Dated: 21 November 2011

The Institute of Directors today (Monday) welcomed the report on a possible GAAR (general anti-avoidance rule) for the UK tax system, prepared by Graham Aaronson QC.

It is reassuring that the report comes out against a broad-spectrum GAAR, which would undermine certainty and make the UK less attractive to multinational businesses.

• The report recommends a well-targeted, moderate GAAR, which should only bite on tax planning that is clearly abusive. As such, most businesses will not be worried by it.
• However, the boundary between abusive and non-abusive planning will remain hazy. It must not be so hazy that people are wary of undertaking the legitimate tax planning that is needed to allow decent post-tax returns to be made in the UK. If that happened, international capital would be scared away.
• Potential secondary advantages include saving judges from having to adopt strained interpretations of existing law – the risk of which itself creates uncertainty – and the possibility of some specific anti-avoidance rules being repealed. But since hard-bitten tax inspectors are unlikely to want to give up any of their weapons, it remains to be seen whether any specific rules would be repealed.

The report expects that a moderate GAAR would not generate much uncertainty, since the Revenue would have to jump through many hurdles before using it. However, this expectation must be tested.

The IoD suggests that a dozen lawyers should each be given the same list of example tax avoidance schemes. Working independently, each would then decide which schemes would be caught by the proposed GAAR. If they do not all give much the same set of answers, then the proposed GAAR will not give certainty.

Commenting on the report, Richard Baron, Head of Taxation at the IoD, said:

“This is a very measured and sensible report on a possible GAAR. The Government should certainly use it as the basis for a formal consultation. But we must test the idea thoroughly before it gets anywhere near being put into law. If we get evidence that it would generate uncertainty, that problem will have to be fixed.”