The background

The government launched the Work Programme in June last year, calling it the biggest welfare-to-work initiative in UK history.

Is it working? With more than 2.5 million people unemployed, there’s a huge amount riding on that question.

And the government now says it wants to export the same kind of model – outsourcing the problem to contractors and paying them according to the results – to other areas including cutting reoffending rates among prisoners.

Since last summer there’s only been one statistical release, which was seized on by ministers as evidence that the Work Programme was working, but failed to pass the FactCheck test.

The first full set of performance figures was due to come out this autumn, but the Department of Work and Pensions (DWP) told us the release date is still to be confirmed. Why the delay?

Channel 4 News’s social affairs editor Jackie Long has now obtained performance figures compiled by the second biggest Work Programme provider, A4E, after the first full year. It’s not perfect but it’s the best data we have seen so far on how the flagship policy is panning out.

The analysis

The data covers June 2011 to June 2012. The five areas of England where A4E acts as the “prime contractor” – the company in charge of how the scheme operates locally – are all included, covering the east Midlands, east London, Merseyside, Cumbria, Lancashire, the Thames Valley, Hampshire, the Isle of Wight and South Yorkshire.

Figures are also included for other regions where A4E works as a smaller partner under a different prime contractor. But here the company would have less influence on how the model operates, so we’ve left them out.

Over the first full year more than 93,000 unemployed people went on to A4E’s boooks. That alone netted the company more than £41m of taxpayers’ money in “attachment fees”.

Of those people, about 3,400 people have found sustained work – crossing the three- or six-month mark that triggers more payments for A4E, depending on what category the jobseeker falls into. That’s a success rate of less than 4 per cent.

The different categories are important because the government has laid out minimum performance targets for three groups: 18-24-year-olds claiming Jobseekers’ Allowance, over-25s claiming the same benefit and the supposedly ill or disabled people getting Employment and Support Allowance who have been assessed as fit to work after all.

DWP carried out historical analysis of the numbers of people who would be expected to find jobs for themselves even if the government did nothing to help them. They settled on 5 per cent and gave Work Programme contractors a cushion of an extra half of a percentage point.

So A4E need to find long-term work for a bare minimum of 5.5 per cent of all these kinds of claimants.

As things stand, A4E is failing to hit the minimum standard with all three categories of claimant. The success rate is just over 3 per cent for 18-24-year-old jobseekers, just under 4 per cent for the older group and barely 1.5 per cent for the Employment and Support Allowance section.

If things don’t improve the firm could lose its contract with DWP, as official departmental advice makes clear:”DWP expects that providers will significantly exceed these minimum levels. However, should providers fail to reach minimum levels for any of the customer groups, it will lead to contractual action up to and including contract termination if improvements to performance are not made.”

These figures are averages for the whole of the regions where A4E is the main contractor. As you might expect, there’s quite a bit of variation between region and type of claimant.

So when in comes to 18-24-year-olds in South Yorkshire, the company is hitting 5 per cent. But look at former ESA claimants on Merseyside, and the success rate is around 1 per cent.

Opponents of the Work Programme warned from the outset that providers would concentrate on the jobseekers who were the easiest to deal with and make less of an effort with the more challenging groups of people, despite added incentives from the government to deal with the latter.

A4E’s data has so much regional detail that it’s possible to look at individual offices, some of which are having an almost negligible impact on local unemployment.

The branch in Merton, south London, for example, has found sustainable work for just eight people in the nine months for which data is available. That’s obviously a little less than one job a month. Other branches have been more successful, clearly.

In total, A4E has received £45,893,535 from DWP, most of that, as we have seen, from the initial attachment fees. The cost to the taxpayer per job outcome so far is £13,498.

On the other hand….

We ought to stress that these figures are not official. They haven’t been checked and signed off by government statisticians, and indeed we’ve spotted what look like a few arithmetical errors in them – though nothing big enough to doubt the trends discussed here.

Most importantly, it’s still very early in the life of this policy, and there will be an inevitable time-lag before we can make a final judgement on the Work Programme.

We’re in a better position than we were back in June when similar figures emerged. A government spokesman said at the time: “To try to draw conclusions when most people have not even been on the programme for six months would be ludicrous.”

We’ve passed the six-month stage for many claimants now but the point still stands that there was always going to be a slow start to this scheme, and the early months where virtually nothing happened will skew these figures.

The percentages are likely to improve over time, but it remains to be seen whether the improvement will be enough to make sure A4E hold on to their contract, let alone make a sustainable profits or put a serious dent in unemployment.

An A4E spokesman told Channel 4 News: “The DWP plans to release official statistics for all Work Programme providers later this year, and we can categorically state that the information obtained by Channel 4 News at this time is out of date, has been previously cited, and is not data we recognise.

“Furthermore this data does not give a full or accurate picture of Work Programme performance.

“The Work Programme is just 18 months into a five-year outcome-based programme. The statistical data will undergo robust validation in DWP, compiled and released in accordance with the Code of Practice for Official Statistics.

“We have always maintained it is correct and proper to share performance data for the Work Programme, on the part of all providers and across all regions, when it is the right time to do so.”

By Patrick Worrall