The figures have been a political battleground since the early years of Margaret Thatcher's government.
This year, David Mitchell, the social security minister, sought to blunt the impact of the figures by publishing the findings of new research showing that about half of those in the bottom tenth in 1991 had moved up the income scale three years later.
Ministers have previously suggested that, because the individuals who make up the poorest tenth are not the same ones from year to year, the misery of poverty is spread around. But the first evidence of "income mobility" failed to support this argument conclusively.
The figures confirmed that pensioners and lone parents at the bottom of the pile tend to be stuck there, according to Paul Johnson of the Institute for Fiscal Studies. "Single young people do move in and out, but they don't tend to climb out very far, usually only to the next groups up. They don't tend to climb to the top," he said.
The figures undermine the Prime Minister's rhetoric, repeated at the Tory conference in Bournemouth last month, about giving people at any level of society the opportunity to reach the top.
The data for the period from 1979 to 1993/94 show that the median income for the poorest tenth fell by 13 per cent after housing costs, and taking inflation into account.
This represented a small recovery from the previous year, when it had fallen by 17 per cent., partly because of unemployed people falling into arrears on their mortgages.
The median income of the richest tenth rose over the same period by 65 per cent.
Mr Johnson supported the government's argument that the estimates for the very poorest are "particularly uncertain". He said: "There are genuinely a lot of problems with that group." But, he said, the fact that the income of the poorest third of the population has not gone up at all in real terms, when the rest has gone up by about half, is "pretty dramatic".
t Households Below Average Income, 1979-1993/4, Stationery Office, pounds 30.