Means tests do help the poor but the Chancellor has still to sort out the snags. By Donald Hirsch

There are two striking features of Labour's social policy that few would have predicted in 1997. The first is that Gordon Brown has played the role of Robin Hood. The second, even more surprising, is that the most direct criticism has come not from rich barons who want to keep their wealth but from the self-appointed champions of the poor.

Brown's budgets have decisively redistributed income to poorer families, not by increasing personal taxation on the rich, but by allocating a disproportionate share of the Treasury-controlled fruits of economic growth to people on the lowest incomes. This has not caused a big change in the distribution of incomes, given that the market-controlled fruits of growth (such as higher earnings) tend to go disproportionately to the better-off. But at the very least, the net effect of Gordon Brown is lower inequality than could have been envisaged under any other Chancellor.

Brown has achieved this largely through means tests. The poorest pensioners' incomes have been lifted through a generous Minimum Income Guarantee. Out-of-work parents have had their child allowances raised sharply. And for people who work but still have low family income, the tax credit system provides a new pillar of means-tested support.

The government has just finished consulting on another piece in this jigsaw, the Employment Tax Credit, due to be introduced in 2003. This credit will, for the first time, give support to all the full-time working poor who fall below a minimum income threshold (as long as they are over 25) and not just those who have children or a disability.

But - and this is what troubles the champions of the poor - means tests have big disadvantages. They are unpleasant to claim: in effect, you have to go to the government and say you are poor. And they tend to create poverty traps: it becomes difficult to rise much above the guaranteed income floor because support is withdrawn as incomes rise. The reply to these long-established arguments is that means testing may be the best policy on offer for the poorest people. It is simply unrealistic, in this day and age, to expect a high basic state pension for all old people and complete job security or high wages for all working people.

But there are several reasons why ministers should think twice about relying excessively on means tests. First, topping up low earnings threatens to entrench low pay. Why should an employer raise wages if workers get them topped up anyway, and lose the top-up as earnings rise? This effect is not inevitable, particularly in a system such as the Employment Tax Credit, where the top-up depends on family income rather than directly on individual wages.

But the government needs to be careful that it does not use means tests as substitutes for an adequate minimum wage - especially as the current minimum wage rate has not endangered jobs. In this year's budget, the Chancellor talked of a "minimum rate" for people on Working Families Tax Credit of £6.40 an hour. It is disturbing if ministers consider topped-up minimum incomes to be the equivalent of minimum wages. They are not. Higher wages increase the net reward for working extra hours; means tests reduce it.

A second, related risk is that an in-work income floor will slow progress towards one of new Labour's cherished objectives - an "opportunity culture". The Employment Tax Credit will increase incentives for people to get a job, but reduce incentives for them to better themselves thereafter. Some people, as they earn more, will lose more than 90 per cent of the extra income through tax and withdrawal of housing benefits (although, in most cases, it will be much lower than 90 per cent). While ministers preach about improving individual opportunities and productivity, they risk sending signals that encourage individuals and companies to go for the low-grade, low- paid jobs that have been so plentiful in the United States in recent years.

Finally, the decision to base entitlement to tax credits on the income of the "family" unit (a single adult or a couple, with or without children) seems to contradict the principle of independent taxation for men and women established a decade ago. For example, a couple with only one person working may receive the credit, but as soon as the partner does a few hours of work, the credit will be cut. This may reduce the incentive for some women to work at all, even when they have no children or other pressing responsibilities at home.

The Treasury is aware of these pitfalls, and has tried in some cases to address them. For example, it plans to use a new form of income assessment so that, where someone starts work part-time and earns below the tax threshold, his or her partner loses only 37p in the pound from the Employment Tax Credit. Yet it will be fiendishly difficult to balance priorities. Can the government provide adequate protection from working poverty without creating new poverty traps? Will its emphasis on incentives to enter work cause it to neglect the quality of people's jobs and of their working lives? Gordon Brown and his team are still a long way from finding the answers.

The Employment Tax Credit and Issues for the Future of In-Work Support by Fran Bennett and Donald Hirsch, is published by York Publishing Services, 64 Hallfield Road, Layerthorpe, York YO31 7ZQ (£9.95 plus £2 p&p)