High income child benefit charge: a number of problems...

High income child benefit charge: a number of problems
27 January 2023

The high income child benefit charge has always been a controversial measure. The case of HMRC v Wilkes revisits the taxation problems surrounding the benefit.

Key Points

What is the issue?

If a family unit is in receipt of child benefit and the recipient (or partner) has taxable income in excess of £50,000, some or all of the child benefit is repayable to the government in the form of a freestanding income tax charge, the high income child benefit charge (HICBC).

What does it mean for me?

There are a number of problems with the HICBC. In many cases, it represents a clawback of all the child benefit received, which is not necessarily suffered by the person who has actually received the child benefit. Some of these problems have also led to certain difficulties in its enforcement.

What can I take away?

If you have a client who is facing a HICBC challenge via a discovery assessment, you should ascertain whether or not the client is protected by the limited transitional provision for pre-30 June 2021 cases.

© Getty images/iStockphoto

The high income child benefit charge  has always been a controversial measure, ever since its introduction by the Finance Act 2012. Its apparent aim was to remove the financial advantages of receiving child benefit from wealthier families without giving the impression that the universality of the benefit was being undermined. In order to do so, it overrode what has been a key principle of income tax ever since 1990 – namely, the tax independence of married couples (and, latterly, civil partners) – by treating couples as a single unit for certain specific purposes.

In short, if a family unit is in receipt of child benefit and the recipient has taxable income in excess of £50,000, some or all of the child benefit is repayable to the government in the form of a freestanding income tax charge, the high income child benefit charge (HICBC). If the recipient of the child benefit is a member of a couple, then the freestanding charge is levied on the partner with the higher income.

Where the payer of the HICBC has income in excess of £60,000, then the HICBC payable is equal to the entire child benefit received (whether it is received by the payer or the other member of the couple). If the payer’s income is between £50,000 and £60,000, then the amount of HICBC payable is simply a commensurate proportion of the child benefit received in the year. In other words, income of £55,000 means an effective repayment of 50% of the child benefit; and income of £57,000 means an effective repayment of 70%, etc.

Some problems with HICBC…

Readers will immediately spot a number of problems with the HICBC.

First, the HICBC is not simply the application of income tax to a source of ‘income’ received by the family. Such an approach would require the child benefit to be added to the person’s total income for the year and tax to be charged at the individual’s marginal rate on that additional income. Instead, the HICBC represents a clawback of (in many cases) all the child benefit received.

Secondly, the clawback is not necessarily suffered by the person who has actually received the child benefit. Indeed, it is a modern example of robbing Peter to pay Paul.

Thirdly, even if the HICBC can be claimed to be a justified exception to the independent taxation of married couples, it does not carry out its task particularly fairly. This is because a couple with two incomes of £50,000 (i.e. total family income of £100,000) will have no liability to pay any HICBC, whereas a single earner family with that earner in receipt of income of £60,000 will be liable to repay all child benefit in the form of the HICBC.

Some of these systemic problems with the HICBC have also led to certain difficulties in the enforcement of the HICBC by HMRC. In the December 2020 issue of Tax Adviser, my article ‘The third Wiseman’ discussed three cases that had been heard by different constitutions of the First-tier Tribunal (with differing outcomes), which each considered whether HMRC was entitled to make discovery assessments to collect the HICBC. One of those cases, HMRC v Wilkes [2022] EWCA Civ 1612 has now reached the Court of Appeal.

The facts of the case

In November 2018, HMRC sent Mr Wilkes a letter suggesting that he might be liable for the HICBC. This was on the basis that his income was in excess of £50,000 and that his wife had received child benefit. In response to the letter from HMRC, Mr Wilkes telephoned HMRC and in late December 2018, based on the information he had provided, was issued with discovery assessments for the 2014/15, 2015/16 and 2016/17 tax years.

Unlike the taxpayers in the other two First-tier Tribunal cases, Mr Wilkes was successful in his appeal against the imposition of the HICBC. However, HMRC then appealed against the First-tier Tribunal’s decision to the Upper Tribunal and, when it lost again, to the Court of Appeal.

The Court of Appeal’s decision

The case came before Lord Justices Newey, Baker and Arnold. The main judgment was given by Lord Justice Newey and the other two judges gave concurring judgments.

The court looked at three issues.

The first was the wording of Taxes Management Act 1970 s 29(1) as it stood prior to the enactment of the Finance Act 2022. Section 29(1) sets out three circumstances in which a discovery assessment could be made, being:

  • where income or gains that should have been assessed have not been assessed;
  • where an assessment is or has become insufficient; or
  • where relief given is or has become excessive.

It was common ground that the latter two limbs did not apply in the present case. Mr Wilkes had not self-assessed his tax liability and therefore there was no insufficient assessment; similarly, this was not a case where excessive relief had been given.

HMRC’s difficulty in respect of the first limb, however, is that the HICBC does not represent a tax on income. Indeed, child benefit is statutorily excluded from being within the definition of income. Accordingly, as concluded by the First-tier and Upper Tribunals, this was not a case where there had been any failure to assess any income.

In the Court of Appeal, HMRC argued that the word ‘income’ should be construed purposively so that it could catch any situation where there was a shortfall in the income tax paid.

However, the Court of Appeal disagreed. It noted that (at least prior to the introduction of certain standalone charges in the Finance Act 2004), s 29 adequately catered for situations where there was an income tax shortfall. In all such cases, there would have to be a source of income that was insufficiently taxed. In other words, the actual wording of s 29(1)(a) was up to the job it was meant to fulfil.

Although new provisions were subsequently introduced that affected how income tax might be levied, the ‘purpose’ of the words in s 29(1)(a) could not be assessed in the light of those later amendments. Accordingly, there was no justification for departing from the actual words of s 29(1)(a). As the HICBC does not represent any charge on income, the first argument was decided against HMRC.

HMRC’s second argument was based on the fact that Mr Wilkes (strictly) ought to have notified HMRC of his liability to HICBC under s 7 of the Taxes Management Act 1970. (When the HICBC was introduced, the usual exception for PAYE taxpayers such as Mr Wilkes was cut back so as to require such individuals to bring themselves to HMRC’s attention. This process would lead to HMRC issuing a Section 8 notice requiring a tax return, which would then have included a self-assessment for the HICBC. In the absence of the inclusion of the HICBC, HMRC could have then been justified in proceeding down the second limb of s 29(1) as the self-assessment would have been insufficient.)

HMRC relied on this failure by Mr Wilkes to start the tax return process, claiming this meant that (ultimately) Mr Wilkes’s other income (his PAYE earnings), which would otherwise have been included in his self-assessment, was not in the end assessed. Therefore, even on the court’s preferred interpretation of s 29(1)(a), it was indeed the case that income that should have been assessed was actually assessed.

However, the court dismissed that argument as well. When the relevant conditions are satisfied, s 29 permits HMRC to issue assessments to remedy the loss of tax that it has discovered. As the court concluded, that loss of tax must relate directly to the income that HMRC has discovered has not been properly assessed. In the present case, the loss of tax (the HICBC) had nothing to do with the unassessed income – it was in fact an additional freestanding charge to tax over and above Mr Wilkes’s actual taxable income. Although there is nevertheless some connection between liability to HICBC and a taxpayer’s income (as the court put it) ‘that limited connection with income cannot, however, render section 29(1)(a) of Taxes Management Act 1970 applicable’. Indeed, as the court continued, it is not as if the HICBC will ‘make good the loss of tax arising from income which ought to have been assessed’. Accordingly, HMRC’s second argument was similarly dismissed.

HMRC’s third argument relied on asking the court to apply a ‘rectifying construction’ to the legislation. This is a rarely used power whereby a court can overlook an obvious error in the legislation and pretend that the statute reads how Parliament had clearly intended the legislation to read. See, for example, my article, ‘Mind the (property) gap’, in the November 2013 issue of Tax Adviser. When considering a rectifying construction, the courts are mindful not to step into the shoes of the legislature (Parliament): instead they are restricting their activities so as to do no more than interpret the statutory words.

However, once again the court disagreed with HMRC’s approach. The main stumbling block was the fact that the court could not be ‘abundantly sure’ that Parliament had intended HMRC to use discovery assessments to recover HICBC. For example, there was no clue in the statute that this was something that had been considered and either overlooked or improperly implemented. As the court noted, it is possible that no actual intention had been formed on this point either by the drafter or by Parliament. HMRC might well have been expected to use its powers to request tax returns. Indeed, HMRC would have had sufficient information (being the payers of child benefit) to enable it to request tax returns within the relevant time limits.

Another obstacle to a rectifying statute was the fact that the ‘problem’ could have been solved in various different ways. Accordingly, the court could not have been sure as to the approach that Parliament would have taken to ensure that taxpayers such as Mr Wilkes paid the HICBC. As a result, it would be wrong for the court to assume that HMRC’s approach represented the only reasonable way forward.

For these reasons, HMRC’s appeal was dismissed.


It should be recalled, as the court noted, that Parliament has indeed sought to rectify the situation. Section 29(1)(a) (as substituted in last year’s Finance Act) now ensures that a discovery assessment may be made if there is a discovery that ‘an amount of income tax or capital gains tax ought to have been assessed but has not been assessed’. Furthermore, Parliament felt that this was an occasion where the statutory revisions will generally be applied retroactively. Accordingly, the full effects of the Wilkes decision will be enjoyed by very few individuals.

This retroactive legislation is in itself an unsatisfactory practice that has become extremely common in recent Finance Acts. What it amounts to is a statement that HMRC might well have failed to get its paperwork right. However, it has managed to persuade Parliament to intervene and allow the paperwork to be remedied after the event. It will be noted that taxpayers who discover errors of their own and who seek to repair the paperwork at a later date will usually be committing fraud.

Adding to the discomfort I have with this retroactive legislation is the fact that some cases are protected from the Finance Act 2022 changes and therefore will be able to take full advantage of the Wilkes decision. These were cases where the s 29(1)(a) point was live (or likely to be so) as at 30 June 2021 (the date on which the Upper Tribunal gave its decision in the Wilkes case). Typically, these will be taxpayers who were better advised and therefore knew of the restrictions in HMRC’s powers to issue a discovery assessment.

Given the inherent unfairness of the HICBC itself and the fact that these were generally cases where HMRC actually had all the information it needed to ensure that HICBC was properly taxed at the time, it is my view that Parliament should have been far more careful before sanctioning retroactive legislation to remedy HMRC’s administrative shortcomings and then protecting only those taxpayers who had access to better advice. However, given that fiscal drag is bringing even more taxpayers within the scope of HICBC, perhaps now is time for the entire HICBC code to be the subject of some serious reconsideration.

What to do next

If you have a client who is facing an HICBC challenge via a discovery assessment, you should ascertain whether or not the client is protected by the limited transitional provision for pre-30 June 2021 cases. If the client is so protected, then (based on my experience in relation to my one HICBC client) I would expect HMRC to be writing to you or your client to in order to withdraw its assessment.