Blog: Non-dom stats: important in what they don’t tell us as in what they do

HMRC have released statistics for the first time showing the breakdown of the tax paid by those not domiciled in the UK (‘non-doms’). The statistics purport to show that in 2014/15 there were 121,300 non-doms; 85,400 of whom were UK resident; 54,600 of whom used the remittance basis and only 5,100 of whom paid the remittance charge of £30,000 – £40,000.

Some things are  clear in the report:

That the non-dom population contributed £9.25bn in income tax, CGT and national insurance contributions in 2014/15
Each UK resident non-dom paid an average of £105,000 in income tax, CGT and national insurance.
The average income tax paid by UK resident non-doms was £76,500; the average income tax paid by taxpayers as a whole was £5,430
UK resident non-doms paid 3.9 per cent of all income tax (£167 billon) yet represented only 0.2 per cent of the taxpaying population (30.7 million)

Yet the statistics are more telling in what they omit to say.

Data is welcome

The first-ever publication by HMRC of detailed data on the UK’s non-dom population and the tax paid by them is to be welcomed – not least because evidence-based policy-making has been in short supply in this area until now. Ever since 2007, successive governments have targeted non-doms with scant regard either for the size of the population, the tax already paid by them or the overall contribution that existing non-doms and potential new arrivals might make to the economy as a whole.

Non-doms have been further vilified in the media. Abolishing non-dom status was apparently one of Labour's most popular policies during the 2015 General Election.
In such a politicised debate, the lack of evidence has been worrying.

Back in 2011 the-then Treasury minister David Gauke informed the House of Commons that government records indicated that there were 7,400 non-doms using the remittance basis in 2011/12 and 9,600 in 2012/13. It now seems that the true figures were 5,500 and 5,100 respectively. The publication of figures – which we are promised will now be an annual event – is a welcome start in rectifying this evidence gap. Yet, despite 35 pages of commentary accompanying the release of the data, the evidence is still sparse and incomplete.

Although the data is a helpful starting point, it is arguably far more revealing in what it omits or fails to say.

… But far from complete

We still do not actually know how many non-doms there are. The figures merely give data for those non-doms who file self-assessment returns. We do not know whether these are the same individuals each year. We do not have figures for non-doms who do not (or have no need to) file a tax return. And we do not have figures for those who could claim non-dom status, but choose not to do so and give no indication of this on their tax return.

On the basis that only around a third of taxpayers (and only around 18 per cent of the total population) file self-assessment returns, this suggests that the total number of non-doms may be three to five times as high as HMRC’s figures. Add in those who could have claimed non-dom status, but gave no indication that they had chosen not to and the figure could be higher still.  Previous guestimates of up to five million non-doms are probably wide of the mark, but the total non-dom population may well be somewhere between 500,000 and a million people.

The figures for non UK-resident non-doms (a further 35,900 individuals) are confusing. After all, everyone in the rest of the world without a UK father or grandfather – so close to seven billion people – is presumably in this category.  One can only presume, therefore, that the figure is for those outside the UK who nonetheless pay UK tax (e.g. on UK employment duties or capital gains on residential property). Aggregating these figures with the resident non-dom population, however, does not give any particular insights and will, in my view, tend to confuse policy.

Implicit in the figures, but not drawn out, is the possible Laffer curve effect of the changes in 2007/08.  Statistical analysis of such tax elasticity is notoriously difficult, so I will leave better analysis to those who know more about what they are doing. However, at face value, total receipts from non-doms fell from £8.4 billion in 2007/08 (prior to the reforms) to £7.2 billion in 2008/09 (after the reforms) – a loss of £1.2 billion. Some of this may be due to forestalling, but the last-minute nature of the 2008 reforms suggest that this is unlikely to be significant – and yields below 2007/08 levels continued for at least the next two years – despite rising overall tax receipts during those years.

A similar reduction in yield occurred when the £30,000 remittance basis charge was increased to £50,000 in April 2012.  Income tax receipts from the remittance basis charge drop £41 million and total tax drops £59 million after the change.

Finally, as is sadly all too normal, the statistics omit to look back at how much tax the changes in 2008 (and 2012) were supposed to bring in. One has to dig back to find these – see p112 in link here – but one discovers that the 2008 reforms were supposed to yield nothing in 2008/09, £700 million in 2009/10 and £500 million in 2010/11. It is not clear whether this was just wrong or can somehow be reconciled with the outturn that the yield fell by £1.2 billion in the first year; and remained £400 million down in the second year; and £300 million in the third year. 

The 2012 changes – note p52 of link here- were similarly supposed to be neutral and then raise money, but revenues subsequently fell.

It is to be hoped that the release of these statistics might cause government to pause – at least once the 2017 legislation has been enacted – – before introducing yet more reforms to non-dom taxation.  Tax policy is better made when it is backed up by evidence and, particularly, when governments listen to stakeholders (who warned at the time of these effects).

By John Barnett. John is a Council member of the Chartered Institute of Taxation