House of Commons - Social Security

Select Committee on Social Security Second Report

The structure of payments to children

32. The present tax credit and benefits system loads greater financial assistance on to the first child. In Income Support, and income-based Jobseeker's Allowance, the presence of the first child entitles a family to an extra 'family premium' worth between £14-£16. The Children's Tax Credit, worth £10.00 from April 2001, becomes payable for a first or only child; no extra amount is payable for subsequent children. Child Benefit is set at one-third lower for second and subsequent children, than for the first child.

33. The imbalance between first and subsequent children is carried into the new system of support for children. The illustrative rates for ICC given by the Treasury[69] show that total support for the first child of a working family, including Child Benefit, would be £50.[70] But a second or subsequent child would only get £36.35, due to lower Child Benefit (only £10.35) and no Children's Tax Credit. The amount paid for a second or subsequent child is therefore 27% less. Sue Middleton told us: "In terms of birth order, Child Benefit assumes that a first born or an only child is going to cost a lot more than a subsequent child. Presumably the reasoning behind that is that there are savings from being able to pass on goods from one child to the other. What the spending evidence suggests is that those savings are not that great in regular spending terms; they are relatively small. The average a second child in a family has spent on it is something like 17 per cent less than an only child, which is not vastly different."[71]

34. Another consequence of the present structure of children's financial support is that children in larger families receive proportionately less help. Yet, as IFS pointed out, rates of worklessness and poverty are significantly higher for families with three or more children than for families with one or two children.[72] Sue Middleton argued that, if ICC was not going to be paid at a single flat rate, it should be structured to pay more for more children in the family: "if we are actually into poverty relief, there would be a case for assisting larger families because that seems to be where the big drop [in income] comes, where you have three or more children."[73] Both Canada and Australia offer large family supplements, Australia for four or more children,[74] Canada, for a third and each additional child.[75]

35. We question whether the present structure of financial support for children, which in recent years has disproportionately channelled support to first or only children, is justified in the face of evidence concerning patterns of expenditure on children, the greater poverty of children in larger families and the decreased likelihood of such families being able to lift themselves out of poverty through work. We recommend that ICC rates do not distinguish between first and subsequent children, and that levels of Child Benefit between first and subsequent children be equalised.

36. Another issue raised was whether there should be higher payments for children according to age. The work of the Family Budget Unit for example suggests that teenagers cost more than young children. In contrast, as discussed above, Sue Middleton's work examining expenditure patterns on children found little difference. However, the inclusion of childcare costs can alter the picture. Should there be an extra amount for children under five years to reflect the direct and indirect costs of childcare? In Australia, for example, families get more if there is a child under five.[76] Jane Millar told us , "In the past we have not in a sense treated Child Benefit as being particularly a part of the contribution towards the costs of care. A supplement for under-fives could be more directly perceived in that way, that somehow you are also contributing not just to the parents' direct costs of those children but also to the costs of caring for those children and, by paying it through Child Benefit, allowing the parents opportunity to chose the sort of care, self-care or paying for care, that they use."[77] However, she added, "once we get into these targeting issues about different types of family we need to think what they are for, what we are trying to do with these sorts of payments, why do we want to introduce this sort of targeting, particularly if part of what we want to do is ensure that we have a system that is simple to administer and simple to understand."[78]

37. The debate about compensating parents for the direct and indirect costs of childcare is one we touched on in this inquiry,[79] but it was not the focus of the inquiry. There are arguments in favour of giving extra financial resources to parents with young children to enable them to choose between paying directly for childcare in the manner they think most appropriate, or using the money to compensate for the drop in income caused by a parent choosing to stop work or work less in order to care for children. The question is whether ICC is the most appropriate vehicle, as opposed to a higher amount of Child Benefit for example, or a separate allowance or tax break. We consider that there is a stronger case for keeping the design of ICC as simple as possible - both from the point of view of the administrators, and also the recipients. We have concluded that in structuring the system of payments to children through ICC, the aim of simplicity - in terms of public understanding and in administration - must be paramount. We therefore recommend that there should be no age-related element in the rate of child credit per child.

ICC and work incentives

Crossing the no-work/work divide

38. Benefits policy has always paid close attention to the 'unemployment trap,' whereby people can find themselves worse off financially by working, compared to receiving out-of-work benefits. But, as the Treasury discussion document on ICC has acknowledged, tensions can arise between policies designed to create better work incentives and those designed to give increased family support.[80] Measures aimed at giving greater financial support to children in the poorest families risk being undermined by criticisms that they increase the unemployment trap, thus reducing work incentives. In evidence to us, DSS officials made it clear that in internal debates about the adequacy of Income Support for children, judgements concerning adequacy took place against the background not only of public expenditure considerations, but also of "maintaining the work incentive issues."[81]

39. ICC seeks to resolve this conflict. One great advantage of ICC is that, by paying maximum ICC to families with children on the lowest incomes, whether out of work or in work, channelling greater support to children does not deepen the unemployment trap. Families who take low-paid work can carry with them the ICC which they received when out of work. Sue Middleton commented: "One of the joys of taking benefits for children out and paying them separately is...that all the work I have done with families would lead me to conclude that it would not have any effect on work incentives...Families meet the needs of the children and it does not matter whether you are in work or out of work."[82]

40. Tackling child poverty therefore does not need to be constrained by concerns about worsening work incentives. Indeed, the ability of ICC to offer a "portable and secure income bridge spanning welfare and work" is likely positively to assist work incentives, in that it will overcome the financial disruption associated with moving off benefits into employment, which often discourages parents from making the transition into work.[83] We welcome the 'portability' of ICC for families moving into work, whereby the maximum rate of ICC will be retained across the no work/work divide. It will enable resources to go to children in the poorest families whilst providing their parents with certainty of income if they move into low paid work.

41. A related issue is the level of earnings up to which the maximum amount of ICC should be paid. Professor Millar pointed out that "many families as they move from unemployment to work move not into high paying jobs, many families move into relatively low paying jobs."[84] She argued it was important to keep ICC steady across the change from unemployment to employment and back again. The question is, across what income band should ICC be kept steady, to prevent financial disruption as circumstances change? The Fawcett Society suggested, for example, that withdrawal rates for ICC should start at a relatively high level, in order to avoid disincentives to work for second earners.[85] The Women's Budget Unit pointed out that, at present, a second earner in a family claiming WFTC faces a high withdrawal rate of 55 per cent if she starts work, giving her little incentive to do so.[86] We have concluded that in order to reduce child poverty, assist the transition into work, and aid administrative simplicity there are strong arguments for maintaining a wide band of income across which maximum ICC is paid, before it starts to be withdrawn.

Making work pay: the question of tapers

42. The question of the level of earnings up to which the maximum amount of ICC should be paid cannot be separated from the question of the rate at which it should start to be withdrawn thereafter. Given finite resources, the more resources are targeted on those on lower incomes, the steeper the withdrawal rate has to be as income rises. Yet high withdrawal rates are traditionally seen as creating work disincentives: people are unlikely to work harder or longer if, as a result of the combined effect of tax and benefit deductions as their earnings rise ('marginal tax rates'), their net income remains relatively static despite their extra work. On the other hand, if withdrawal rates are made less steep, resources have to be spread higher up the income scale, potentially resulting in less help for those on low incomes. And the effect is that more people across a wider income band are brought into the tax/benefits trap, albeit at a less severe rate. Andrew Dilnot of IFS explained:

"If our principal concern is child poverty, then it is reasonable to think that you might want to focus Income Support help further down the income distribution. It may well be that a world with even higher marginal tax rates than we have now, so that we get out of the benefit region more quickly, would allow us to be even more generous to those at the very bottom without extending the potentially bad consequences of means-testing further up the income distribution. On the other hand you might take a view that it is simply iniquitous that people should face these very high marginal tax rates and so seek to reduce the taper rate even below the current taper rate. The trouble with that is that...to make a big difference in the aggregate marginal tax rate people are facing means either extending entitlement to means-tested benefits into the higher rates of income tax or making the benefit much less generous at low levels of income. There is no escaping that conundrum and therefore there is no absolute right answer to how you should do the tapering."[87]

43. The balance to be struck between greater targeting and high marginal tax rates, and less targeting and lower marginal tax rates depends on the view taken of the work disincentive effects of high marginal tax rates. The DSS told us, "Whatever the technical marginal deduction rates are, it is the way you deliver the thing to the customer that makes the real difference...Even if the Marginal Deduction Rate is not as good as we might wish it to be, with support and a vision for a better career in the future, that is what makes the difference for individuals..."[88] Peter Whiteford suggested that, on the one hand, there was not much evidence in Australia that people had reduced their hours of work in order to get higher benefits, because in reality they were still better off with longer hours of work.[89] On the other hand, in the mid-nineties, when marginal tax rates were particularly high, "it led to a sort of perception which I think is wrong that 'I would be better off if I did not work at all'. It is not so much that it changed work behaviour but it... made people think incorrectly that they would be better off if they were on benefits rather than in low paid work."[90]

44. How have other countries with similar integrated child payments resolved the question of withdrawal rates? Australia has a model similar to that proposed for ICC, with an initial 'plateau' of income across which maximum ICC is paid, followed by a fairly steep withdrawal rate across a narrow income band, then a second plateau at a lower level, before a further tapering away. Canada in contrast has a lower withdrawal rate across a wider band of income; there is a gradual tapering away of child payments with no second plateau.[91] Peter Whiteford's view, however, was that, although the Canadian system implied higher marginal tax rates over a much broader range of income, in practice this did not happen, because child payments were based on annual income in the previous year and were not affected by changes in income in the course of the year.

45. In the UK, as a number of respondents to the inquiry emphasised, the withdrawal rates for ICC cannot be considered in isolation from the tapers which apply to Housing Benefit and Council Tax Benefit.[92] With a taper of 65 per cent for Housing Benefit and twenty per cent for Council Tax Benefit, a family currently in receipt of Working Families Tax Credit can find themselves as little as five pence better off for each additional £1 of gross earnings. In our recent report on Housing Benefit,[93] we called for a review of the relationship between Housing Benefit and the proposed new Child and Employment Tax Credits, and for consideration of the inclusion of a housing credit as part of the planned reforms in 2003. The response of the Government has been disappointing.[94] Ms Ghosh of the DSS told us that they intended to look at the relationship between Housing Benefit and ICC: "We do not want to make marginal deduction rates worse than they are at the moment."[95] We are pleased that the Government is looking at the interaction of ICC and Housing Benefit, but consider that the aim should be to improve marginal deduction rates rather than simply not make them worse than they are at present. Nick Macpherson of the Treasury argued that the Government were addressing the "multiple taper effect"[96] by floating people off Housing Benefit and Council Tax Benefit altogether through higher payments of Working Families Tax Credit: "I think this will continue to inform our thinking. The more money is put into these particular credits the fewer people will be both on these credits and on Housing Benefit."[97] Nevertheless, the problem will remain and will bite particularly hard on families in areas of high housing costs, or who are living in private rented accommodation or in temporary accommodation.

46. It is in the nature of means-testing that a balance has to be struck between greater targeting and high marginal tax rates, and less targeting and lower marginal tax rates but over a wider proportion of the population. The role of marginal tax rates in affecting people's decisions to work can be exaggerated, but it does adversely affect the Government's key objective: to make work pay. We repeat the recommendation made in our report on Housing Benefit that consideration should be given to removing as many people as possible from the necessity of making claims for Housing Benefit alongside ICC, either by increasing the earnings disregard or examining the possibility of including a housing credit as part of the reforms of tax credits in 2003.

The nature of the means test

Period of assessment, length of award and responsiveness

47. One of the key design parameters for ICC is the length of an award and the period of assessment. ICC will bring together three systems, each of which have made, for good reasons, different decisions for determining length of award and period of assessment. Income Support and income-based Jobseeker's Allowance provide immediate income for poor families with little other income. These benefits are assessed weekly and paid fortnightly. They are heavily means-tested, in the sense that any change in income or family structure is likely to have an immediate effect on the payment due. In contrast, WFTC and DPTC are based on a 'snapshot' of income in the weeks before the claim, and are then paid for six months at a fixed rate, regardless of changes in income or family size. The relative inflexibility of WFTC and DPTC reflect the fact that they are a supplement to families who have other income through working, albeit low. The Children's Tax Credit goes much higher up the income scale, and for most families will be a flat rate reduction in their annual tax bill, regardless of income. For higher rate taxpayers, the amount of the Children's Tax Credit, if any, will be calculated only when the person's income for the year in question is known. Andrew Dilnot of IFS suggested that in bringing the three regimes together, there were likely to be "some awkward corners."[98] Jane Millar commented, "I think one of the challenges to the UK is, we have to think about income-testing in a different sort of way. We have been used to using income-testing or means-testing to identify current needs very precisely and we have very detailed rules in those. I think you might need what you might call 'means-testing with a light touch'...to make this sort of thing work."[99] Andrew Dilnot's view was "We can do means-testing with a lighter touch, but the consequence will be more expenditure. If we do not want to spend more, then some of the means-testing we have to implement will in certain circumstances seem a little more aggressive than we might otherwise want. Those are some of the sorts of decisions which the Government has to make."[100]

48. Both Canada and Australia assess entitlement to their child credits on an annual basis. In Canada, National Child Benefit is assessed yearly through the tax system, based on income in the previous tax year. Nick Macpherson of the Treasury commented, "if you look at the Canadian system, where it is done on the previous year, a lot of the information is pretty out of date. If you had a good year last year, and you were a couple both of whom were working you might be eligible for very little payment. But your economic circumstances might change the following year and under the Canadian system there would be very little redress, that is just rough justice."[101] When we travelled to Canada, we did indeed hear complaints about rough justice in this situation from recipients of National Child Benefit in British Columbia. Nevertheless, experts on the Canadian system based at the independent Caledon Institute in Ottawa told us that their research had concluded that recipients were willing to accept a degree of rough justice in return for a lack of intrusive inquiries into their personal circumstances. Jane Millar thought the annual basis of assessment in Canada worked because there is a relatively "long plateau" across which maximum payments were made, so a family who suffered a drop in income might already be getting maximum payments; and because there was a safety net, in the form of Income Support administered at provincial level, which could step in if people became unemployed.[102] Tony Orhnial of the Inland Revenue pointed out that, in the UK, children's Income Support would be absorbed into ICC , "therefore we have to design an ICC that can deal within the system itself with sharp falls in income."[103]

49. In Australia, since July 2000 Family Tax Benefit has been based on an estimate of current year's income, with an annual reconciliation at the end of the year. The system has not yet gone through a complete cycle. Potentially, the system appears problematic, in that some people might end up underpaid during the year, whilst others might face large bills at the end of the year because they have received too much. Peter Whiteford thought that when reconciliation took place in July 2001, "there are going to be some nasty scenes around the country...any reconciliation it seems to me must produce more people with debts and a lot of unhappiness."[104]

50. The Treasury argues that there is a case for setting annual awards of ICC in the UK: "An annual award would minimise administrative work for both public and government. It also means that information on incomes could be verified against data collected for income tax purposes."[105] In contrast, Andrew Dilnot put the case forcefully that "it would be a very substantial change to go to a system where at the end of a year those on low incomes identified as being low income by having received some entitlement during the year were asked for a repayment." He added that, "if we are not going to go down that road, then we shall end up paying benefits to people, which if we were to measure their income over a longer period, we might well not think they were entitled to."[106]

51. It is difficult to separate the question of the length of award and period of assessment from the degree to which an award should change in response to fluctuations in income or family size. A balance has to be struck between immediate responsiveness to changes in income, which is administratively burdensome but which ensures that the credit accurately responds to falls or increases in income, thereby containing costs; and less responsiveness, which is administratively simpler, but which may be more expensive, and can lead to hardship on the one hand and accusations of abuse on the other. In Canada, changes in income during the year are generally ignored, although in-year re-assessments can be made for changes in family composition. In Australia, Peter Whiteford pointed out that a crucial design feature of the system was the high level of maximum payments of Family Tax Benefit, which represented 60 per cent of average male earnings. Thus for many recipients, a drop in income does not require an adjustment because they are already receiving the maximum. For those with higher family incomes, from 1996 families have been able to apply for a reassessment if their income went down by any amount. If their income increased by 10 per cent or more, they were required to have a reassessment.[107]

52. We have concluded that the best model for ICC is likely to be a relatively unresponsive structure, with fixed awards of at least six months duration, but with a safety net of 100 per cent ICC when income drops to ICC/JSA level. Otherwise there should be adjustment only for major changes in family circumstances such as the birth of a child; a child leaving the household; a new partner; or loss of a partner. Our conclusion that there should be a relatively unresponsive structure reinforces our earlier recommendation that there should be a wide band of income across which maximum ICC is paid.

69 HMT 2000, Chart 2.6. Back

70 The illustrative rates are based on April 2001 rates, as known in March 2000. Taking into account the March 2001 budget, the total support in April 2001 for the first child of a working family including Child Benefit would in fact be £51.50. Back

71 Q 122. Back

72 Ev. p. 73, para 17. Back

73 Q 127. Back

74 Peter Whiteford, Ev. p. 19. Back

75 Jane Millar, Ev. p. 1. Back

76 Q 72. Back

77 Q 9. Back

78 Ibid. Back

79 Q 8. Back

80 HMT 2000, para 2.22. Back