Workless household statistics: The Mail gets it wrong

New figures released by the Office of National Statistics (ONS) have been seized on by several newspapers as evidence of a rise in worklessness among the UK's "Shameless generation".

The ONS report on work and worklessness by household shows that 3.9 million UK households, or 19.2 per cent of all households include no working adults - a rise of 0.6 percentage points on 2009.

But have newspapers been reporting the figure accurately?

The Claim

This morning's Daily Mail contained a striking claim, apparently based on the official figures, that "the number of workless households has soared by 22 per cent, with 700,000 families joining the total".

Our curiousity about the claim only increased after checking the ONS bulletin behind the story which shows a much smaller rise of 212,000, or 5.6 per cent.

Analysis

Full Fact contacted the story's author Jason Groves to ask how he had arrived at the figures used in his report. He explained that rather than using Table 1.1 of yesterday's release to get a figure for workless households, he had taken figures from an earlier ONS briefing.

Sure enough, the 2003 release to which we were referred includes figures for workless households in 1998 that were roughly 500,000 lower than the level for 1998 reported in yesterday's bulletin — thus suggesting a rise of 700,000 over the period.

We contacted an ONS spokesperson to get to the bottom of this discrepancy, and were directed to a section in yesterday's report drawing attention to "major changes to the way the estimates have been complied in this release since its publication in 2009". The Mail had arrived at erroneous figures by comparing two incompatible data sets.

Mr Groves has assured Full Fact that in light of the error he has requested corrections to the online version of the story. Although statistics showing such a dramatic rise in workless households should have triggered caution, the Mail appears to have been honestly mistaken rather than intentionally misleading, and such willingness to fix inaccuracies is encouraging.

Nevertheless, whatever changes are made to the online version, the print paper carries the invalid comparison.

Furthermore, this was not the only dubious figure contained in the report. The headline of the article asserts that 7 million people live in jobless households, but the first line then claims that this constitutes "a fifth of all Britons".

Given the UK's population of over 60 million, these two statistics clearly fail to add up.

The ONS figures do back up the claim that "the percentage of households where no adults work was 19.2 per cent". However this does not mean that those living in these households account for an equivalent proportion of the British population.

Mr Groves explained that this error had surfaced in changes made by subeditors after the submission of his original article.

Other Newspapers

The Telegraph appears to fall prey to the same confusion between percentage of workless households and percentage of population who live in workless households when it claims that "there were 3.9m households in Britain where no one was working — 19.2 per cent of the total population."

On the other hand, The Sun's claim that the number of children living in workless households has risen by 100,000 since 2007 is borne out by the official figures.

Yet the claim is slightly disingenuous given that the 2010 figure has fallen since 2009: the paper's suggestion that the data presents an upward trend is not seen in the figures.

Conclusion

In their attempt to condense yesterday's figures into a story about soaring unemployment and mass reliance on benefits, these papers have either misused the data, or used it selectively to imply more defined and uniform tendencies than the official statistics actually show.

At the time of writing the Mail's report has not been corrected, and we will watch to see when it is, given the inaccuracies the article contains.

But even with these amendments in place, the inaccurate impression these articles could create though their use of statistics cannot so easily be undone.

Edgar Gerrard Hughes