tax planning is said to merge indistinctly into tax avoidance in that one-dimensional continuum of increasing taxpayer aggression that one so often hears about.
What I do in my paper is plot that one-dimensional continuum (expressed in terms of amount of tax said to be saved) onto a two-dimensional graph with filing position certainty on the other axis, and show that what is supposed to be an undistinguished linear continuum in fact has a distinct kink in it where the "legitimate" planning stops and the avoidance starts. Since this challenges the conventional idea that tax avoidance cannot be objectively distinguished from "legitimate" tax planning, it has caused some to wonder if the entire category of tax planning (at whatever degree of aggression) can really be said to have an objective real-world referent, distinct from the other inputs into the form of business transactions and structures.
One such post-structuralist is Iain Campbell, who wrote up his thoughtful and very welcome response to my paper in a comment on Andrew Jackson's "Render Unto Caesar" blog here. "Where" Iain asks "is the evidence the structure was put in place, not from commercial/business considerations, but from acting in accordance with tax advice that created [tax] risks?"
I should emphasise that my paper was a theoretical one; it does not make any practical recommendations. If it were proposed, however, that my theoretical definition of tax avoidance be converted into a real-world legal test to be applied in a forensic context (for example as a component of a penalty regime in circumstances where a taxpayer is found to fall foul of his or her own deliberately inserted tax risk) then Iain's question would be an extremely pertinent one. By way of answer, I would suggest that the evidence will probably be on a server farm in someone's virtual filing system.
I say this because, in internal communications regarding a proposed transaction or business structure, the tax planning input almost always emanates from either an internal tax function or an external adviser with a specific remit to advise on tax, and it therefore flags itself up as such. Business decisions are taken by business people who take into account all relevant factors (of which tax will be one), and so their motives or purposes can be hard to unpick with clarity, but the content of the tax advice they are relying on should be discernible by reference to the documentary record. Indeed if it isn't then someone somewhere has probably been negligent.
Iain asks (with reference to the example of Google's UK tax structuring) "if the act of providing goods or services cross-border creates a tax risk, does that arise from following tax advice, or from the commercial decision?" It seems to me the question is probably best analysed as breaking down into two components: (1) what was the content of the actual documented tax advice insofar as it tracks through to the actual eventual form of the transaction or structure, and (2) did adopting that tax advice increase tax risk or introduce tax risk factors as compared with the form the transaction or structure would otherwise have taken? Clearly as bystanders without access to the full documentary record we can't answer these two questions, but I agree that we would need to feel that they are capable of being answered if we want to apply my definition in practice.
Certainty
The main purpose, or one of the main purposes, of my paper was to correct the (more-or-less ubiquitous, but false) perception that the effect of tax avoidance is to reduce the amount of tax legally payable. This ignores the fact that some tax avoidance (and it is not possible to say as at the point of self-assessment which tax avoidance) goes on to be found by the courts to fail. And tax avoidance doesn't just fail where it falls foul of anti-avoidance law: tax avoidance can fail by falling foul of any risk factor that it introduces, whether the risk factor be to do with the law, the facts, the accounting assumptions, the mechanical effectiveness of the implementation, the valuation of an asset, whatever. And if the tax avoidance is found by the courts to be ineffective then the tax avoider turns out to have assessed itself as owing less tax than turns out to have been legally payable. The tax reduction was not "legal"; it was a figment of the taxpayer's imagination. By my definition, therefore, "tax avoidance" means any tax planning in which there is a risk that the saving that the planning is said to yield turns out to have been imaginary.
The avoidance apologist will counter that there is nothing wrong with the taxpayer claiming the benefit of imaginary tax savings; the law is uncertain and the tax avoider cannot be blamed for having wrongly applied it to its position. The flaw in this argument is that it was the taxpayer itself who introduced the risk factor in question, as part of its tax planning. If you have understated your tax liability by (a) deliberately introducing a tax risk factor and then (b) wrongly assessing yourself to fall on the right side of it, the fact that tax risk factors arise naturally does not protect you from opprobrium. The risk factor in question was artificial. It would not have been there were it not for the deliberate prior act of tax risk creation.
It is for this reason that I have always thought the business sector's constant calls for "certainty" in tax law to be deeply bogus. In a world of infinite possibility but finite tax law, there will never be certainty of treatment in all circumstances, and tax law is no more to blame for seeking to apply its distinctions to commercial reality as commercial reality is to blame for coming too close to those distinctions. No, what the corporate sector seems to me to be calling for when it calls for "certainty" in tax law is something more specific than that. The rhetoric about "certainty" always seems to me to be specifically about certainty in tax planning. The business sector likes the savings that tax planning delivers but doesn't like the risk of those savings being challenged by HMRC and found not to exist.
I am happy to announce that I feel richly vindicated in this scepticism about the business sector's habitual rhetoric on certainty by Jason Piper's quietly radical recent paper Certainty in Tax, which acknowledges that aggressive tax avoidance is an extreme form of risk-creation, and that any deliberate creation of risk for the public exchequer is "open to censure". What struck me most about Jason's excellent paper, since he was writing in his formal capacity with the Association of Chartered Certified Accountants, was this sentence from its conclusion:
Tax systems should be designed so as to minimise unfair outcomes - but if the 'fairness' of tax certainty led to economic stagnation then that would be too high a price to pay.
If business sector organisations are starting to acknowledge that deliberate risk-creation is a necessary component of the tax savings that business is accustomed to obtaining, are they starting to tone down their rhetoric about the absolute desirability of "certainty"? It would appear so. This is a far more startling development than Jason's measured and unassuming prose would suggest.
There is one aspect of Jason's paper which (if he will forgive me for saying so) falls short of giving credit where credit is due, i.e. to the UK government. The paper is framed as a series of recommendations for policy-makers, but in this respect policy-makers are way ahead of business organisations. UK tax policies like DOTAS, Follower Notices, Accelerated Payments, and the freshly proposed GAAR penalty regime, are all recognition in practice that tax avoidance is a species of risk-creation, and we of the tax commentariat are only just catching up by having this debate about how risk and tax avoidance relate to each other on a theoretical level.
Abuse by the state
A key issue which Iain raises in his commentary is the question of how useful the risk-mining analysis is in the context of abuse of the international corporate tax system by multinational enterprises. I readily accept that in many cases the risk-mining analysis will only be half the story in this context. There is always more than one jurisdiction involved with double-non-taxation, and while there might be risk mining going on in both jurisdictions, there might equally be risk-mining going on in one jurisdiction and deliberate exercise of the state power to not tax going on in the other. There would be little point in Amazon risk-mining the UK exchequer from Luxembourg, for example, if Luxembourg was operating a proper domestic corporate income tax regime so as to tax the booty.
This latter category of tax abuse - states exercising their power to not tax in such a way as amounts to an abuse - is not one addressed in my risk-mining paper; my paper is about circumstances where the taxpayer is the abuser and it simply assumes that tax havens are available for the purposes of international planning. In order to theorise international tax abuse fully, both taxpayers and states need to be considered in their role as abusers.
Indeed when theorising abuse by the taxpayer, it is probably for the best to acknowledge that the state can be the abuser too, as a matter of general principle, so as to avoid giving the impression of having taken sides. I suspect I may have failed in this regard. Or, at least, if I had acknowledged the role of the state as abuser I might have attracted a slightly less hostile critique from Michael O'Connor, who raises in his guest-post below two very interesting questions about my "risk-mining" analysis, the answer to both of which is "no, that is not risk-mining by the taxpayer, that is (or may be) abuse by the state".
Before discussing those questions I should emphasise that Michael and I are talking at complete cross purposes. When he talks about tax risk, he is talking about the risk of HMRC challenge, and he focuses on circumstances where HMRC challenges filing positions which are legally correct. I adopt for the purpose of my paper a conception of tax risk which, frustratingly from the perspective of being able to have a coherent debate, excludes the very category that Michael is most interested in: I am talking about the risk of an HMRC challenge having the outcome that the tax liability turns out to be greater than the one claimed in the filing position. I am, in other words, talking about the risk of successful HMRC challenge. "It is not meaningful to describe a taxpayer making a filing that is correct in law as creating tax risk, let alone deliberately so," says Michael, and I completely agree. That would fall outside the creation of tax risk as I characterise it for the purposes of my paper.
I would urge Michael to re-read the first three paragraphs of my paper carefully, which I hope make this absolutely clear. I fear that he has been led to misunderstand my entire argument by taking the flow-chart on p.15 as his entry-point. The flow-chart exists only to make the point that wrong filing positions are wrong ab initio, rather than (as is sometimes suggested) being made wrong by HMRC challenge. The significance of this is that, where wrong filing positions go unchallenged, tax which is legally payable is lost to the exchequer. Michael himself says: "when David talks about Exchequer risk he can only mean the risk that an incorrect treatment in law will not be detected by HMRC". Indeed - and that is what my flow-chart illustrates!
Another misunderstanding between us is the one I try to forestall at the outset of this post. Risk-mining is about the prior structuring that you implement with a view to taking positions as at filing, and not about the positions you take as at filing per se. I am not saying that taxpayers can't take filing positions that HMRC might challenge, or that they have to adopt filing positions which maximise the amount of tax payable, or that they have to eschew reliefs unless it is 100% certain they are available, or anything like that. Still less am I saying that taxpayers have a positive obligation to minimise tax risk. This is about drawing a distinction between "legitimate tax planning" and "tax avoidance" at the tax planning stage, and it is not saying anything at all about the positions you might take at the self-assessment stage except to assume that, if you have structured for a tax saving at the prior tax planning stage, you are going to proceed to claim it at the subsequent self-assessment stage.
There are, however, interesting questions raised in Michael's post notwithstanding these misunderstandings. The principal question is the one raised by the difference between our usages of the term "tax risk" i.e. is it "tax avoidance" where the taxpayer takes steps to minimise tax which introduce a risk of unsuccessful HMRC challenge? My answer to this is a resounding "no". Where HMRC is known to adopt a position which is wrong in law, having the consequence that planning which does not create a risk of successful HMRC challenge nonetheless creates a risk of HMRC challenge, that is an abuse by HMRC and not by the taxpayer.
The other question which he seems to me to be raising is the question of whether it is an abuse by taxpayers to exploit strategies which are widely considered to be abusive, but which are known to be legally effective, so that no risk of successful HMRC challenge is introduced when those strategies are adopted. Again, my answer to this is another resounding "no". If a loophole is knowingly left open, it becomes deliberate policy. The continued existence of the loophole may be an abuse by the state, but once it is known to succeed the use of it ceases to be risk-mining by the taxpayer.
Editor's Note: Follow @_DavidQuentin on twitter. And here's a link to his original 'Risk Mining the Public Exchequer' Post.