Analysis - New Deal performance 2006-2007

We recently acquired the New Deal results for the first year of the new 'prime contract' regime in London and the South East. Rather than keep it to ourselves, it seemed only fair to share them more widely, along with some explanation and analysis.

Providersort iconDelivery DistrictCumulative StartsCumulative LeaversCumulative JobsJER TargetJER Achieved
Action for EmploymentBerkshire, Buckinghamshire and Oxfordshire1588112422545%20%
Action for EmploymentCentral London3316191325445%13%
Career Development GroupHampshire and Isle of Wight4157260843345%17%
Career Development GroupNorth & North East London (Delivering in North London only)3409184138045%21%
Career Development GroupSouth London9623580580945%14%
Career Development GroupSurrey & Sussex2993218638545%18%
Community LinksCity & East London2862158852245%33%
Skills Training Uk LtdLambeth, Southwark & Wandsworth4322268152045%19%
Skills Training Uk LtdWest London5126315159345%19%
As the figures show, New Deal performance is below target in all of the 10 New Deal contracts that cover London and the South East regions. The worst performer, A4e Central London, is achieving less than a third of the job entries it promised in order to win its contract, and the best is still only getting 3/4 of the jobs it promised. So what's going wrong? Are these really as bad as they look? What's a start? What's a leaver? What on Earth is a JER? The first thing to realise is that New Deal is not a single course, and these figures are not the number of people entering New Deal and the number getting jobs. Jobcentre Plus advisors still deliver a key part of New Deal, and send the customer on various courses run by the Prime Contractor. These figures show the performance of the Prime Contractor. Thus, the actual meaning of each of the columns is as follows:

  • Start - Each time a customer starts a new course while on New Deal. Thus a customer attending 3 courses on New Deal counts as 3 starts
  • Leaver - Each time a customer leaves a course they're on, whether to get a job, because they didn't turn up, or because it finished. Repeating the same course twice doesn't usually count
  • Job - When the provider can prove that the customer starts work within 6 weeks of leaving, in the Jobcentre-approved method, they claim a job outcome
  • JER Target - Job Entry Rate Target. The percentage of people who should enter work after taking each course in New Deal. This overall figure of 45% is presumably an average, as each course typically has its own expectation
  • JER Achieved - Job Entry Rate Achieved. Jobs divided by leavers. To give an example of the impact of this, if somebody takes two courses and gets a job after the second, then the job entry rate is 50%. If two people take a long course and one gets a job at the start, while the other stays on the course and does not get a job, then the job entry rate will be 100% until the second leaves, when it will go down to 50%

So why are the Prime Contractors below target? The performance figures give us a useful indicator of how each provider is doing, but to understand the reasons behind the figures, we need to look at the wider picture. I've split the analysis into three sections for further discussion, based around the following explanations:

  1. the contracts are too new
  2. the providers are doing a bad job
  3. the contracts were unrealistic

Are the contracts too new to provide accurate performance information? No. The figures are taken from July 2006 to June 2007. New Deal should never last more than a year for one person, and courses generally don't last more than 13 weeks. Given that performance is measured for leavers, there is potential for skewed results arising from people who either drop out early, or who find work during a course. However, these act in opposite directions on the figures, and are unlikely to be major compared to a full year's results. Are the providers doing a bad job? Probably, in some cases. The degree of variation between contracts indicates that something is amiss, and it would be worth investigating Community Links to see what they're doing that other providers aren't. However, every provider is below target, providing scant evidence that a 'good job' is achievable by anyone. Shortfalls in delivery by providers can be split into:

  • Prime contracting - Providers are still getting used to managing a range of subcontractors in large contracts. Much of this was previously carried out by Jobcentre Plus staff, and requires new skill sets that take time to develop
  • Individual performance - One bad or good course can sway the overall job entry rate disproportionately. Gateway to Work, a short course that every New Deal entrant should take, has far more starts than other courses. A separate table is available in Further Reading that excludes Gateway to Work from results, leading to anywhere from a 4% (in Hampshire) to a staggering 51% (in Berks Bucks & Oxon) improvement in contract results. If I were A4e, I'd be looking closely at Gateway to Work in that area

Were the contracts unrealistic in the first place? Yes. Compared to previous New Deal contracts:

  • There's less money - Providers get the same money for each start and each job outcome, but Prime Contractors have to take 5-15% of this to cover management costs. Jobcentre Plus previously managed individual local contracts with in-house teams
  • The targets are higher - This is most evidently the case for Gateway to Work, which moved from 20-25% JER targets to 45%, higher than anyone in the contract areas had been achieving
  • The job entry evidence is tougher - To claim a job outcome, a provider needs a declaration signed or stamped by the customer's employer, within 6 weeks of the customer leaving their course. Previously, the employer details and a 13 week period were considered acceptable.
  • Jobcentres no longer track performance for providers - In previous contracts, it was possible to link with Jobcentre Plus tracking systems. JCP no longer track job outcomes directly, instead linking with Inland Revenue contributions data for statistics

The JER targets also failed to take account of local circumstances, or of previous performance by providers at local level, as is evident from the use of a single target to cover some of the most and least deprived Districts in the country. Some areas undoubtedly have greater deprivation, and customers face greater barriers to employment. However, the proportionately higher performance of City & East, probably the most deprived of all the Districts, does not lend weight to this complaint. Conclusion - looking deeper In a briefing event for the New Deal bidding that led to the current contracts, the then head of WorkDirections UK, Richard Johnson, told the assembled providers and Jobcentre Plus staff that he had no intention of submitting a tender. He'd done their sums, and they would lose money. The unspoken corollary is that they would lose money if done properly. Many of the prime contractors are essentially dependent on Jobcentre Plus / DWP funding, or a combination of that and LSC funding. Some had almost all of their turnover tied up in New Deal. They had to bid, despite the financial and performance problems they would undoubtedly encounter. At this stage, it's probably too late for the DWP (who took over contract management earlier this year) to take away any New Deal contracts. The existing ones only have another year or two to run, the premises, subcontracting and TUPE regulations would make them unattractive and awkward for other providers, and there's no evidence that any provider is exceeding government targets in these regions. However, a cycle of overpromising and underdelivering benefits nobody, especially when it reaches such extremes. Hopefully, the next New Deals will see experience of the Prime Contract model bringing greater realism on both sides. Further reading: The same New Deal performance figures, without those for Gateway to Work courses included, are available here. Jobcentre Plus has stats for New Deal Options (the middle part of New Deal delivery) here, but only cover a three month period at the time of writing. Agree? Disagree? E-mail me your comments, and I'll publish them next week!