£1.2 billion saved? The troubled families programme

"And I can announce today that almost all of the 117,000 families which the programme started working with have now been turned around — in terms of either school attendance or getting a job or both. This has saved as much as £1.2 billion in the process."—David Cameron, 22 June 2015

The last time the £1.2 billion figure turned up was in this press release, when the government had 'turned around' 106,000 families.

While the government have published updated figures for the number of families 'turned around', the £1.2 billion figure seems to be the same one we saw before.

We don't know how much the programme is saving

Spending on the families helped by the programme fell after intervention. But we don't know that spending fell because of the intervention.

Some of these families might have changed their behaviour without taking part in the troubled families programme, and until we know how many families the programme actually caused to change their behaviour, we won't know how much it saved.

A fuller evaluation is currently underway. In the meantime, £1.2 billion is not a credible estimate of the savings of this programme.

There are more problems with this figure

The £1.2 billion figure doesn't count the cost of the intervention, and it's based on a sample of families from just seven of the 152 local authorities taking part in the scheme.

Most of the seven areas saved about £6,000 to 10,000 per family, but Salford saved £18,000 per family while Staffordshire saved £49,000 per family in the first year of the programme.

If the other 145 local authorities are more like the £6,000 or £10,000 regions than Salford or Staffordshire, then the total savings will be much lower than the £1.2 billion figure.

The department says that it thinks the estimate for savings is on the low side. Not all of the savings made are included (for example, they don't have an agreed cost for police call outs), and the councils reporting back didn't collect information on all the services where they could make savings.

Some areas actually ended up spending more per family on things like social services and housing, so some of the services that they didn't have information on might have seen an increase in spending.