Universal Credit

Inconsistent Incentives

How the overlap between Universal Credit and the High Income Child Benefit Charge limits work incentives

It is well known that some groups in the UK face high effective tax rates, but the most punitive rates of personal tax arise in a situation that has been almost completely overlooked. Families with an earner on £50,000 to £60,000, whose Child Benefit is withdrawn and who are also eligible for Universal Credit (UC), find themselves in this position. The combined effect of Child Benefit and UC withdrawals leads to truly punitive effective tax rates of at least 80 per cent for those with one child, 83 per cent for those with two children and 87 per cent for those with three (and higher still if we include student loan repayments or pension contributions). About 50,000 families will be in this situation from April, but this will nearly double by the end of the decade. At that point, there will also be another 250,000 families where the highest earner is on between £40,000 and £50,000 who could easily be brought into this zone of punitive marginal deduction rates if they receive a pay rise.

This problem is growing over time due to successive government decisions to freeze the income thresholds at which Child Benefit is withdrawn (first set in 2013), which brings more families into this income zone. Compared to rising destitution and increasing child poverty, this is clearly not the most pressing issue facing the social security system, but it certainly amounts to a serious design flaw; no rational policy maker would ever have drawn up the current system. Unless we are to accept that ever-more families will face a £10,000 stretch of income where there is no point in seeking higher earnings, the Government will have to fix this situation. But there are no cheap and easy solutions: there will be difficult trade-offs involved, whether we solve this either by paying Child Benefit to more families, or bringing Child Benefit into UC. The most expensive of these options: making Child Benefit universal would cost around £4 billion per year.

Benefits, not just taxes, affect work incentives

Who faces the highest marginal effective tax rates in the UK? Not the super-rich, nor those living off inherited wealth; rather it is families with children.[1]

It is widely known that families receiving Universal Credit (UC) face high marginal effective tax (or marginal deduction) rates: as their earnings rise, not only do they pay Income Tax and National Insurance contributions, but so is UC withdrawn. As a result, for those earning enough to pay the basic rate of Income Tax, there is an effective tax rate of 69 pence and only 31p of each extra pound is kept. As of April next year, around 3 million working adults will be eligible for UC and face high marginal deduction rates (MDRs) of 69 per cent or higher. Where adults are paying the intermediate rate of Income Tax (21 per cent) in Scotland, MDRs will be higher (starting from 70 per cent).

Families typically further up the income scale can also see high marginal deduction rates as Child Benefit is withdrawn through the High Income Child Benefit Charge (HICBC). This affects any family where one person receives Child Benefit and (at least) one person has a gross income above £50,000 per year; those with incomes above £60,000 have their Child Benefit completely offset by the HICBC, and those with incomes between £50,000 and £60,000 effectively lose a proportion of the Child Benefit on a sliding scale.[2] This mechanism means that, from next April, those with incomes between £50,270 and £60,000 face a marginal deduction rate of 55 per cent if they have one child (40 per cent from Income Tax, 2 per cent from National Insurance contributions and 13 per cent from from the HICBC), 63 per cent if they have two children and 71 per cent if they have three children (with a further 8 percentage points for each additional child). Rates for those living in Scotland will be 2 percentage points higher, reflecting their higher rate of income tax (42 per cent from April).[3] From April, we estimate that around 2 million families, or 1-in-4 (27 per cent) of those with children, will have some Child Benefit effectively partially or fully withdrawn because one person has an income over £50,000, up from 1-in-8 when the policy was introduced in 2013. Around 1-in-13 (600,000 families, or 8 per cent) families with children have someone earning between £50,270 and £60,000, and – unluckily – experience the high marginal deduction rates described above.[4]

Usually these two issues are thought of as affecting two distinct parts of the population. What has gone almost totally unnoticed is that a relatively small, but rapidly rising, number of families are in the position of being entitled to UC while also having their Child Benefit withdrawn. This combination of UC and Child Benefit withdrawal results in truly punitive marginal deduction rates. As discussed above, anyone with an income between £50,270 and £60,000 and paying the HICBC is already facing a marginal deduction rate of 63 per cent if they have two children, meaning that they keep 37p of each additional pound earned. If these families are also receiving UC (and paying the HICBC through an adjustment to their PAYE code), then UC will take away 55 per cent of what is left: this will lead to a marginal rate of 80 per cent for those with one child, 83 per cent for those with two children and 87 per cent for those with three (these all assume that the taxpayer is living outside Scotland; see also Figure 2).

These MDRs may sound very high, but for some of this group they are a significant understatement. First, if the person paying the HICBC is repaying student loans we need to add 9 percentage points to each of these figures.[5] We do not have good statistics on how many people caught in the £50,270 to £60,000 bracket will also be repaying student loans, but two-thirds (65 per cent) of employees with children in this earnings bracket are graduates. Clearly, many in this group will not be repaying student loans today, but over time we would expect a rising percentage to do so, not least as higher student fees leave more graduates repaying student loans for longer. [6]

Second, these MDR calculations do not take account of any individual pension contributions. These are slightly different in nature, in that they are effectively a transfer of income across an individual’s life-cycle, but still represent a state-backed reduction in today’s pay packet that reduces current living standards. If, on top of student loans, we include a pension contribution of 5 per cent of gross income, we get to MDRs of 89, 93 and 96 per cent respectively.[7] This means that a hypothetical earner – with pension contributions and student loan repayments – with an income of £50,000 and two children would only take home £800 more if they received a £10,000 pay rise.

Lastly, as if this wasn’t all complicated enough, the effective tax rates quoted so far assume that those affected will pay the HICBC through an adjustment to their PAYE code. If instead the payment was made via a one-off transfer to HMRC then the MDRs would be even higher (reaching over 100 per cent for a family with three children even before we consider student loans for pension contributions).[8] Table 1 summarises some of these scenarios.

Table 1

The number of families affected by a Child Benefit withdrawal and receiving UC is currently small, but growing

Some may be surprised that families containing someone earning over £50,000 can receive UC. This isn’t, of course, the typical position of most families on UC (who are further down the income scale) but for those with high housing rents, or (if a single parent family or a couple with a second earner on low earnings) with significant childcare costs, it is perfectly possible. As Table 2 shows, a single parent with three children and typical housing costs will be entitled to receive some UC on pre-tax earnings of anything up to £66,000 (assuming the person had less than £6,000 in savings). If we also consider childcare, then a single parent with one child, typical housing costs of £150 per week, and childcare costs of £460 per month, will be entitled to some UC with earnings up to £62,000. Indeed, Rishi Sunak’s (welcome) decision when Chancellor to cut the rate at which UC was withdrawn as earnings rise inevitably resulted in more middle- and upper-middle- income households becoming eligible for UC.

Table 2

At the moment, the number of families in this position is relatively small, but it is set to rise rapidly. There are no official figures, but we estimate that from April roughly 50,000 adults will be entitled to UC and liable to the HICBC; we also forecast that the number will come close to doubling to around 90,000 by the end of the decade (see Figure 1). At that point, there will also be another 250,000 adults earning between £40,000 and £50,000 who are entitled to UC and are the highest earners in their family (so not currently facing Child Benefit withdrawal) that could easily be brought into this punitive MDR income zone if they receive a pay rise.

Figure 1: By the end of the decade, 87,000 adults are set to be affected by the High Income Child Benefit Charge while being entitled to Universal Credit

The reason for the rise in numbers is the long-term freeze in the threshold at which Child Benefit is withdrawn. The £50,000 to £60,000 band over which Child Benefit is effectively withdrawn is still the same as it was when first announced by George Osborne almost a decade ago, and is currently set to be frozen forever.[9] If £50,000 and £60,000 were the right figures back then, then the ‘right’ cash figures in 2023-24 – allowing for CPI inflation – would be £64,000 and £77,000, which would be uprated to around £72,000 and £87,000 by the end of the decade.[10] Meanwhile UC uprating (though this, too, has been frozen some years) plus the cuts in the ‘taper rate’, mean that families further up the income scale have become eligible.

Bad policy design leads to perverse outcomes

No rational policy maker would ever have designed the system that we have now. Its distortions should serve to remind us that, as a general rule, long-term freezes in tax or benefit thresholds make for poor policy. But stepping back from the details, what does this problem say about our current support systems for families with children?

First, and most obviously, we have inadvertently created a £10,000 stretch of income – at an increasingly busy point in the income distribution – where, for the higher earner in a family with children, it is simply not worth getting a pay rise. There is now a iron cap on aspiration for this subset of families with children.

Second, the coming together of the income band where Child Benefit is withdrawn with the income band where UC is claimed means that families who rent can face a (very) long, continuous stretch of income where really high MDRs apply (see Figure 2 for an example). For example, those who are the sole earner in a family with children need to reach at least £60,000 (adjusted net income) before their deduction rate falls to 42 per cent (or 51 per cent for graduates; and 54 per cent if we include pension contributions). For many of these households, high deduction rates will be the norm until their children grow up.

Figure 2: More families will experience a very long stretch of high effective tax rates

Third, we are asking families on UC and the Child Benefit taper to navigate two different means-tested income support systems run by separate departments – one via a UC claim administered by the DWP and one via the tax system and self-assessment, run by HMRC. This is precisely the sort of complex interaction that the introduction of UC was intended to remove (and the detail in endnote 8 gives just a glimpse of the complexity of this interaction). As if that wasn’t bad enough, families paying for formal childcare will also need to work out whether they would actually be better off claiming tax-free childcare (a separate process) rather than support for childcare through UC (families cannot receive both). And, as a final twist, households with an earner above £60,000, who therefore will not benefit (at all) from Child Benefit, still need to register for the benefit if the person caring for a child under 12 is to receive their NI credits and therefore protect their future state pension entitlement. It is something of an understatement to say that this is not what a family-friendly tax and benefits system looks like.

Fourth, if the Child Benefit threshold freeze continues alongside UC uprating, eventually a greater share of other family types will be dragged into this high MDR trap (i.e. more dual earners). For example, by 2030-31 a dual-earning couple with high rent or childcare costs will be entitled to some UC: for example, with one earner on £50,000, and a partner working part-time on the minimum wage and earning £15,000.


[1] We have benefited greatly from conversations with officials at the Low Incomes Tax Reform Group, and the Child Poverty Action Group. All errors remain the authors’ own.

[2] The HICBC applies to individual adjusted net income, that is gross income minus pension contributions and charitable donations. For simplicity we will refer to ‘adjusted net income’ as just ‘income’ for the rest of the note.

[3] These calculations assume no pension contributions and no student loan repayments. Individuals in families with incomes between £50,000 and £50,270 will be liable for the basic rate of Income Tax, and so face lower MDRs (this may also be the case where individuals have income composed of dividends or capital gains as well, cases which we do not consider in this note).

[4] Analysis of: DWP, Family Resources Survey using the IPPR Tax Benefit Model. These families can avoid the HICBC (as well as avoiding having to be part of income tax self-assessment) by opting to register for but not receive Child Benefit, but they would be financially worse off if they did this and the main earner has income of less than £60,000, compared to a situation where they receive Child Benefit and pay the HICBC.

[5] UC entitlement is not affected by student loan repayments, so the MDRs are additive.

[6] Analysis of: ONS, Labour Force Survey, October 2021 to September 2022.

[7] MDRs are only very slightly affected by pension contributions, because contributions to a pension reduce income tax liability, increase UC entitlement, and reduce how much is paid in the HICBC.

[8] When those on UC and paying the HICBC do so through an adjustment to their PAYE code, the HICBC reduces post-tax earnings. This creates an interaction with the UC system which responds to the apparent reduction in household income by increasing the family’s entitlements to UC. In effect, UC cushions the income hit of the HICBC. But this interaction does not apply to anyone who pays the HICBC back as a one-off transfer to HMRC. For those people, the UC system will (typically) not take into account the fact that a rise in earnings will lead to a higher HICBC, and consequently the resulting MDRs will be even higher: 86, 95 and 103 per cent (without student loan repayments or pension contributions) for families with one, two or three children. As a result, it will usually be in the family’s interest to pay the HICBC via a change to their PAYE code if there is a possibility that they may be entitled to UC. The situation is even more complicated for self-employed individuals, who may be able to argue that a one-off payment of a HICBC should be deducted from their post-tax income for UC purposes in the month in which the payment is made (thereby increasing entitlement to UC).

[9] Unlike most parts of the social security system, and some parts of the personal tax system, there is no statutory requirement to increase (or even to review) the £50,000 and £60,000 thresholds in the HICBC. The fact that the band of income over which CB is withdrawn is by default fixed in nominal terms, even though the value of CB usually rises with inflation, means that the marginal rate implied by the HICBC increases each year (so that an ever-increasing entitlement to CB is always withdrawn over a fixed band of income).

[10] Using forecasts from: OBR, Economic and Fiscal Outlook, November 2022.