Measuring Tax Gaps 2017 edition has been published and estimates the tax gap for 2015-16 to be in the range of £34bn, representing 7% of all tax liabilities. If you enjoy statistics, this is the report for you.
In summary, the tax gap is attributed as follows:
By customer group
By type of tax
£13.7bn IT, NICs and CGT
£6.1bn Failure to take reasonable care
£9.7bn Large businesses
£12.6bn Value Added Tax
£6.0bn Legal interpretation
£3.3bn Corporation Tax
£3.0bn Excise duties
£3.5bn Hidden economy
£1.5bn Other taxes
The ‘tax gap’ is the difference between the amount of tax that should, in theory, be collected by HMRC, and what is actually collected. It provides a useful tool for understanding the relative size and nature of non-compliance. The causes range from simple mistakes and errors to deliberate criminal evasion.
It is interesting that the tax gap attributable to legal interpretation is getting slightly worse year on year. Legal interpretation losses arise where the taxpayer’s interpretation of the law and how it applies to the facts in a particular case and HMRC’s interpretation, results in a different tax outcome. Examples include the correct categorisation of an asset for allowances, the allocation of profits within a group of companies, or VAT liability of a particular supply. An indication of the increasing complexity of UK tax law perhaps?
The report notes that the corporation tax percentage tax gap in 2015-16 is at its lowest level and there has been a steady downward trend for the avoidance tax gap since 2012-13, to £1.7 billion in 2015-16.
For the tax gap relative to amount of tax collected, by type of tax, the prize for the worst culprit is awarded to VAT, where the VAT gap of £12.6 billion, equates to 9.8% of VAT theoretical tax liabilities. Perhaps inevitably, tobacco duty remains a particular problem area and the tax gap relative to duty thought to be due is, at 16.7%, at its highest rate for the ten year period under review.
On a brighter note, fewer people are operating in the hidden economy than were estimated previously and their average incomes are lower.