The government is very happy that its "troubled families" intervention programme is having results, with the BBC reporting that :
Early intervention by a dedicated case worker has reduced crime among those people involved by 45%… Anti-social behaviour has gone down by 59%.
Those are good results, even if the vast majority of the report is case studies of a few of the families involved. As for the actual results, the vast majority of quantitative data presented is percentage changes. This is clearly important; but it's also crucial to know how many troubled families actually exist, and how many can be helped. After all, a programme which is targeted at just a handful of families isn't particularly useful in the grand scheme of things.
This is an area the government, and the BBC, fall down on severely. The Department for Communities and Local Government's report  claims, three times, that there are 120,000 families.
This is incredibly unlikely to be true. We've explained before, in detail, why this is the case , but the short version is that the DCLG claimed there were 120,000 troubled families defined with one set of criteria, but then changed the definition and continued claiming 120,000 families existed.
Unless two markedly different groups of people both add up to 120,000, it seems likely that this number was just pulled out of thin air (none of the research which the DCLG has made available explains where it came from). And yet today's report, and the BBC write-up, repeats it.
The BBC also claims that 40,000 families are expected to be helped this year, which would be a twelve-fold increase from the 3,324 families who were actually helped in 2011-2012 (and, of course, would still be just a third of the claimed eligibility). That figure of 3,324 is not mentioned anywhere in the BBC's report, nor the DCLG's press-release .
The trouble families programme does seem to be a great help to those families successfully referred to it, as Casey's report makes clear. But it is helping far, far fewer families than media reports make out; and part of that may be because no-one seems to actually know how many families are even eligible.