HMRC’s estimate of the ‘tax gap’ is chiefly valuable in preventing credence from being given to the inflated estimates of tax campaigners. However, there are good reasons to think that the department’s calculation is significantly overestimated, and it is questionable whether HMRC should be responsible for calculating a figure which it has a clear interest in overestimating.
On 16 October 2014, HMRC published its annual calculation of the ‘tax gap’ in a paper entitled 'Measuring tax gaps 2014 edition: Tax gap estimates for 2012/13’ (all references below are to this paper, unless otherwise stated). As always, the headline numbers have elicited substantial comment in the tax technical and financial press. As is usual, however, the detail of the paper reveals a very different tale to those told by HMRC, by the press and by tax campaigners.
The introduction to the paper explains that: ‘The “tax gap” is the difference between the amount of tax that should, in theory, be collected by HMRC, against what is actually collected.’ (page 3)
It is only when one arrives at page 16 of the paper that it reveals that HMRC defines ‘the amount of tax that should, in theory, be collected by HMRC’ (the ‘theoretical liability’) as: ‘the tax that would be paid if all individuals and companies complied with both the letter of the law and our interpretation of Parliament’s intention in setting law (referred to as the spirit of the law)’ (emphasis added).
So the theoretical liability is not the amount which would be collected if everybody paid what the law required, but that amount and the additional amount that they would pay if they agreed with HMRC as to what the ‘spirit of the law’ required.
It will also be seen that the spirit of the law only works to increase the theoretical liability. Where the ‘letter of the law’ imposes a liability which is not in accordance with its spirit, as the First-tier Tribunal seemed to accept that it did in Lobler v HMRC [2013] UKFTT 141 (TC), the liability will form part of the theoretical liability.
What is more, it appears that HMRC assumes that taxpayers should voluntarily pay amounts which are not due under the law at all, whether or not payment would be required by compliance with HMRC’s view of the ‘spirit of the law’. £4.5bn of the tax gap is due to ‘legal interpretation’ (pages 4 and 13). This, the paper explains, is: ‘the potential tax loss from cases where HMRC and individuals or businesses have different views of how, or whether, the law applies to specific and often complex transactions. 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 item.’ (page 15)
Either there has been a loss or there might be one in the future. How can an element of a ‘loss’ be a potential loss? There can only have been a ‘loss’ due to legal interpretation where, after argument, HMRC has concluded that the taxpayer’s construction is correct, or where the taxpayer has been successful in the tribunals or the courts. So it appears that in the tax gap, HMRC has either included tax which it has not actually lost but might lose in the future; or that ‘the amount of tax that should, in theory, be collected by HMRC’ includes tax which it would collect on an interpretation of the law that it has adopted, but which is incorrect.
It seems, therefore, that the tax gap is the result of a mixture of reality, of the tax actually collected, and of fantasy – the amount which HMRC thinks that it should have been able to collect, if everybody paid tax in line with its view of what the law ought to be, including where that view does not accord with what the law actually is.
This is not the only area where the paper’s analysis is, to put it mildly, opaque.
Analysis by ‘behaviour’
For example, in analysing the tax gap by ‘behaviour’, it has one category which is described as ‘criminal attacks’ (page 13). This category is not actually defined but, in respect of it, the paper says:
‘Organised criminal gangs undertake coordinated and systematic attacks on the tax system. This includes smuggling goods such as alcohol or tobacco, VAT repayment fraud and VAT missing trader intra-Community (MTIC) fraud.’ (page 15)
It would seem that ‘criminal attacks’, on a literal reading, would not include attacks by individuals, attacks by criminal gangs which were uncoordinated even if systematic, or attacks which were unsystematic even if coordinated. HMRC, it appears, is only interested in the better class of criminal.
The analysis by ‘behaviour’ also includes the category ‘evasion’, and a further category, the ‘hidden economy’ (page 13). Such categories do not appear, at first, to be mutually exclusive. If undertaken by an organised criminal gang, smuggling goods, for example, would seem to fall into the category of ‘criminal attacks’; however, it involves, at least in ordinary usage, evasion and it is part of the hidden economy. In a telephone conversation with a very helpful HMRC analyst, I was assured that HMRC takes considerable efforts to ensure that there is no double counting.
When one looks a little further one finds that evasion is defined as: ‘illegal activity, where registered individuals or businesses deliberately omit, conceal or misrepresent information so they can reduce their tax liabilities’ (page 15).
It is not entirely clear what the term ‘registered’ means to HMRC here. It does not seem to be defined in the paper, but one presumes that it means something like being recorded in HMRC’s systems in some form, rather than bearing a more restricted meaning, such as a person being registered for VAT or as an employer or employee for PAYE purposes. HMRC’s analyst confirmed to me in our telephone conversation that is indeed the case.
According to HMRC, the hidden economy is defined as: ‘Undeclared economic activity that involves what we call “ghosts” – whose entire income is unknown to HMRC; and “moonlighters” – who are known to us in relation to part of their income, but have other sources of income that HMRC does not know about’ (page 15).
By these definitions, evasion would not seem to include the tax lost by a failure to declare a taxable amount by an individual who has managed to avoid generating any sort of record on HMRC’s systems, however much tax may be due on his undeclared activities. It seems, however, that this would fall within the ‘hidden economy’; by HMRC’s definition, this includes both a criminal evader in the ordinary meaning of the phrase, and those who omit a source of income from their returns in simple ignorance that it is taxable. The unfortunate Mr Lobler – who found, to his surprise, in Lobler v HMRC, that he had lost most of his life savings in income tax on a chargeable event gain, in spite of having made no economic profit at all on the policies concerned – would, by HMRC’s definition, fall into this category as a ‘moonlighter’. HMRC’s analyst, however, said that generally tax lost through innocent omissions would be recorded under the category of ‘error’, rather than that of ‘hidden economy’.
Analysis by customer group
The paper also analyses the tax gap by ‘customer group’. The groups are ‘large businesses’, ‘SMEs’, ‘individuals’ and ‘criminals’ (page 12). If criminals are a separate category from the other categories of ‘customers’, it appears that they must all be persons who are not individuals and who do not undertake any business – a rarefied group indeed. In any event, although we have grown used to HMRC regarding its ‘customers’ as criminals, it is rather alarming to find that they regard criminals as their ‘customers’. Perhaps it is because, as we have seen, they only take cognizance of the better class of criminal.
It is clear from the paper that, however odd its definitions and categories may be, the tax gap remains of relatively small significance to the UK economy. At £34bn, it amounts to just 6.8% of the theoretical liability for 2012/13 (page 3), 5% of forecast government expenditure, and about 2.2% of UK GDP for that year. Even if the whole of the tax gap were to be eliminated, it would not solve the problem of the UK’s budget deficit, which was £80bn in 2012/13 alone.
Further, when one looks at the paper’s analysis of the tax gap by behaviour, it is clear that avoidance is of small significance to the tax gap itself. Tax ‘lost’ through tax avoidance (however one may understand loss in relation to an amount which is not legally due) amounts to just 9% of the tax gap or just £3.1bn (pages 3 and 13).
Interestingly, the paper reveals that in the previous year’s report, the avoidance element of the tax gap had been significantly overestimated. In the 2013 paper, the avoidance tax gap for 2011/12 had been estimated to be £4bn; but it is now estimated to have been only £3.4bn in that year (page 14). It would be unsurprising if the figure for 2012/13 were similarly revised.
Illegal activities
45% of the tax gap, being £15.4bn, is accounted for by what appear to be illegal activities: criminal attacks, evasion and the hidden economy (although, as we have seen, the paper’s definition of the hidden economy, read literally, would seem to include some innocent error) (page 13).
Taxpayer error
20% of the tax gap, or £7.1bn, is accounted for by simple taxpayer error (page 13). 40% of this part of the tax gap is in respect of errors where the taxpayer has taken reasonable care. If errors arise where a taxpayer has taken reasonable care, one might think that the blame lies with the design of the system, rather than with the taxpayer who struggles to comply with it.
Failure to collect tax: 13% of the tax gap, being £4.4bn, arises from ‘non-payment’ (page 13). One would think that this must relate to situations where the fact that the tax is legally due has either been unchallenged or has been established after challenge; otherwise, it would fall also into the other categories of ‘behaviour’. So one might think that this must relate primarily to situations where the taxpayer is insolvent and to situations where HMRC has failed to collect the tax due through incompetence.
The paper does indeed say that it arises mainly as a result of insolvency (it does not admit to incompetence) and ‘does not include debts that are eventually paid’ (page 15). This, however, is subject to the exception that VAT non-payment ‘is based on the difference between new debts arising and debt payments’ (page 15). Later in the paper, HMRC says that: ‘The contribution of debt to the VAT gap is defined as the amount of VAT declared by businesses but not yet paid to HMRC’ (page 25).
It is difficult to believe that this can mean what it appears to say; that VAT non-payment merely measures VAT which happens to be outstanding at the year end, whether or not it is eventually paid; and that this figure has been aggregated with a figure for other taxes, which represents the tax actually due but that HMRC fails to collect at all. What significance can be attributed to an aggregate of two figures measuring such different things?
Inclusion of VAT debt
If that is the case, this peculiar method of computing the non-payment element of the tax gap must result in its being overestimated. It is not the only cause of the tax gap being overestimated which is revealed in the paper.
Legal interpretation
As we have seen, the inclusion of the category ‘legal interpretation’ also indicates that the tax gap is overestimated.
Time lag
Another factor likely to result in overestimation is that much of the tax gap must, perforce, be calculated not by reference to experience in respect of 2012/13, but in respect of prior years. For example, the income tax, NIC and CGT components of the tax gap are mainly based on information relating to periods before 2012/13, and which are simply extrapolated (page 46). Therefore, in effect, the tax gap for 2012/13 does not fully reflect what happened in that year, but rather what happened in prior years. If, as there is good cause to think is the case, the increased resources given to HMRC’s compliance activities and the increased aggression with which they are pursued are effective, then this lag in the data will have the effect of overestimating the tax gap.
Taxpayer overpayment
As the Common’s Treasury Select Committee identified in 2012 (see below), one other very obvious reason why HMRC’s tax gap figure is likely to be an overestimate is that it contains no estimate of the amounts which are collected which are not due under the law. This is where taxpayers make errors in HMRC’s favour of which it is unaware or which it makes no effort to correct, and where tax is collected due to HMRC’s errors. This includes errors in respect of PAYE codings, which are not discovered or not resisted by the taxpayers concerned; and where taxpayers concede on assessments, because the costs and risks of resisting them through the courts are disproportionate to the tax at stake.
As we have seen, taxpayer error, whether or not due to the taxpayer’s failure to take reasonable care, accounts for £7.1bn of the tax gap. In contrast to non-payment, evasion and avoidance, which are deliberate, errors are by definition inadvertent. That being the case, one would expect that they are as likely to result in overpayments as in underpayments. I can find nothing in the paper suggesting that HMRC attempts to identify those taxpayers’ errors which are in its favour, or to take account of the possibility of such errors in its estimates. Nor can I find any attempt to identify overcollection through taxpayers acquiescing in assessments, not because they are due under the law but because of the risks and costs of resisting them. HMRC’s analyst confirmed that, although in some components of the tax gap overpayments have been netted off against underpayments, there is no attempt generally to estimate the amount of overpayment by taxpayers.
Any serious attempt to measure the tax gap would have to include a reduction to take account of overcollection. That suggests that, through this cause alone, the tax gap may be overestimated by an amount which may exceed £7.1bn.
Illustrative tax gaps: Another potential source of overestimation in the paper’s estimates is the reliance placed on what it calls ‘illustrative’ tax gaps. The paper explains that: ‘Where robust methodologies have not yet been developed, an illustrative tax gap estimate is given based on our operational experts’ opinion or calculated by selecting the nearest equivalent measured gap’ (page 52).
An example of the latter – ‘selecting the nearest equivalent measured gap’ – is that the tax lost through incorrect self-assessment is estimated from the employer compliance random enquiry programme (ECREP). Large employers are not subject to the ECREP, so the authors of the paper simply assume that large employers’ returns will have a similar level of errors to those estimated for small and medium-sized employers (page 49). That seems a strange assumption to make when one considers the great differences in the sophistication of accounting systems and resources available to large employers, and the greater harm which reputational damage can do to them.
The former method of arriving at an illustrative tax gap – basing it upon ‘our own operational experts’ opinion’ – appears to mean no more than asking an HMRC employee to guess a figure. Giving all due credit to such employees’ efforts to be objective, this approach must be at risk of overestimation due to institutional bias; the natural tendency of the members of an institution to see the world through the prism of the institution’s interests.
The fact that the paper ignores these obvious risks of overestimation might be thought to call into question the propriety of HMRC having control of the production of the tax gap estimates. Although the paper does cite a favourable review of ‘HMRC’s tax gap analysis program [sic] by the International Monetary Fund’, the paper is not, I understand, specifically subject to review by the Office for National Statistics or the Office for Budget Responsibility, although some of the subsidiary totals used in compiling the estimates are. HMRC, like all government departments, must justify the allocation of resources it receives. The tax gap has provided a convenient argument that its expenditure should not be restrained. It is surely time for the oversight of tax estimation to be the responsibility of a body that is independent of both the Treasury and of HMRC.
Even if one takes no account of those factors which suggest that HMRC’s calculation of the tax gap is an overestimate, HMRC’s figure of £34bn is, as in previous years, vastly less than the wilder estimates quoted in the popular press. One tax campaigner recently produced for a trade union an estimate of the tax gap of £122bn, which received wide publicity.
What is more, the trend is for the tax gap to reduce. It is true that this year, for the first time, the figure for the tax gap has increased over last year both absolutely (from £33bn to £34bn) and as a percentage of the theoretical liability (from 6.6% to 6.8%). However, it seems that this level of increase is unlikely to indicate a change in the overall trend. There are several reasons for this.
First, the paper reveals considerable revisions in the estimates for every one of the prior years for which the tax gap has been calculated (page 18). The figure for 2011/12 was originally estimated as £35bn, for example, but was reduced in this year’s paper to £33bn (page 6). A similar level of revision to this year’s estimate would mean that the tax gap would have reduced from last year, rather than increased.
Secondly, the paper acknowledges that the calculation is the result of very broad estimates. It says: ‘These are our best estimates based on the information available, but there are many sources of uncertainty and potential error’ (page 3).
Finally, the tax gap is rounded to the nearest £1bn. This allows for a considerable degree of convergences between the figures of £33bn for 2011/12 and £34bn for 2012/13.
Taking all of these factors into account, it is unlikely that this year’s year-on-year increase is statistically significant in isolation. Over the long term, there has been a significant reduction in the tax gap; since it was first calculated in 2005/06, it has reduced by £3bn (or 8% in absolute terms), and from 8.5% of the theoretical liability to 6.8% (page 8). The paper says that the tax gap is £8bn ‘lower than it would have been if the percentage tax gap had remained at the 2005/06 level of 8.5%’ (page 6).
Is the tax gap worth calculating at all? The Treasury Select Committee in its report entitled Closing the tax gap: HMRC’s record at ensuring tax compliance, published on 6 March 2012, thought not. It recorded that:
‘HMRC’s calculation of the tax gap includes the tax it judges has been lost owing to a number of different behaviours. These are: error, criminal attacks, evasion, operating in the hidden economy, avoidance, legal interpretation, non-payment, and failure to take reasonable care. It is unhelpful to aggregate these different behaviours ... A large loss of tax arising from the hidden economy requires a quite different kind of intervention to a difference in opinion between HMRC and a taxpayer on the tax due under an ambiguous law.’
The Committee’s overall conclusion on HMRC’s annual calculation of the tax gap was that:
‘[HMRC] risks focusing its employees’ attention on the wrong task. HMRC should not be aiming to collect more tax at any cost, but should be ensuring that all taxpayers pay the correct amount of tax ... For HMRC to collect the right amount of tax, it would also need to address the many people who pay more tax than they need to by law…
‘The tax gap can be a useful concept for assessing trends in the amount of possible unpaid tax. We are not, however, convinced that the process of calculating, publishing and publicising an aggregate figure for the tax gap is a sensible use of HMRC’s limited resources. The aggregate tax gap figure is misleading and risks focusing HMRC on the wrong task as it only provides an order of magnitude.
‘[I]t would be more useful for it to identify ambiguities in tax law rather than employ resources in calculating how much tax would be collected if everyone shared its interpretation of the law.’
The tax gap has become part of the annual fiscal calendar, however; and if HMRC were not to calculate it, the fanciful estimates cited in the popular press would acquire a spurious authority. If only for this reason, however unfortunate the decision to begin to publish these annual estimates may have been, it would not be helpful to cease to do so now.
HMRC’s estimate of the tax gap is chiefly valuable in preventing credence from being given to the inflated estimates of ‘tax campaigners’. A detailed analysis of the paper would seem to suggest that there are good reasons to think that: HMRC’s calculation of the tax gap is significantly overestimated; part of the tax gap is due to the fact that HMRC wishes to collect more tax than the law allows; some part of it is caused by taxpayer error in situations where the taxpayer takes all reasonable care but is defeated by the complexity of the system; avoidance forms only a relatively insignificant portion of the whole; and the most significant single constituent part is the result of criminal behaviour.
The lessons one might draw from this are that HMRC: should not be left in control of calculating a figure which it has a clear interest in overestimating and which is so politically charged; should refrain from inflaming public anger about tax avoidance, which is an insignificant part of the tax gap; should try harder at drafting tax law accurately and at simplifying the tax system so as to reduce taxpayer error; and should accept that the law is not determined by what HMRC thinks it ought to be and should concentrate on its proper task of detecting and deterring the serious crime of tax evasion.
