I initially wrote this as a comment to Jolyon Maugham’s post on increased penalties for the GAAR. It got a bit long so I thought I’d just let the pingback suffice…
Whilst I think there is a certain justice in imposing additional fines on those who embark on a clearly abusive scheme, I’m not sure a distinct regime for the GAAR would be effective.
The GAAR is primarily a deterrent, often described as an electrification of the fence. It is supposedly there to catch the minority of schemes that are surviving challenge under pre-existing legislation and principles developed through case law. With the prevalence of anti-avoidance legislation and, for example, the recent resurgence of Ramsay through the firm adoption of Ribeiro PJ’s apothegm from Arrowtown, the chances of success are already low.
That low chance of success is cut down further by the threat of the GAAR and in one sense, that is all it need do to be effective: cut down the perceived chance of success.
The investment of time and resources into testing a scheme through the courts is already a significant disincentive to most, let alone the reputational damage that accompanies the public pursuit of defending artificial and abusive arrangements.
I have no problem with trying to increase the disincentive, but somebody who is currently contemplating taking on the GAAR is already highly likely to not be easily disincentivised from adopting abusive arrangements. Therefore, the main benefit in a separate penalty regime seems to be the punitive element, to my mind, rather than to increase the warnings surrounding our electric fence.
I have no issue with that rationale either but on the deterrent point I think the additional disincentives might be outweighed by the potential incentives that it creates.
Where a scheme might appear to fail by the GAAR, a harsher penalty regime creates an incentive for HMRC to use the GAAR over other methods. There will always be public and political pressure to administer increased punishments for aberrant behaviours. Will future HMRC inspectors worry about whether the PAC will ask why higher penalties were not sought by applying the GAAR?
I think it’s important to remember that the GAAR is primarily designed as a deterrent. To my mind, a deterrent that is called into action has failed in that instance. Anything that encourages systematic use of a deterrent seems rather counter-productive.
Of course the deterrent does need to actually work because it needs to be seen to be effective. I imagine HMRC will carefully pick test cases to that end. But we do not want the deterrent being used too often.
One of the GAAR’s key attributes as a deterrent is uncertainty. That needs to be preserved as much as possible. The more instances in which the GAAR is tested, the more certain we become over its scope. And the more certain we become of where its reach fails.
To push the electric fence analogy to breaking, quite literally, do we wish to encourage HMRC to publicly test our fence for weaknesses simply to sate our appetite for watching trespassers get electrocuted?
Finally, there is potentially an incentive for the taxpayer who is caught by the GAAR. Facing stiffer penalties might make them more likely to challenge the effectiveness of the GAAR in order to reduce their penalties to the level attached to other, “normal” measures.
This appears slightly contrary to my earlier point about the lack of disincentive to test the GAAR, but to resort to the electric fence analogy yet again, I’d point out that the incentive of escaping the pain of the electrocution is not the same as the disincentive given by the warning signs once your hand is on the wire.
However, if the penalties were exactly the same for being caught by any other method, what incentive would there be to struggle against the GAAR specifically?
Ultimately, if we simply want to punish more harshly abusive behaviour, I think we ought to amend the entire penalty regime so that abusive behaviour is punished more harshly irrespective of the mechanics of how the arrangements fail to operate as intended.