So who are ?
Figures out today from the have been seized on by the Conservatives.
The shadow work and pensions secretary, Theresa May says they are "scandalous".
They show that the number of people of working age who live in homes where no one has a job has gone up by half a million in the past year. That sounds like a sharp rise and takes the total to 4.8 million.
And for all the people who have lost their jobs, leaving their household without income from work, the effects can of course be devastating.
The figures also show a rise in the number of children living in households where there is no earner - up 170,000 to 1.9 million - an increase which puts the government's target of cutting child poverty further out of reach.
For any government, a rise in the number of people not earning, not contributing tax to the Treasury's coffers, and perhaps requiring financial support from the state is a concern - especially at a time when the screw on public spending is tightening.
But what might surprise you is that while unemployment has been growing fast, the figures show that the increase in the rate of workless households (the proportion of homes in which no adults work) is not in fact so rapid - it's increased by only 1.1 percentage points in the last twelve months to 3.3 million.
That may be the highest year on year increase since 1997, but 1.1% doesn't sound like a lot.
So what does that really tell us? Is this a rise due to recession or a more long term problem?
Well, one social policy expert suggested to me is that what is striking about the figures is how high they were before the recession really began to bite - even this time last year, in roughly one in six homes no one had a job.
So given that, it's hard for the government to blame the number of homes where no one has a job simply on the results of the recession.
But what is a realistic level to expect? It may help to take a closer look at which households are counted in the "workless" totals?
For instance, they don't include households where the occupants are only pensioners. Nor do they include homes where people don't work but live with someone that does, for example a mother or father who stays at home and cares for children while their partner goes out to work.
But the figures do include people who care for a relative at home full time in a household with no other earner, and single parents who don't work (although there's been an increase of more than 10% in the rate of single parents going out to work since 1997).
They also include people who can't work because of disability, people who are recuperating after illness and those who have been lucky enough to be able to choose to retire early.
So, taking that into account, it is clear that there are always going to be households that qualify as workless. But experts suggest that if everyone who could work was working the remaining households might only represent about one in 10, significantly fewer than the current level even before the recession.
Out of the three million or so people claiming out of work benefits on the grounds of disability, research has suggested that at least third of them could do some form of work, and want to.
The employment minister Jim Knight says the government has made real progress in tackling worklessness and points out that there are 2.5 million more people in work than in 1997.
Yet today's figures show that major obstacles clearly still exist for any future government to tackle, if and when the current downturn ends.
PS Theresa May will be attacking the government on its welfare record in a speech tomorrow. But a word of caution - the research the Conservatives are circulating ahead of the speech uses figures from the 2001 census.
Of course she may well still make a compelling argument, but the figures she's using are eight years old. No political party though can quibble with today's stats from the ONS.