Troubled Families estimates: a troubled history
We said yesterday there seemed to be confusion in the media about the source of an estimate that 'troubled families' are costing the taxpayer £30 billion a year, with some suggestions that it came from the government.
It's actually been made by the Sunday Times following an interview over the weekend with Head of the programme Louise Casey.
The figure was reported as the estimated cost of 520,000 'troubled families' identified by the government. This includes an initial estimate of 120,000 families who were thought to cost in the region of £9 billion a year, and the cost of an additional 400,000 families identified by the government in June last year and to which its Troubled Families programme is being expanded.
The Sunday Times told us it's based on a working assumption of the second tier of troubled families costing about two thirds of the first tier cost (£75,000), informed by guidance given by professionals, government sources and Freedom of Information responses.
Questions around the £9 billion estimate
The original estimate of 120,000 troubled families comes from a Cabinet Office report from the now-defunct Social Exclusion Task Force, which explains that "approximately 140,000 families with children experience five or more disadvantages". A Department for Education (DfE) FoI disclosure explains that the 117,000 figure (rounded to 120,000) is an estimate of how many of the 140,000 families in the UK as a whole are based in England.
When Full Fact first looked into the original estimate of the £9 billion cost (£75,000 per family), the Department for Communities and Local Government (DCLG) confirmed it was based on a study by the DfE. The study looked at the burden placed on taxpayers by a 46,000 subset of the 120,000 families with multiple problems, who also had 'problem' child behaviour. The estimated cost of these 46,000 families was £4 billion.
Since that article, the department has published a more detailed explanation of their calculations. But it acknowledged that the priority for the estimate was getting a sense of scale rather than an exact costing:
"While the figures in this analysis are significant and informed government's decision-making process, the critical point for the Government was not necessarily the precise figure, but whether a sufficiently compelling case for a new approach was made. In this context, the indicative distribution of reactive to targeted spend was as important as the total figures estimated."
The analysis included estimates by DfE, the Department for Work and Pensions, the Department of Health, the Ministry of Justice, the Home Office and DCLG of the portion of departmental expenditure going to the 120,000 troubled families (directly or indirectly).
To qualify for assistance from the extended programme, a family must now have two out of six problems to be deemed a 'troubled family'. That's a shift from the previous requirements under the first wave where families qualified if they had three out of three issues, or two out of three issues and were considered a "cause for concern" by the local authority. Neither of these are the original five or more problems used to make the initial 120,000 estimate so whether the £9 billion estimate should be used to describe the cost of the families that are actually being helped through the programme remains an unanswered question.
We also don't know the extent to which this is an additional burden placed upon the taxpayer by 'troubled families' in particular. In other words, even if they were costing £9 billion this doesn't necessarily mean £9 billion more than an average family.
We know even less about the new families to be helped under the scheme
We don't know that much about the additional 400,000 families to which the programme is being extended. Previously they were said to have fewer problems: when giving evidence to the House of Commons Communities and Local Government Committee the then Secretary of state for DCLG, Eric Pickles, described this group as "not as acute as the 120[000]". But in a press release this morning, DCLG described the expanded group as having a "wider set of problems including domestic violence, debt and children at risk of being taken into care". We don't know whether this means that each family's problems are more or less acute than those in the first group.
Based on guidance given to it by professionals and government sources, the Sunday Times told us its £30 billion estimate is based on an assumption that each of the additional 400,000 families will cost about two-thirds of what the initial 120,000 families cost - working out at about £50,000 each (rather than £75,000). It told us this assumption was an underestimate if anything as it could well be that these families have the same amount of problems as the first.
We're not aware of any evidence published by the government which could provide sufficient detail to this assumption. Mr Stephen Crossley, one of our readers who is a phD student exploring the Troubled Families Programme, pointed out to us that he's tried to get more detail of this additional group through FoIs but with no success so far.
Scale of the problem for the initial 120,000 could be bigger than estimated
The Sunday Times told us a report released in July on the scale of the problems faced by these families, and its interview with Louise Casey, showed how families identified so far have had more than the three serious problems first needed to qualify for the programme.
"On average, Casey said, the families she tries to help have nine significant problems, most of which are linked".
On this basis it stressed how uncertain the figures are and noted that its analysis, although grounded by evidence it has gathered, comes with a very strong caveat that it is an estimate and only with more evidence can we say whether it's close to the actual figure.