On Monday, Child Benefit will effectively become an income-related benefit for the first time, reducing public spending by an estimated £1.5 billion in 2013–14. We estimate that about 820,000 families, in which at least one adult has a taxable income exceeding £60,000 per year, stand to lose all their Child Benefit via a new income tax charge unless they change their behaviour in response. And about 320,000 families in which the highest-income adult is on between £50,000 and £60,000 would have some, but not all, of it clawed back. The affected families stand to lose an average of about £1,300 per year. The remaining 85% of families currently receiving Child Benefit will be unaffected for now, although more will be affected in time because the £50,000 threshold is planned to be frozen in cash terms.

In the context of a large fiscal consolidation, one can understand why the government is looking at universal benefits going to those on higher incomes as one candidate for cutbacks. But various features of the design of this particular policy look problematic.

First, much attention has understandably focused on inequities of treatment between 1-earner and 2-earner families. The means test uses information only on the taxable income of the highest-income family member so, for example, a 2-earner couple with taxable income of £100,000 split equally between them would retain all Child Benefit, but a 1-earner couple or lone parent with taxable income of £60,000 would lose all of it. This kind of situation does not arise with existing means-tested welfare payments, which are based upon family income. (But note that, because a family’s Child Benefit entitlement will depend upon the income of the highest-income individual, this is not simply an argument about family versus individual assessment: this policy is an unusual hybrid of the two.)

A second unusual feature of the reform relates to its impact on incentives for individuals to reduce their taxable income by, for example, working less or contributing more to a private pension. All families’ Child Benefit entitlements will be exhausted once the taxable income of the highest-income adult reaches a fixed level – £60,000 – regardless of how much Child Benefit has to be withdrawn. To achieve this, the rate at which it is withdrawn as income rises above £50,000 has to vary with the number of children in the family, i.e. with the amount of Child Benefit the family receives. Specifically, affected taxpayers will pay back one per cent of their family’s Child Benefit for every £100 by which taxable income exceeds £50,000. One per cent of Child Benefit is £10.56 per year for a 1-child family, and an additional £6.97 per child for larger families. Hence the marginal tax rate between £50,000 and £60,000 is increased by about 11 percentage points for the first child and by an additional 7 percentage points for each subsequent one. So, for example, while about 320,000 people will find that their marginal income tax rate increases to more than 50%, about 40,000 of them - those with three or more children - will find that it jumps to at least 65%. This is illustrated by the Figure, which shows marginal income tax rates for taxpayers with different numbers of children.

Figure: Marginal income tax rates in 2013-14 for the highest-income member of families with children

Source: Author’s calculations.

This unusual feature of the policy has another consequence in the long run. Since higher Child Benefit entitlements imply higher withdrawal rates, marginal tax rates between £50,000 and £60,000 will also have to rise over time as cash-terms Child Benefit rates rise in line with prices (or indeed in line with anything else). This may sound obscure and technical, but in time it would be important. For example, indexation of Child Benefit rates at 2% per year would increase these marginal rates by a further 5 percentage points for a 3-child family over one decade. To avoid this, some families with an adult on more than £60,000 would have to be allowed to retain some of their Child Benefit. But there are currently no provisions to do this: the £60,000 withdrawal end-point is not a parameter that is planned to be uprated over time.

That is an important detail, but the government could and should address it without any structural upheaval. Other problems are more fundamental to the design of the policy. It creates a series of administrative complexities, including the need for up to 500,000 more individuals to fill in income tax self-assessment forms according to HMRC. But perhaps the biggest concern is the incoherence it creates in the welfare system. We already have the Child Tax Credit, and soon its imminent replacement, namely the child additions within Universal Credit. The reform to Child Benefit will mean that we have two systems of income-related support for children. But the relationship to income will be completely different in each case: based on family income in one case and the income of the highest-income family member in the other; withdrawn at different rates as income rises; and with the withdrawal starting at very different income levels.

With the introduction of Universal Credit, the government will integrate six of the seven existing means-tested benefits and tax credits into one. IFS researchers have argued that the basic principles behind this move have much to commend them. But almost simultaneously, the government is introducing a new and separate means test for Child Benefit which will work in a completely different way, as well as making the seventh existing means-tested benefit – Council Tax Benefit – much more complicated by asking every Local Authority in England to design its own scheme. It is unclear whether the net effect of all this will be to improve the welfare system.