The third Wiseman

One of my earliest tax memories was the announcement in the 1988 Budget concerning the proposed introduction of ‘independent taxation’ of married women, to come into effect from 6 April 1990. More than three decades later, it probably seems inconceivable to many advisers that married couples had to combine their income for tax purposes (and, perhaps, even more so, that the wife was far from an equal partner in the relationship so far as the former Inland Revenue was concerned).

As anachronistic as it might sound, it should be recalled that dependent taxation is not completely confined to history. Those who advise on social security benefits will be aware that household income remains an important feature.

I shall not consider the respective merits of the two approaches, leaving that for politicians to resolve. However, from a practical perspective, they tend not to clash because, even to the extent that benefits are taxable, it is clear whose income they represent and, so, the tax system can properly cope.

The high income child benefit charge

The one exception to this, however, is child benefit. Typically, this is paid to mothers (rather than fathers) and is (strictly speaking) a tax-exempt benefit. However, the 2010-15 coalition government wanted to make the benefit means-tested.

It was recognised that this would be politically difficult so the government then opted for a mechanism whereby the benefit would be clawed back from high income families. Leaving aside the potential political difficulties of imposing a tax charge on a predominantly female subset of the population, it was also acknowledged that such a move would not be particularly efficient. This was because, in the majority of qualifying couples, the benefit recipient was not the partner with the higher income, thereby reducing the effectiveness of such a measure, if based solely on the recipient’s income.

To overcome these political hurdles, the high income child benefit charge (HICBC) was introduced. It is imposed on the recipient of the child benefit unless the recipient’s partner has a higher level of income in the year, in which case it is the partner who is liable. The HICBC is, as its name implies, a standalone tax charge. In other words, it is not determined by applying tax rates to an amount of income. Instead, the charge is calculated as a percentage (which can range from 0% to 100%) of the total child benefit received by the taxpayer (or the taxpayer’s partner), the applicable percentage being determined by reference to where the taxpayer’s adjusted net income (as defined in the Income Tax Act 2007 s 58) falls in the £50,000 to £60,000 range (with the 100% charge applicable whenever the income exceeds £60,000).

It is far from a fair imposition. Even amongst couples where one partner receives child benefit, some couples with a combined income of £100,000 can escape the charge, whereas others with income totalling £60,000 can be required to repay the benefit in full. It can lead to infinite marginal tax rates (putting into shade the politically-sensitive 60% rate suffered when income exceeds £100,000) and also risks being non-compliant with human rights law.

Furthermore, there have been a large number of tribunal cases in which taxpayers have contested penalty assessments for failure to disclose their liability to the charge. This is because of the fact that, on top of the counterintuitive aspects of the charge, Parliament has provided that HICBC liability is an exception to the rule that PAYE taxpayers do not generally have to worry about notifying HMRC about their potential exposure to income tax.

In other words, HICBC joins the increasing list of areas where an obscure (and far from obvious) tax obligation generates a steady stream of penalty income for the exchequer. As was noted in one case, this failure to notify HMRC arises in a context where HMRC actually knows all the relevant information in any event but still expects taxpayers to join the dots for it.

However, this article focuses on a further practical issue that has arisen. How does HMRC assess taxpayers when it decides that the HICBC needs to be imposed? This has been the subject of several conflicting cases in recent months, the two main ones being Wilkes [2020] UKFTT 256 (TC) and Haslam [2020] UKFTT 304 (TC).

In Wilkes, the tribunal noted the various situations where HMRC has powers to issue discovery assessments. Section 29(1)(b) of the Taxes Management Act (TMA) 1970 is applicable in cases where an assessment has been made but is insufficient. That is inapplicable, however, in cases where there has been no self-assessment. Whilst that might lead to an assessment under s 29(1)(a), that paragraph is limited to cases where income has not been assessed but should have been. Because the HICBC is a standalone charge and not an imposition of tax on specific income, the tribunal concluded that the discovery assessment powers were inapposite.

In Haslam, the tribunal agreed with much of the analysis in Wilkes but then decided that Parliament must surely have intended discovery assessments to be available to HMRC in such cases. It proceeded to give the legislation a ‘rectifying construction’ so as to read into s 29(1)(a) words so as to gloss over the fact that the HICBC does not actually represent income.

This article considers a third case (Wiseman [2020] UKFTT 383 (TC)), where it was decided that both Wilkes and Haslam were wrongly decided.

The facts of the case

Mr Wiseman was previously self-employed and submitted tax returns. However, from about 2004, he has been subject to PAYE and has not been required to submit tax returns. This remained the case for the 2013/14 and subsequent tax years.

Mr Wiseman had received a notification from HMRC issued shortly before the implementation of the HICBC that alerted readers to their possible exposure to HICBC if their income exceeded £50,000. In 2013/14, Mr Wiseman’s salary was just below £49,100 and, accordingly, he assumed that the HICBC was of no relevance to him. However, he did not realise that the HMRC notification had been written by someone with more than basic tax knowledge and that, when it referred to ‘income’, it meant taxable income which, in Mr Wiseman’s case, included taxable benefits in kind. Because of a car and medical benefit, Mr Wiseman’s taxable income exceeded £50,000 and he was thus liable to the HICBC. (Mr Wiseman’s pension contributions were insufficient to take the figure back below £50,000.)

Mr Wiseman’s partner had been in receipt of child benefit since 2004 and this continued into the tax years which were the subject of the appeal.

Following an awareness campaign by HMRC in 2018, Mr Wiseman found out that he was indeed liable for the HICBC and advised HMRC accordingly. A few weeks later, HMRC issued the discovery assessments now under appeal (for the 2013/14 to 2016/17 tax years). Mr Wiseman appealed against these assessments, predominantly in relation to concerns about the process. However, by the time that the case reached the tribunal, the attention of the tribunal was focused on the validity of the discovery assessments.

The tribunal’s decision

The case came before Judge Rupert Jones and Ian Menzies-Conacher. They set out the principles deriving from the main cases concerning discovery assessments and the key parts of the main cases that discussed the interaction of the HICBC and discovery rules.

The hearing took place a few days after the Haslam case was decided, but it is unclear whether the parties at the hearing were aware of that decision at the time. However, the tribunal invited further written submissions to address the issue, which were received in early August 2020.

The tribunal noted the various remedies available to HMRC in cases where taxpayers had not paid enough tax. For taxpayers within the Self Assessment system, HMRC could issue determinations (in the absence of any tax return), enquire into the return, or issue a discovery assessment under s 29(1)(b). However, for taxpayers outside Self Assessment, HMRC would normally have a choice: to issue a discovery assessment under s 29(1)(a); or to bring the taxpayer within the Self Assessment regime by issuing a notice to file under TMA 1970 s 8.

If one follows the approach taken in Wilkes (and to some extent adopted in Haslam), however, the s 29(1)(a) option is not available in HICBC cases, leaving just the possibility of issuing a s 8 notice. However, TMA 1970 s 34A imposes a four-year time limit on self-assessments, meaning that this remedy will not be of any practicable assistance in many cases.

Recognising a potential anomaly here, the tribunal then looked at the wording of s 29(1)(a) and, in particular, the word ‘income’ therein. It considered that the word covered the totality of the taxpayer’s income and not merely income from a specific source. In Mr Wiseman’s case, this was his net employment income. The tribunal continued to consider that that (employment) income should have been subject to income tax, as calculated by the various steps in the Income Tax Act 2007 s 23, which incorporate the HICBC at step 7. As that (employment) income was not so subjected to tax, the tribunal considered that the condition in

s 29(1)(a) was in fact met. (In Wilkes, the tribunal had considered that the phrase ‘income that ought to have been assessed to income tax’ covered the first six steps but not the seventh step because that related to standalone tax charges.)

As a result of its own reading of the legislation, the tribunal concluded that the discovery assessments were valid and that it did not have to consider whether a rectifying construction was appropriate.

Commentary

The tribunal’s approach might provide the necessary breakthrough for these cases. It will be interesting to see how the Upper Tribunal responds when HMRC’s appeal in the Wilkes case is heard next year, as it is likely to look carefully at this decision.

Although it gives rise to a neat solution, there are a number of arguments that militate against the tribunal’s approach in this case.

One part of the tribunal’s reasoning was that, even though step 7 brings in standalone charges, it is still a part of the assessment of the taxpayer’s income, because the amount added to the liability at step 7 is dependent on the amount of that income.

Whilst that approach is superficially attractive, it does give rise to some difficulties. First of all, it is an oversimplification to say that the calculation of the HICBC depends on the amount of the taxpayer’s income. More strictly, it depends on the amount of a statutory term, known as the taxpayer’s ‘adjusted net income’. Although broadly equivalent to what many taxpayers might consider to be their ‘income’, it is not income per se. In much the same way, Mr Wiseman was understandably misled by the HMRC guidance that told him he could ignore the HICBC because he read ‘income’ as amounting to his salary.

The second reason is related to the first. The condition in s 29(1)(a) is that income that ought to have been assessed has not been assessed. However, income itself is widely regarded to amount to separate components of income, which themselves are derived from specific heads of charge (for example, employment income and bank interest). Indeed, the wording of the steps in s 23 reflects this meaning of income. First, step 1 refers expressly to the different ‘amounts of income’ that the taxpayer might have. Secondly, these components are never actually amalgamated and are subject to tax at (potentially) different rates, according to their income type.

This more conventional meaning of income allows s 29(1)(a) to be used if one (or more) such sources of income has not been fully assessed to tax. This is reflected by the use of the word ‘any’ in the phrase ‘any income’ in s 29(1)(a). The tribunal’s approach glosses over the word ‘any’ and treats income as if it reflected an amorphous whole.

Thirdly, there are some standalone charges (for example, unauthorised payment charges in relation to pension schemes) which are taken into account at step 7 but which are not income dependent. The tribunal’s logic would not apply to those charges, leaving an uncomfortable mismatch. In this regard, it is noteworthy that the pensions rules in Finance Act 2004 expressly provide for a modification of s 29 to address such concerns. It is my view that Parliament should have made a similar provision in the Finance Act 2012 when the HICBC was introduced.

It should also be remembered that the problem that HMRC complained about – being the inability to take remedial action after four years because of the time limits in s 34A – is rather overstated. In cases involving HICBC, is it really reasonable for HMRC to want to go back further than four years? After all, it does have all the information at its own disposal and in some cases should know better than the taxpayers whether or not the HICBC is due. (In the present case, the four year time limit would have taken out only one of the four years being targeted by HMRC. And if HMRC had taken action earlier, that year would not have been a problem either.) Furthermore, the time limit in s 34A was introduced by Finance Act 2016 and should not be a relevant factor in determining how to implement provisions introduced four years earlier.

What to do next

In an ideal world, advisers will ensure that their clients either self-assess any exposure to the HICBC or provide a timely notification to HMRC of their potential exposure.

However, if you have a case where a discovery assessment has been made, an appeal would seem worthwhile as this issue is not yet at the end of the road. The First-tier Tribunal would be likely to stay any such appeal, pending the outcome of the Wilkes case.

The alternative is to advise clients to consider opting out of the child benefit (whilst retaining the NI credit).