Means tests and tax credits

The tax credit is New Labour's big idea in welfare policy. It heralds an extension of means testing and a start to a merger of the tax and benefit system. It should also make work pay and reduce, if not abolish, child poverty. But is it too complex?
May 19, 2003

Little more than a year after taking office in 1997, Frank Field became one of the first ministerial casualties of the New Labour government. The leading advocate of the "contributory" principle in welfare policy had gone. The conflict with his ostensible boss, Harriet Harman, began as an intellectual dispute but ended as a personal one. Social policy was thereafter made by Gordon Brown's treasury, introducing the long-heralded device of the tax credit. "Targeting" became the orthodoxy of new-left social policy-advocates of contributory and universal benefits over the means test had been routed.

This victory is now so complete that it is rarely noted how remarkable it is. Outrage at the means test was once a badge of honour of the left. In the 1930s, there was nothing more degrading for the poor, apart from their poverty, than a visit from the officials of the Public Assistance Committee. EW Bakke complained in I"nsurance or Dole?" (1935) of "the recurrent knocking of the investigator at one's door, the knowledge that all eyes in the street are on the investigator... the experience of having the public eye on one's private affairs." Aneurin Bevan attacked the household means test as "a principle that eats like an acid into the homes of the poor. In the small rooms and around the meagre tables of the poor, hells of personal acrimony and wounded vanity arise."

In 1971, opposition to Keith Joseph's family income supplement inspired Frank Field and David Piachaud, in a famous article in the "New Statesman", to describe "the poverty trap" that was an inevitable feature of means-tested benefits. The poverty trap closes when somebody starts to pay tax and national insurance at the same time as their benefit is withdrawn, resulting in a minimal increase to net income and destroying the incentive to find a better paid job or to work longer hours. This incentive problem has informed the critique from the left (and right) ever since. The 1989 Labour policy review stated: "We want to reduce the need for means-tested benefits by providing insurance benefits as of right." In 1993, Gordon Brown told the Labour conference: "I want the next Labour government to achieve what... has never been achieved. The end of the means test for our elderly people."

Nearly a decade later, this rhetoric is an unrecognisable guide to Labour policy. The contributory principle has continued its late 20th-century decline and more people have been brought within the ambit of the tax credit-based means-testing state than ever before. Total spending on benefits and tax credits for 2003/4 is ?126 billion, 30 per cent of total public spending of ?420 billion. Only ?55 billion, 45 per cent of the total amount spent on social security (basically benefits and credits), is covered by national insurance contributions.

The chancellor has called his tax credit-based strategy "progressive universalism." In a speech in December 2000, he said that "Our approach helps all and is at the same time progressive-a progressive universalism... as we build a fully modern tax and benefit system under which for the first time the taxman can give money as well as receive it." Progressive universalism: neither means testing, nor universal benefits, but both at the same time. Could there be anything more third way? A universal base to the welfare state but resources targeted at the least well off. And, into the bargain, the claim that this is a step towards the goal of an integrated tax and benefit system. But what is progressive universalism, and is it quite the departure from Labour history that it seems?

INTEGRATING TAX AND BENEFITS

Progressive universal social policy is not just about the balance between means-tested and universal benefits. If that were all there was to it, then we could describe any British government since Attlee's as progressively universal. Since 1945, both means-tested and universal benefits have increased as a proportion of the welfare bill while benefits paid directly from individual contributions through national insurance have declined. This decline became very steep during the Thatcher years and has continued, at a reduced gradient, since 1997. In 1964, 73 per cent of the social security bill was paid out in contributory benefits compared with 45 per cent now.

So what is the claim to novelty? It is partly the method of benefit delivery-the tax credit. Moreover, both means-tested benefits and universal benefits such as child benefit have been rising sharply-a big difference from the Tory years. And underlying these reforms are two ambitious social policies. First, always to make work more profitable than benefits. Second, to eradicate child poverty.

The integration of the tax and benefits system is the holy grail of British social policy. It makes no sense for the inland revenue to take with one hand what the benefit service is giving with the other. Far better to face a single assessment of means. A tax credit is a means-tested benefit delivered through the tax system or the pay packet. Rather than having separate benefits which people claim individually, the tax credit allows government to pay out in one lump. However, that simple formula obscures the many difficulties which have broken several previous attempts at integration.

The Labour manifesto of 1964 offered an "income guarantee" for pensioners. Douglas Houghton, the minister in charge, hoped to extend his scheme, which he called an income tax in reverse, to all claimants and those in low-paid work. He soon encountered the problems. Tax returns, for example, are annual but circumstances change more frequently.

Undeterred, the Conservative party had a go as well. Geoffrey Howe had advocated an integration in a pamphlet in 1961 and some of his ideas found their way into Keith Joseph's green paper in 1972. The cost was high-over ?5 billion in today's prices-but legislation would have followed if the Tories had won the 1974 election. The argument resurfaced in 1983 when Norman Fowler, an advocate, had a losing row with Nigel Lawson, the chancellor.

Then in 1999, Gordon Brown introduced the working families tax credit (WFTC). This was the first time that a benefit had been administered through the tax system and paid through the pay packet. It was a more modest integration than previous governments had envisaged and was inspired by Bill Clinton's earned income tax credit for low-wage earners. The WFTC was a benefit for people in low-paid work who had children. It guaranteed a weekly income floor of ?200 for families with a full-time earner. Next, in 1999, the childcare tax credit (CCTC) was introduced, payable to those eligible for the WFTC who also incurred childcare costs. It was worth 70 per cent of childcare costs up to ?150 a week. Then, in 2001, the children's tax credit (CTC), worth up to ?442 a year per child, replaced the married couples allowance.

Barbara Wootton once described complexity as the sixth giant evil (after idleness, squalor, want, disease and ignorance). The unintended consequence of the fiscal welfare state has been that the tax system has been made ever more difficult to navigate. Hence, no sooner have the tax credits been invented than they have been merged. This month the three tax credits have been reduced to two: the WFTC and the CTC have been melded into the child tax credit. The CCTC has been folded into a new working tax credit (WTC) which extends the principle of the WFTC to the childless. (Some analysts fear that tax credits have created a "purse to wallet" effect-with more benefit entering the home via the man rather than the woman. The WTC will usually go to the man, but the new child tax credit will go to the main carer.)

SOME REAL LIFE WELFARE

It is easier to pick a way through this thicket with some (imaginary) real-life examples. Imagine Sidney works full-time and earns ?200 a week, while Beatrice cares for their two children and earns ?100 a week working part-time. They pay ?60 a week for childcare.

If Sidney and Beatrice earned, between them, only ?97 a week, they would be entitled to the full amount of combined tax credits. They would receive the maximum ?177.85 each week. Given that they earn, in fact, ?300 a week, they are ?203 a week over the threshold. For every pound over the threshold, their credit entitlement is reduced by 37p. This is a deduction of ?75.11 from the maximum entitlement of ?177.85, leaving ?102.74 in tax credits-comprised of ?65.95 in child tax credit and ?36.79 in the childcare element of the WTC (that does not include their ?26.05 in universal child benefit).

Remember that dual goal of making work always more profitable than benefits and tackling child poverty. Consider Barbara and Ted who do not work. They have two children and their total allowance from income support, including extra for their children, comes to ?160 a week. If Ted then gets a job at which he earns only ?97, his total tax credit support rises to ?175 a week, making his weekly income ?272. Barbara and Ted receive more from the state when Ted gets a job than they did when both of them were unemployed.

That is not to say that the poverty trap has disappeared. It hasn't. It has been pushed further up the income scale-more of a "middle-income trap." Ed and Yvette have two children and a combined income of ?250 a week. Their child tax and childcare tax credits give them a further ?90 a week, bringing them up to ?340 a week. Ed is then given a 20 per cent pay rise which takes their combined pay up to ?300 a week. For every one of those extra fifty pounds he earns they lose 37p of their tax credit entitlement at the same time as they are liable for a higher tax bill. After his pay rise, Ed and Yvette receive ?71.50 in tax credits which brings their income up to ?371.50 a week. When the extra tax they pay is taken into account the increase in their net income is only 6 per cent a week. Their combined income has, though, gone up, which it would probably not have done under the old system.

Perhaps pushing the trap up the scale makes social policy sense. Better a trap high up the scale, where the marginal impact of every extra pound is less than at the bottom where every pound really counts. At the bottom end, the minimum wage (now ?4.50 an hour) and the new in-work tax credits have put down a floor that ensures that work will always be more profitable than benefits. (Though there are fears that unscrupulous employers may exploit the credits.) But if this floor is the new "right," it comes attached to a new "responsibility" to seek work. This pairing has been given institutional expression: the department of social security has been renamed the department for work and pensions to symbolise the change. The Benefits Agency, which pays out social security benefits, and the Employment Service, where the jobless go to search for vacancies, are being merged into the JobCentre Plus.

Work-friendly benefits and credits may have contributed to keeping the jobless figure down even now the economy is not so buoyant. The other major government effort on joblessness is the New Deal, introduced in 1998 and financed out of the windfall tax levied on the privatised utilities. The scheme was at first directed at young people who had been unemployed for six months and provides them with one of four options: a job, training place, voluntary work or a place on an environmental task force. The failure to take one of these options leads to a deduction in benefit.

It is hard to say how many of the jobs created by the New Deal-which in total has cost more than ?1.7 billion-would have been created anyway. It is, however, true that long-term youth unemployment is now at the historically low level of 36,000. (Although a worrying 1.5m people of working age are neither in work, education or receiving benefit; some are likely to be criminalised young people.)

The New Deal is significant for another reason. It has introduced a second element of "conditionality" into the means-tested welfare system-benefits depend not only on your means but also your behaviour. Hitherto the political left has taken the TH Marshall view that benefits are a right of citizenship. Certainly, this was the Labour party's response to the Conservatives' Restart programme in 1986 and to the obligations to seek work introduced in 1996 under the jobseeker's allowance regulations. These obligations now look timid next to those imposed as a requirement of the New Deal. And there may be more such conditions ahead. The idea of withdrawing child benefit from the parents of truants seems to have been abandoned, but there is some sympathy within government for the principle. Reducing housing benefit to rowdy tenants is also being actively discussed. Perhaps progressive universalism means, amongst other things, that the wider community cannot be expected to finance the welfare state if they disapprove of the behaviour of those who draw most from it. It is too early to say that the "good behaviour" welfare state is emerging, but this is a space worth watching.

TAX CREDITS: THE BALANCE SHEETS

For all the wizardry of the system, the examples above show the tax credit to be, in some respects, a familiar friend. Most credits are administered through the pay packet, but the test of means remains. Indeed, it may be that the means test is noticed but the credit is seen as an artefact of the tax system. At least the old means-tested benefits involved having to go somewhere to collect it. The tax credit slips silently into a bank account and there are fears in government that Labour will receive little credit for its credits.

Few people today would use the 1930s language of Aneurin Bevan but there are still good reasons, apart from stigma, for reservations about the means test. We have seen already that the system is highly complex. The different thresholds for eligibility mean that claiming a credit is not as easy as one might think. Even the simplified child tax credit form is 12 pages long. This is one of the reasons take-up is so low-indeed, take-up appears to be falling and last year nearly ?5 billion in benefits was unclaimed. Three years after it was introduced, only 62 per cent of those eligible claim the WFTC. The latest figures suggest that nearly a third of those pensioners entitled to claim the minimum income guarantee (MIG) fail to do so.

Means tests are also expensive to administer. The administrative cost of income support is 11 per cent of total expenditure (which is ?16 billion) and of housing benefit 15 per cent, compared to 4 per cent for the universal child benefit. The complexity of the system also contributes to a high error count and encourages fraud. In 2002, the department of work and pensions estimated that ?1.15 billion was lost due to fraudulent claims of jobseeker's allowance and income support. It is estimated that 5.4 per cent of income support claimants and 8.6 per cent of those claiming jobseeker's allowance do so fraudulently. Only 1.7 per cent of the child benefit caseload is thought to be fraudulent.

Also, the modern means test has not only not abolished the poverty trap, it has made another trap, the savings trap, even worse. If Lucy retires and her only source of income is the basic pension of ?77 a week, she will then be eligible for the income support top-up, the minimum income guarantee. This takes her weekly income up to ?100 a week. Liam, by contrast, has done as the government wants and saved for a small private or occupational pension, from which he receives an extra ?25 a week. This takes him to ?102 a week. For all the saving he has done, he receives the princely sum of ?2 a week more than Lucy.

The government is now responding to this by introducing, in October, a new pension credit. Amira receives a basic state pension of ?77 a week. She has ?23 a week which comes from her occupational pension. Her combined income of ?100 is supplemented with ?13.80 a week from the pension credit. This will benefit single pensioners with income up to ?135 a week and couples with income up to ?200 a week. It is calculated by crediting Amira with an extra 60p in pension credit for every pound of her own pension money.

The tax credit regime has softened the trap problem. But traps are inevitable in a means-tested system. All that an ingenious chancellor can do is to play with the rate that benefits are withdrawn when someone gains extra income. If income is withdrawn too fast it creates a disincentive to earn more. The loss of credit makes the worker no better off and fearful of the transitional period between the time of the final payment of the credit and the first pay cheque. If it is withdrawn too slowly the cost to the state begins to escalate.

The full effect on incentives and public spending of these new tests of means are not yet fully apparent. The absolute amount spent on means-tested benefits actually fell between 1997 and 2001 by ?4 billion (from ?36 billion to ?32 billion), mainly because unemployment fell from 1.6m to 1m over the same period. Neither have means-tested benefits yet increased as a proportion of the whole of spending on welfare. In 1997/8, means-tested benefits accounted for 34 per cent of the total spent. By 2001, they accounted for 33 per cent. Projections from the Institute for Fiscal Studies (IFS) suggest that they will grow gently, to 35 per cent of all benefits in 2003/4 and 36 per cent in 2005/6. The absolute numbers of people on means- tested benefits has, though, already started to rise. The IFS has calculated that, whereas 17 per cent of all households were entitled to some means-tested support in 1997, 29 per cent of households will be so entitled in 2003.

So, the government has increased the scope of the means test and it has increased the value of universal benefits paid to everyone in a certain category, such as mothers or retired people. The main universal benefits are the basic retirement pension and child benefit. In 1997, the first child was entitled to ?11.05 and each subsequent child received a further ?9. In 2003, the first child now receives ?15.75 and each subsequent child ?10.55. Universal benefits will grow from 18 per cent of all benefits in 1996 to 20 per cent of all benefits by 2005.

Contributory benefits, above all the state pension, have continued their secular decline. Incapacity benefits have been reduced, maternity allowances have been extended to mothers without a complete contributions record and the link between entitlement and contributions has been weakened with respect to the second state pension. The minimum income guarantee (MIG) is a re-titling of income support for the elderly which has targeted help on the less well-off instead of the universal state pension. By 1997, only 47 per cent of all benefits were contributory. Under this government that decline has continued and contributory benefits now comprise 45 per cent of the total, (compared with about 35 per cent means-tested and 20 per cent universal).

PAST AND FUTURE

How new is all this? Both the 1964-1966 and the 1974-1979 Labour administrations increased the proportion of benefits that were means tested; rate and rent rebates being Labour's means tests of choice in 1967. Those two administrations did nothing very much to the overall amount paid in universal benefits, so the increases in the value of universal benefits since 1997 represent a modest improvement on its Labour predecessors.

But even if the numbers are not drastically different from earlier Labour or Tory administrations, this is the first Labour administration that has actively embraced the means test as a tool of progressive social policy. And means testing has turned out to be a rather different tool in the hands of a Labour chancellor than it was in the hands of his Conservative predecessors. The Thatcher governments coupled means tests with a refusal to increase the value of universal benefits even to keep pace with inflation. The objective was to reduce cost, to create what Richard Titmuss once called a residual welfare state. In this conception, only the poorest are assisted by the state. The majority middle class is expected to secure private solutions to its needs. Hence, the point of the means test is to set the safety net at a relatively low level. Contrary to this, it was once believed on the left that the middle class would only continue to bankroll the welfare state if they themselves saw a decent return from it. This idea has not exactly been abandoned-rather New Labour has combined it with its opposite: the idea that the middle class will now continue to pay only if they do not benefit too much themselves and see the money going to those most in need. Again, not either/or but both.

To set the safety net at a low level has never been Gordon Brown's intention or his effect. New Labour redistribution through the benefits system is a story that is only now beginning to be told. Since 1997, means-tested income support has increased by 33 per cent in real terms for lone parents with two young children, by 44 per cent for low-income working lone parents with young children, by 31 per cent for single pensioners below 75 and by 25 per cent for couple pensioners over 75. Between 1996/7 and 2003/4, spending on children will increase by almost 40 per cent. In his William Beveridge commemoration lecture of March 1999, the prime minister famously pledged himself to eradicate child poverty in 20 years, halve it in ten and cut it by a quarter by 2004. These targets are extremely unlikely to be met. The government managed only half of the 1m children it promised to take out of poverty during its first term. The number of children living below the poverty line fell from 4.4m in 1996/97 to 3.9m in 2000/01 with a further fall to just 3.8m in 2001/02.

But the rising inequality of the 1980s and 1990s has been at least slowed if not reversed. Because of the way the poverty threshold is calculated, at 60 per cent or less of median income, the fact that the floor is rising does not necessarily make a great deal of difference to the number below the threshold. Even a shakeout of bonus-rich top earners in the financial sector, as we have seen recently, may make little difference to the poverty measure-it all depends on how it affects the bunching around the median point. But these are just statistical artefacts. The truth is that incomes for most people at the very bottom of the pile have risen quite sharply since 1997.

So, is there something emerging in social policy which deserves the title progressive universalism? The stress on work and families with children is well known. So too is the tax credit as the new version of targeting-"civilised selectivity" as Roy Jenkins once put it. We are a long way yet from complete integration of the tax and benefit system. But a start has been made. And overall social policy has certainly been progressive.

But there is also something odd going on. The New Labour rhetoric of "rights and responsibilities" and "the end to the something for nothing state" slots in rather neatly with the idea of insurance-type contributory welfare. But in policy there has been no sign of this linkage. It seems that when Frank Field went, the contributory principle and opposition to the means test went with him.

Perhaps a distinctive conception of a conditional welfare state is emerging. No longer a method for allocating resources tied to unconditional entitlement, but instead a corrective against capitalism, a huge leviathan for the mitigation, indeed eradication, of poverty. No longer grants as of citizenship right, but conditional: today on income, tomorrow, perhaps increasingly, on behaviour too.

Benefits and tax credits

CONTRIBUTORY
Retirement pension; Bereavement allowance; Jobseeker's allowance; Statutory sick pay; Incapacity benefit; Statutory maternity pay
MEANS-TESTED
Jobseeker's allowance*; Minimum income guarantee; Income support; Council tax benefit; Housing benefit; Social fund; Working tax credit; Child tax credit
UNIVERSAL
Child benefit; Retirement pension*; Industrial injuries; Disablement benefit; Christmas bonus; Disability living allowance; Severe disablement allowance; Maternity allowance; Winter fuel payments

*Appears under two headings as it contains elements of both

Tax credits from April 2003

WORKING TAX CREDIT (WTC) The following people are eligible:
Couples who earn below ?280 a week
Single people who earn less than ?200 a week
Those over 25 without children working a minimum of 30 hours
Those with children who work a minimum of 16 hours
Disabled people who work a minimum of 16 hours a week
THE WTC CONSISTS OF:
A maximum of ?29.20 for all adults who have an income up to the threshold of ?5,060 per annum
A maximum of ?11.90 for families working at least 30 hours
A couple and lone parent element up to a maximum of ?28.80

CHILD TAX CREDIT (CTC) The following people are eligible:
Couples or lone parents with responsibility for a child under 16 or under 19 and in further or higher education
Recipients of income support or jobseeker's allowance who have children will automatically receive the maximum entitlement
THE CTC CONSISTS OF:
?10.45 per week for families earning up to ?50,000 a year. It is gradually withdrawn by 37p for every pound above a threshold. In the first year of the child's birth this amount is doubled
?27.75 per child per week for families earning up to ?13,000.
It is withdrawn by 37p for every pound above a threshold
Taken together with universal child benefit, the CTC will guarantee support for the first child of ?26.50 a week for the 85 per cent of families with an income of less than ?50,000 per annum
The 25 per cent of families with an income of less than ?13,000 per annum will be guaranteed ?54.25 a week