Following recent commentary on tax return software glitches, I wanted to reflect a little more on the implications of errors in the light of a recent judgment in the tax tribunal.
The case concerned the liability of a taxpayer for 2006-7 and 2007-8. The precise make up of his income and gains does not matter for our purposes here, because the specific tax rules which applied then are no longer in effect; but what does matter is the principle. The taxpayer’s agents used third party software based on HMRC’s specifications to produce tax computations, which were submitted with his returns.
HMRC accepted the figures in the returns but said that the computations were incorrect. This was no minor blip in the system: the difference between the tax shown by the software and the tax which HMRC said was actually due was £4.8m. This high figure was due to a very unusual combination of income gains and losses incurred by the taxpayer. In evidence, HMRC said that there were 12 other identified cases with the same problem; the total tax at stake in those other cases was £23,000.
At the tribunal the taxpayer had two arguments. The first was that the calculation produced by the software was in fact correct. The tribunal rejected that and eventually decided that HMRC’s higher figure was the right answer. The second, more important argument from our point of view, was that HMRC had no right to enquire into the calculations accompanying the taxpayer’s return. As his counsel argued, those calculations were done by HMRC itself, because they had specified the way that the software should calculate the liability. Effectively, the counsel argued, HMRC was enquiring into its own work, and there was no power for it to do so.
The judge disagreed and said that HMRC has the power to enquire into anything contained in the return, and that extended to the tax calculation. HMRC, he said, had done nothing in relation to the taxpayer’s return: all it had done was to supply technical specifications to the third-party software providers whose software happened (my emphasis) to be used by his accountants.
I don’t like the word ‘happened’ here. It suggests that it was just bad luck that the agents chose that particular software. But all such software would have given the same result because it was built to HMRC specifications.
This brings me to Making Tax Digital. We know, because HMRC has said this many times, that it will not be issuing its own software and that all users will have to rely on third party software. We also know that HMRC will be specifying how that software should operate - so similar problems may occur again.
In the end I have to reluctantly agree with the tribunal: the law is the law and problems in software design can’t be allowed to override it. But this sort of thing does leave a bad taste in the mouth. If people are forced to use third party software designed to HMRC’s specification then surely HMRC must have some responsibility for that software. I’d like to think that this was an exceptional case because of the amounts involved, and that in future, any small glitches in favour of the taxpayer are allowed to stand. But somehow I doubt it.
For more information please get in touch with Andrew Hubbard, or your usual RSM contact.