Robert Chote: Will the Government have the means to make child poverty history in Britain?

Tax credits have been Gordon Brown's big idea for the British welfare state. He has used them to deliver unprecedented sums of money to poorer families and begin to reverse the steep rise in child poverty seen during the 1980s.

Tax credits have been Gordon Brown's big idea for the British welfare state. He has used them to deliver unprecedented sums of money to poorer families and begin to reverse the steep rise in child poverty seen during the 1980s.

But the Chancellor, like all welfare reformers, must struggle with the "iron triangle" - the conflict between reducing poverty, encouraging work and keeping down costs. With the Government having set itself increasingly challenging targets for child poverty right out to 2020, these conflicts can only intensify.

Since 2003, the system has had two main elements. The child tax credit is paid to the main carer in a family, whether she or he is working or not. The working tax credit then tops up low wages for people doing a set number of hours a week, whether they are parents or not. The two credits, aggressively means-tested to focus help on the poorest, paid out £13.5bn in 2003-04.

Total financial support aimed at families with children has doubled in real terms and risen significantly as a share of national income since the 1980s. But how has the Government secured political support for this remarkable generosity?

Lower unemployment has helped, containing other welfare costs. And, despite the means-testing, the Government has also ensured that most families are entitled to at least some payment. This gives them a stake in the system as a whole.

Another critical decision was to use the tax system rather than raising "family credit", the in-work benefit inherited from the Conservatives. In part, this reflected the belief that tax credits carry less social stigma than means-tested benefits and that poorer families were therefore more likely to take up their entitlement. A similar argument was made initially for paying the credits through the wage packet.

An analysis by Stuart Adam and Mike Brewer at the Institute for Fiscal Studies (IFS) - published by HM Revenue and Customs last week - shows take-up of the tax credits by lone parents was low when first introduced in 1999, but that they were at or above the level that would have been expected for family-credit payments of equal generosity by 2002.

There is no official estimate yet for take-up of the new credits introduced in 2003. But Kate Green, the chief executive of the Child Poverty Action Group, told a Nuffield Foundation seminar last week: "There is cause for optimism that it will be quite high, despite their incredible complexity." It certainly looks higher for the child tax credit than the working one.

Another reason the Government chose to go down the tax-credit route is that it thought voters would be happier to pay for a form of "tax cut" than an increase in benefits. That may well be so, but the way this has been implemented since 2003 has also created big practical problems for recipients, advisers and administrators.

Most of the difficulties stem from the tax system being based on annual assessments of income. This is not a problem for employees on PAYE or for people who can budget relatively easily for year-to-year fluctuations in their tax bill as it catches up with changes in income.

However, to respond flexibly to changes in circumstances, tax-credit payments are now provisionally based on the previous year's income and then finalised against the current year's. Even in the absence of errors by the claimant or the taxman, this means that over- and under-payments are built into the system. For people with little access to savings or credit to smooth their spending, recovering over-payments (sometimes within the year) can cause real hardship and uncertainty in a whole variety of situations.

Rosina Pullman, the director of TaxAid, an advice charity, points out that "Administrative mistakes may disproportionately affect those on low incomes who are more likely to be in and out of employment, self-employment, or benefits. Like anyone else, they may also leave or join relationships, have children and use formal childcare - not to mention older children who start and leave school. They visit their parents in far-off lands, rent out rooms, give to charitable appeals and look after their children over the holidays. All may affect their eligibility for tax credits or the amount of their award, requiring successive claims."

It does not help that income is defined differently for income tax and tax-credit purposes - and that the former is assessed for individuals and the latter for families, requiring duplication of information.

Of 5.7 million tax-credit awards in 2003-04, 1.9 million were overpaid and 0.7 million underpaid. Overpayments totalled £1.9bn, almost 15 per cent of the tax-credit bill. It is important to remember that these reflect improvements in claimants' circumstances as well as errors, but retrieving the money may still cause hardship, disruption and uncertainty.

Dawn Primarolo, the Paymaster General, has signalled the Government's determination to tackle the problem. But, committed to accelerating the decline in child poverty over the next 15 years, it may come to face an important strategic choice. Will use of the tax system to channel money to poorer people be sustainable as the amounts involved rise - however useful it has been politically in recent years?

And don't forget the iron triangle. For a while, the Government was able to cut poverty and get more people into work simultaneously. Another IFS study, published last week, suggests that tax credits had boosted employment by around 80,000 by 2002 and cut the number of workless families by 100,000 (although other tax and benefit changes partially offset these gains).

But while recent increases in the child tax credit are well focused on the poorest families and cutting poverty, they impose higher effective marginal tax rates on growing numbers of people as they are withdrawn with rising income. This worsens work incentives - especially for parents in couples - and may tempt the Government to use more stick as well as bigger carrots to encourage people into work.

The Government has ambitious goals for cutting poverty further and faces big challenges in shaping the welfare system to deliver them, even if the economic environment stays benign. The going will get tougher, but few would have predicted the progress it has made so far.

John the Baptist moves in for Gordon Brown

John the Baptist's mission was to "prepare the way of the Lord and make his paths straight". Sir Gus O'Donnell, Permanent Secretary to the Treasury, must feel a similar responsibility after the announcement last week that he is to become Cabinet Secretary, Britain's top civil servant. A central task will be to ready Whitehall for Gordon Brown's widely predicted move from the Treasury to Number 10.

Mr Brown began his chancellorship relying on a small group of loyal advisers and distrustful of the civil service machine. When Sir Gus became Mr Brown's third Permanent Secretary in 2002, he knew officials had to adapt to the Chancellor's way of doing things, and they have established a good working relationship.

Some complain that Treasury officials have gone too far in embracing Mr Brown's modus operandi. Institutions with which the department interacts complain it is unnecessarily secretive, and that some civil servants now brief against individuals and ideas to which Mr Brown objects in a way now typical of explicitly political advisers. Within the department, most officials feel empowered working for a Chancellor who effortlessly dominates domestic policymaking, although some feel excluded from what they see as a laddish inner circle.

But it is tempting to hark back to a nonexistent golden age. Officials naturally absorb some of the mindset of their political masters, and drawing the line where effective support for a minister begins to compromise civil service impartiality is much easier to demand in principle than it is to apply in practice. In any event, there would have been no point choosing a Cabinet Secretary with whom Mr Brown could not work. Sir Gus meets the bill in this respect, beside his many other strengths.

He also has an important potential advantage in getting the government machine to operate more coherently. The relationship of near-equality between Messrs Blair and Brown has contributed to an unstable equilibrium between their camps of officials and advisers. This has even prompted some ill-advised proposals to emasculate the Treasury.

But if Mr Brown becomes Prime Minister, there will no longer be a partnership of near-equals in Downing Street. His Chancellor will presumably be a close, but obviously subservient, ally. Mr Brown can't emulate William Gladstone in taking both jobs himself, but it might well feel like that.

Robert Chote is director of the IFS

Hamish McRae is away