The General Anti-Abuse Rule is a step in the right direction, but we have a long way to go to beat tax avoidance

Posted on

It seems some people are surprised that I have had anything good to say about the proposed General Anti-Abuse Rule and the Guidance notes, announced yesterday. They miss the point. As I said to the House of Lords Economic Affairs Committee in January (page 88):

The Exchequer Secretary was surprised that I welcomed Graham Aaronson’s report in November 2011, I think. But I pointed out that if you are campaigning  for street lighting for a whole street, if you at least get the first light you will welcome it, even if you are a long way from achieving your whole objective. So of course I welcomed this but it is a long way from being the lighting for the whole street that I would like.

So let me step back and say what the problems with this GAAR are, even though I would rather have it than not. Much of this I have already said in evidence I presented to the same House of Lords Committee in January. My objections to the GAAR made then, and still, are:

The General Anti-Abuse Rule as proposed for inclusion in the Finance Bill 2013 suffers from three fundamental problems:

1) It is not general in nature;

2) It does not tackle tax avoidance as that term is now widely understood, including by the Prime Minister[i];

3) Its construction makes it unlikely to succeed in its objectives.

Those are big claims, but I think they are right despite the work I have done on the Guidance.

The first point is obviously true: the General Anti-Abuse Rule only tackles the most egregious of schemes and the Guidance notes are littered with references to HMRC having the right to tackle tax avoidance falling outside the scope of the GAAR with other measures available to them. By definition that means it is not general in nature.

The second point is also obviously true. The GAAR does not tackle Google or Amazon. It fail the Prime Minister's own test of acceptability, by design. On that basis it is flawed from the outset.

The third point hinges on the "double reasonableness" test inherent in the GAAR, with which I have major problems. As the new Guidance notes say, to apply the GAAR HMRC has to show that the arrangements:

“cannot reasonably be regarded as a reasonable course of action”. This recognises that there are some arrangements which some people would regard as a reasonable course of action while others would not. The ‘double reasonableness’ test sets a high threshold by asking whether it would be reasonable to hold the view that the arrangement was a reasonable course of action. The arrangement falls to be treated as abusive only if it would not be reasonable to hold such a view.

This appears so, well, reasonable, but it isn't. That is because the test is not one of fact; the test is whether it is reasonable to think in the light of current prevailing thinking that the activities undertaken abuse tax law. The question to then ask is who might hold reasonable opinion on this matter. Two tests apply. One is that of a judge. It may be hoped that they might reflect the opinion of the person on the Clapham omnibus, but I do not hold out a lot of hope there. More importantly though  it is unlikely that any case will reach the point of being considered by a judge unless it has been considered unreasonable by the GAAR Advisory Panel. That panel will be drawn almost exclusively from the  tax profession and those who work in it for the companies likely to enter into tax arrangements: outside HMRC they are the only people likely to have an informed opinion on these issues, and HMRC is barred from the panel. And since tax avoidance of the sort the GAAR tackles only happens because the prevailing sentiment of many in the tax profession and in commercial enterprises both encourages and permits it this test is inherently biased in favour of tax abuse. That is my fundamental concern with this GAAR. What is needed in place of this is an objective economic test. I have proposed one in the General Anti-Tax Avoidance Principle Bill Michael Meacher put to the House of Commons. I continue to think it better than the General Anti-Abuse Rule.
There are others problems too. The first is a lack of a clear penalty regime for the GAAR. The second is that burden of proof that an arrangement is abusive falls on HMRC. This is contrary to all tax logic where the burden of proof falls on the taxpayer.
In summary what is needed to put the GAAR right is that:

1. HMRC should be able to commence GAAR action on its own initiative without the advice of a panel;

2. The deeply subjective double reasonableness test should be replaced with an objective economic substance test which would assess whether tax was paid in the right place at the right rate under the right tax code at the right time given the economic impact of the transactions that actually occurred;

3. The burden of proof should be on the taxpayer to show that this has happened;

4. Penalties should be imposed on those who sought to avoid tax in a way that falls foul of this test to deter those seeking to do so;

5. A clearance system should be provided so that any taxpayer could obtain a prior agreement to their proposed arrangements from HMRC, albeit at a price for the certainty that could supply them with.

The General Anti-Abuse Rule is a step in the right direction. But we have a long way to go to get this right as yet.