So many of the poorest people are without work that a minimum wage - one of the Labour Party's concrete policy commitments - would do very little to relieve poverty.
What is more, even if it were set at a level which did not destroy jobs initially, uprating the minimum wage in line with the typical level of earnings could raise unemployment in future.
Increasing inequality has boosted support for the introduction of a legal floor for wages, but a paper published today by the Institute for Fiscal Studies, a respected independent think-tank, suggests that it is not a useful weapon for tackling unfairness in incomes.
Amanda Gosling, author of the paper, concludes: "Because most of the poor are not in work, a minimum wage is not a good way to redistribute income from the rich to the poor. Most of those who gain will be couples where both partners work and young single people living with their parents. These people are richer than the unemployed or many pensioners."
However, a minimum wage would benefit twice as many women as men, as low pay is a predominantly female phenomenon.
The incomes of the lowest paid in the economy have fallen further and further behind during the past two decades. The number of families on incomes below 60 per cent of the average doubled between 1977 and 1991.
Supporters of the minimum wage have argued that it would help to reduce income inequality, although debate about whether or not it would cost jobs has been fierce. Ms Gosling says there is evidence that a legal minimum would not raise unemployment within a certain range of wages.
She calculates the effect on income of various levels of minimum wage, assuming that it does not raise joblessness. Significant numbers of people on low pay would see their incomes rise. A minimum wage of pounds 3 would boost pay for 7 per cent of the workforce, while a pounds 4.50 minimum would help about 30 per cent - 1.8 million people and 7.8 million people respectively.
A higher proportion of the poorest families where someone has a job would gain than of the better-off families.
A pounds 3.50 minimum would help two-fifths of the bottom 10 per cent of the earnings distribution compared to one-fifth of those in the middle. Yet nearly a tenth of the richest 10 per cent of families would also benefit.
Looking at total household income, which includes those with no work, there would be no redistribution.
Even a pounds 3 minimum wage benefits those at the top of the income distribution far more than those at the bottom.
Ms Gosling also warns that if the minimum wage is tied by some formula to average earnings, then a decline in the productivity of the low-paid relative to the average would be likely to cause unemployment.
What the policy would do, by replacing in-work benefits such as family credit, is stop subsidising exploitative employers.
Scrapping benefit to boost education
Children from poor backgrounds could be encouraged to stay on at school if they were paid an allowance from the age of 16, the institute's report also suggested.
Abolishing child benefit for 16- to 18-year-olds and paying them the pounds 10.80 a week allowance instead would be a relatively cheap initiative, it said. Such a move - based on a scheme introduced in Australia a decade ago - would not make poorer families worse off because their income support would be increased, it was argued. The scheme - costing pounds 170m and affecting a million 16-18-year-olds - "should be given serious consideration if school participation is to be increased among children from poor backgrounds".Reuse content