PbR is simple in theory…
Payment by results is quite a straightforward concept.
Its chief attraction lies in its ability to incentivise providers to deliver exactly what a commissioner wants.
For example, any PbR contract concerned with reducing reoffending should ensure that organisations receive the biggest payments when they succeed in getting prolific offenders to give up crime.
This saves the commissioner – the Ministry of Justice – and the country money and is to the benefit of everyone in society.
However, getting the contract right in practice is proving rather more challenging – indeed, I’ve yet to go to a PbR event where at least one speaker hasn’t said:
“The devil is in the detail.”
PbR is complex in practice
A good example is the current Work Programme contract which does reward providers with much larger payments when they place people such as recovering drug users who’ve been on sickness benefit for many years into jobs.
However, despite this clear incentive, the overall design of the contract has meant that most providers have concentrated on getting the easiest to place jobseekers into work as this ensures them a continuous cash flow.
The balance of the contract is wrong and as a result we have increasing evidence that providers are “creaming” the easiest cases while providing a minimal service to those in greater need.
The Sweet spot
Just how difficult it is to get a PbR contract design just right was emphasised to me at a recent workshop I was running at the Probation Chiefs’s annual conference.
The Ministry of Justice faces a considerable number of challenges in designing the new reducing reoffending contracts that are scheduled to start in 2015.
The major ones are:
- The contracts must be long enough to encourage a wide range of providers to bid while still providing break clauses for those that under-perform.
- The proportion of the contract paid on a successful outcome/PbR basis must be big enough to drive innovation and improved performance but small enough to be a realistic financial model that medium and smaller providers can bid for.
- While it is reasonable enough for price to be part of the awarding criteria of any public contract, it will be hard to assess likely quality, especially since most providers won’t have a track record in reducing reoffending.
- The contracts should promote innovation with the intention of sharing best practice, but at the same time may have to allow providers to keep at least some of their intellectual property rights.
- There is also a substantial challenge in designing an outcome metric which includes both binary and frequency measures of reoffending.
- The contracts must be large enough to have substantial cohorts so that outcome metrics can be valid, but flexible enough to reflect local needs and existing effective partnerships.
When you factor in all these different variables (and the list above is far from exhaustive), you get a pretty complicated looking Venn diagram with a tiny sweet spot:
If you have any other factors that you think should be built into the reducing re-offending contracts, please suggest them in the comments section below:
2 Responses
Russell – you’re right that outcome payments are tilted heavily in the direction of Work Programme clients who are furthest from the labour market. But that doesn’t mean there are any financial incentives to help these people.
When you weight the outcome payments of easier- and harder-to-help groups by the chances of getting each into work, the ‘risk-weighted revenue’ is probably less for the hardest to help. WP offers up to £4.4k for helping a standard adult jobseeker into sustained employment. Let’s say you need to help three such people in order to succeed with one, then your risk-weighted revenue is around £1.5k. For longer-term sickness benefit claimants, WP offers up to £13,700 per outcome. But if you need to help ten of these people to succeed with one, then the risk-weighted revenue is around £1.4k.
So the revenue side of things doesn’t look great in terms of incentives to help those most disengaged from employment. But the costs side makes the decision for providers a no-brainer. Since providers would have to spend substantially more on the hardst-to-help than on job-ready people, the finances just don’t stack up, unfortunately.
Thank you Russell. .At the risk of being controversial, I wonder if the current focus at the less severe end is ok, though costly. Whilst a perfect algorithm would weight both ends equally, isn’t the ‘easier’ end today the ‘hard end’ of future years?
The very valid reasons that people have for not being in work do not reduce over time, they tend to strengthen and new ones are added. We all know that the ‘workshy’ theory is bunkum, people want to work,
By helping people before they become entrenched we might prevent at least some of the low self-esteem, hopelessness and lack of self-belief that traps the long-term unemployed. If front-line systems become more efficient then over time focus will turn to those harder cases that are left.