Defining tax avoidance

I think that if I’m going to talk about tax avoidance, I’d like to know what it is I’m talking about. So I’d like to havea look at what other people use “tax avoidance” to mean, and then come up with my own preferred interpretation of it.

Robert Maas, in a recent article in Taxation magazine, had a look at a number of recent “tax avoidance” cases from the news. He concluded:

“It seems to me that most of the so-called avoidance schemes that are being publicly criticised are not avoidance at all, and those that are were unofficially endorsed by HMRC at the time that they were entered into. So what is there to debate?

Obviously that last question is rhetorical. There is a lot to debate but there is also a need to refocus the debate to real avoidance.”

OK, focusing the debate. To me, the first question is: how do we define “tax avoidance”?

The Oxford Dictionary on-line says tax avoidance is:

“The arrangement of one’s financial affairs to minimize tax liability within the law.”

That’s a bit broad, really. I think it needs narrowing down a bit.

I’m going to refer to HMRC and Parliament here because I live in the UK and work mostly on UK taxes, but the arguments will extend to overseas authorities too. And I’m coming here from the perspective of a tax advisor who is going to have to tell a client whether they’re paying the right amount of tax. I want to be able to tell my client what tax they should be paying, why it’s more than they think they ought to (it always is), and how HMRC might be able to argue that it’s insufficient. To do that I need some certainty over the position.

Existing views

There are three sets of people involved in this debate:

– Those directly involved: the authorities (HMRC and Parliament) and the taxpayer;
– Those involved in the process as agents; and
– Those who are interested in the process but not involved: the media, tax campaigners, and the public

Taxpayers and the public don’t seem to say much collectively, largely because their voices are wrapped up in one of the others.


HMRC say you can’t define avoidance properly. They say it’s part of the tax gap, which is “the difference between the tax collected and the tax we think ought to be collected”; this presumably includes evasion, which is “when people deliberately don’t pay the tax they should” and is criminal. So by elimination we find that tax avoidance is paying less tax than HMRC think you should, without doing anything which is clearly illegal.

However, they’ve also said (in “Lifting the lid on tax avoidance”, a consultation document mostly about improving the DOTAS scheme):

“Tax avoidance is bending the rules of the tax system to gain a tax advantage that Parliament never intended. It often involves contrived, artificial transactions that serve little or no purpose other than to produce a tax advantage. It involves operating within the letter – but not the spirit – of the law”

So it’s about using the law in a way that Parliament didn’t intend it to be used, and the implication is that HMRC know that intention.


Parliament don’t really seem to have a definition of avoidance. The PAC has talked about it, and backbenchers have debated it, but those discussions either use the HMRC definition or skate over the details.

Which is interesting for HMRC’s definition: how can anyone know what Parliament intends if the MPs themselves don’t?

I also find it interesting (and rather worrying) that the PAC seems to regard itself as the voice of the public, rather than as being part of Parliament. Margaret Hodge doesn’t seem to think she has any legislative power, for example.

Tax agents

In their response to the HMRC consultation document, the Chartered Institute of Tax says:

“The CIOT fully supports the tackling of ‘…contrived, artificial transactions that serve little or no purpose other than to reduce tax liability’. We do, though, think that not all avoidance falls into this sort of category.”

Which seems to be saying that although you can recognise some schemes as avoidance, the term is broader than that – but it’s hard to see quite how much so.

Other organisations seem to follow much the same line, or stay fairly quiet about it.

Tax scheme promoters tend to say that if the legislation doesn’t stop you then go for it, although they seem to be happy to call what they do avoidance: they just don’t think avoidance is wrong.  That’s another debate, of course.

The media

The media tend to either go with quotes from HMRC or tax campaigners, or to simply quote tax as a percentage of profit – or of turnover, which is a bit irritating.

Tax campaigners

Richard Murphy has a lot to say about tax avoidance: so much that I’ve carved out my analysis of some of his comments into another post.

To summarise: he generally uses the “Parliamentary intention” formula when defining tax avoidance, but when identifying particular cases of it he normally compares tax paid to a notional “fair amount”, generally using his “unitary taxation” concept.

So I think the best way to bring this together is to say that he seems to consider tax avoidance to consist of paying less tax than Parliament would intend if they adopted his principles.

Christian Aid, in a recent survey on tax avoidance (which found that people mostly regard it as bad), defined it as:.

“Tax avoidance is the use of legal loopholes to alter a person or company’s financial position in order to lower the amount of tax that they are obliged to pay. This differs from tax evasion where a person or organisation does not pay tax by illegal methods.”

This is slightly broader than the “parliamentary intention” definition, but notice that we’ve got the loaded term “loophole” being used. No-one can ever use a loophole for anything good these days, so it’s no wonder that people generally disapprove of tax avoidance.

Action Aid, when talking about ABF in Zambia, complained that ABF was following the law but this wasn’t a fair share. ETA: They refer to this as “tax dodging”, which seems like a close synonym for avoidance. The main factors in this are treaty shopping, the group’s transfer pricing position, claiming capital allowances, and obtaining a preferential tax rate aimed at farmers. All these are acknowledged to be following the letter of the law, but defying the intention of the Zambian Government whan making the rules or signing the tax treaties. For capital allowances at least (and possibly other factors) it’s not at all clear to me that the government’s intention has been defied rather than fulfilled. So the message seems to be that tax dodging certainly involves defeating Parliament’s intention, but also goes wider than that.


Pulling this together, the question “what is the right amount of tax?” is answered in a few different ways:

1) What the legislation says
2) What Parliament intended you to pay
3) What Parliament should have intended you to pay
4) A fair share
5) Your headline profit multiplied by the headline rate

Now I don’t like 2-4, as they’re subjective – 3 doubly so.  As soon as you start looking at what should be the case, or what Parliament meant to do but didn’t actually do (for whatever reason – accident, oversight, times changing) then you have a wide range of possible answers from a wide range of sources.  Fertile grounds for debate, but not so good when telling a client what he ought to be paying on 31 January.

5 is of course very simplistic (it rather begs the question of what “profit” is); I’d quite like it if we could get closer to it, but that’s a different debate.

The only answer that gives you any certainty is 1. And given that Parliament writes the laws, then if 1 doesn’t reflect 2 or 3 then that’s Parliament’s fault.  It doesn’t give complete certainty, as legislation occasionally admits of more than one (or no) sensible interpretation, unfortunately, but it’s much better than the others.

The current judicial position is that 1 is the way to go, with 2 used as a tie-breaker in cases of doubt, and I think that’s reasonable.

Ideally of course 1 and 2 should be identical, and areas where they’re not is where we get problems.

Historically one has taken an Act to perfectly reflect Parliament’s intentions, but since Pepper v Hart and the adoption of purposive constructions it has become accepted that perhaps the legislation isn’t as perfect as it ought to be in this regard. So I’d like to have the concept of Parliament’s intention somewhere in the definition, but I think that should be taken as being the wording of the legislation unless there’s a clear indication otherwise.

Even better would be if all (except maybe 5) were completely aligned, but give that we have three sets of subjectivity in there it seems unlikely to be achieved. To eliminate the differences would require clearer legislation (to bring 1 in line with 2), some decent debate about tax in general (to bring 2 in line with 3 and 4), and maybe some simplification and elimination of reliefs and incentives (to bring the whole closer to 5).

Semantic quibbling

Another thing I think needs to be taken into account is the word “avoidance” itself. Taking a man-in-the-street definition, you avoid something by taking steps to go around it. So I think that somewhere we need to be taking that into account: to avoid tax, you need to take steps such that it doesn’t impinge on you, just as to avoid a lamp-post you need to take steps such that you don’t impinge on it.

I think if you take no steps, you can’t be avoiding something; or, rather, if you take no steps that you wouldn’t otherwise have taken anyway, you can’t be avoiding, you’re just missing it. If I walk down the left-hand side of the pavement I miss the lamp-post; if I walk down the right I hit it; if I start on the right then change to the left then I avoid it. It is the moving to the left that is the avoidance; if you were on the left anyway when the lamp-post hove in view, that is not avoidance.

I think steps taken to avoid tax differ in quality, as well as quantity. Some things you can do, like incorporating a company, have wider economic impacts than simply tax; others, such as setting up a chain of three holding companies such that A owns B owns C, have little or no economic impact: A might as well own C, or even C’s assets, directly. If you do something that has no effect other than to change the tax position, I think that’s a bit aggressive towards the tax authority.

The question of what steps you’ve actually taken (which for tax purposes may not be the ones you say you’ve taken) is where the Ramsay doctrine comes in, and all the fun and games that are to be had with that.

Definition at last

So for purposes of this blog, I’m going with the following definition of avoidance:

Tax avoidance comprises taking a step which results in less tax being paid than if that step had not been taken.

This includes as a “step” taking a step in a manner which is not the default. So if you’re a UK-resident individual setting up business, incorporating a company to do it through is a step, as it’s something you didn’t need to do; incorporating that company in France is also a step, as the default would be to use a UK company. Similarly, using a guarantee company or unlimited company would be regarded as step, as the default would normally be to use a limited company.

And I’m going to classify avoidance in five ways:

Evasive: ignoring the tax effect and just not paying it. Carousel fraud.
Abusive: achieving an advantage that goes against a stated intention. K2 scheme.
Aggressive: achieving an advantage with no associated economic impact. Some off-shore planning structures.
Simple: achieving an advantage by accepting an economic impact. Incorporating a sole trader
Benign: achieving an advantage that has been stated to be intentional. Investing in an ISA.

That really breaks down to a flow-chart made up of three key questions:

Q1) Are you failing to pay tax that’s clearly due under the legislation?
If yes, that’s Evasive (or just Evasion)
If no, go to Q2
Q2) Has Parliament stated an intention?
If yes, and you’re fulfilling it, that’s Benign (or Planning)
If yes, and you’re achieving something else, that’s Abusive
If no, go to Q3
Q3) Are you taking steps with no economic impact?
If yes, you’re being Aggressive. How aggressive depends on the number of such steps and their impact
If no, it just Simple avoidance. Join the Duke of Westminster club.

The question then is how one feels about each type of avoidance.

As far as I’m concerned:

Evasion is clearly illegal, however one slices it, and should be eliminated as far as possible.
Abuse is unacceptable, and should be eliminated as far as possible – say with a GAAR in the first instance, and then by improving the legislation
Aggressive avoidance is a concern, and steps should be taken to reduce the scope for it by improving the legislation
Simple avoidance is an artefact of a complicated economic system. Parliament should review whether the tax advantages it permits are appropriate
Benign avoidance is a complication. Parliament should consider looking for ways of achieving the same result outside the tax system

Now I also think that this works both ways, and that often tax disadvantages arise in ways that are abusive or aggressive. Not many people say much about them, though I intend to flag some of them up in the blog.


5 thoughts on “Defining tax avoidance

  1. Hi Andrew,

    I think this last part is an extremely useful anatomy. I agree that a simple tripartite scheme of ‘evasion/avoidance/everything’s fine’ often doesn’t do justice to the facts on the ground.

    I should point out, though, that ActionAid’s report on ABF isn’t simply saying that “ABF was following the law but this wasn’t a fair share”, as you suggest. If it was, I don’t think that would be “avoidance” at all.

    What we’re saying is that while the group haven’t done anything unlawful, they have established entities and transactions to generate tax advantages that the relevant laws never intended. For example, routing U.S. and South African bank loans via an Irish entity in order to avoid Zambian withholding taxes on the interest repayments. This is a transaction which has no Irish substance: there’s no investment *from* Ireland, and no interest income *staying* in Ireland. And it takes advantage of a piece of law – the Zambia-Ireland tax treaty – which was never intended to apply to transactions between Zambia and South Africa, or Zambia and the USA. Newer tax treaties, indeed, commonly contain anti-avoidance provisions (beneficial owner tests, limitations of benefits clauses) specifically to prohibit this kind of ‘treaty shopping’: individuals or companies using conduit entities in a third country to gain treaty benefits not due to residents of their own country.

    It’s not simply ActionAid’s view that this constitutes tax avoidance: the OECD itself labels the use of such conduit entities as “abuse”, and devotes a substantial part of the Commentary on the OECD Model Tax Convention on ways to stop such abuse. This Commentary, agreed by OECD Member States, is clear about what constitutes such abuse: “if a person…acts through a legal entity created in a State essentially to obtain treaty benefits that would not be available directly.”

    It’s quite right to criticise tax campaigners who fall into the trap of making up new or ideal rules, applying them to companies or individuals, and then calling them “tax dodgers”. But I’m often struck by the extent to which tax campaigners sometimes highlight arrangements, as in this case, which have already been labelled “abuse” by the international rules established by comparatively conservative bodies like the OECD. It’s our own collective amnesia that often leads us to overlook the toolkit, however limited, that already exists in these rules and laws.

    What does this mean for your anatomy of avoidance? I think it highlights that while there’s a legal gap between evasion and abuse or aggressive avoidance, there’s also a political gap between what (farily conservative) international tax standards label as abuse or avoidance, and what tax authorities will tolerate. We often argue that the problem lies with the international standards. But I think there’s also an inhibition in many countries – for fear of losing investment, or upsetting important taxpayers – to implement in national laws the toolkit against abuse that already exists in those international standards.

    • Hi Mike,

      Sorry, you’re right, I was a bit brief on your bit of it. Nothing personal 🙂

      I think the impression I get from looking at your infographic and reading the press releases is that ABF isn’t paying enough, and at first glance it’s hard to see quite why not. When one gets into the details, as with Ben and Maya’s dicussions, I think the question of why it’s not enough gets a little confused: there are issues of capital allowances and favourable tax rates in there as well as treaty shopping and transfer pricing.

      To me that reads as though “fair share” is the main end you’re driving at, and treaty shopping/TP are some of ABF’s ways of not getting there. Apologies if that’s not the impression you meant to give, but it was what I got at first.

      You’re right that some of the big questions are about the intentions of the tax authorities when entering into treaties, and obviously Mngoloa has had some second thoughts there recently. I omitted them from the bit where I mentioned your views mostly because “Parliament’s intention” already seemed like a well-established principle. I think it’d be worth going back to add it in, to point out that avoidance isn’t solely a UK issue. So I will 🙂

      • On the bit about pre-existing definitions of abuse, I think you’re right but the problem is to show that the entity exists solely to access treaty benefits. It’s much the same as the economic form question: if the entity is only in Ireland (say) to get treaty benefits then that’s what I’d classify as aggressive avoidance and the OECD would call abuse.

        But as the entity could consist of anything from a brass plate to a reasonably-sized office, and it may be hard to agree exactly what it’s doing, there’s a spectrum there. The OECD is talking about the agressive end and calls it abuse but is quieter about the middle and the other end, whereas I’m looking at the whole thing and calling it mild agression through to severely aggressive.

        So I agree that it’s avoidance, no question there. The question is how to characterise that avoidance, and then how to decide what (if anything) needs to be done about it. That’s where the political will somes in I think.

  2. It’s probably fair to add that in some countries and cases there’s also clearly a political ‘overlap’ rather than gap – i.e. activist tax authorities seeking to bend international standards to recover more tax than those (fairly conservative) international tax standards would suggest is due.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s