Under pressure from budget cuts and the need to evidence that public funds actually deliver results, local authorities are looking into the possibility of using payment by results contracts.
The term payment by results (PbR) can encompass a large variety of arrangements for the delivery of public services. In the NHS it means tariff-based, unit pricing, but in the social care sector it means payments for social, economic or behavioural outcomes.
PbR arrangements typically have the following features:
- The service provider is paid only if pre-agreed outcomes are achieved. This can be immediately apparent or take a long time to realise.
- The provider decides how the outcomes are achieved. The specification for the service will therefore be relatively light- an expression of the outcomes may be all that’s required, for example, a percentage reduction in unauthorised absences from school, as in the Troubled Families programme).
- Objective data determines whether (and how much of) the contract price is payable.
The benefits of using PbR are:
- Risk of performance on provider. If they do not perform well, they will not get paid.
- Value for money and accountability- the public body only pays if the service delivers.
- Enables more freedom on the provider to innovate. The idea is that if the provider bears the risk, they decide how to achieve the outcomes with as little prescription as possible from the commissioner.
Ways of incorporating PbR into social care contracts
As David Hunter explained in his blog on Social Impact Bonds, PbR can be understood on a spectrum. Many arrangements use PbR for only part of the contract price through paying bonuses or withholding part of the payment to incentivise performance. The following mechanisms may be used (in descending order of risk to the provider):
- All of contract price at risk, i.e. if the outcomes are not achieved no payment will be made (for example, some Social Impact Bonds).
- Some payment up front to represent mobilisation costs then payments for outcomes (for example, the Troubled Families programme).
- A specified percentage of contract price at risk (for example, Serco’s Doncaster prison contract, in which 10% of the fee is paid if there is a specified reduction in reoffending results after 3 years compared with national averages).
- Service credit schemes (as is common in IT and other back office services contracts, see Practice note, Service levels and service credit schemes in outsourcing).
- Bonuses/ service debit schemes, in which additional funds are available for hitting targets.
If it’s so good, why isn’t everyone doing it?
PbR arrangements present a number of challenges. The extent of the challenges depends on many factors, the most important of which are:
- The availability of clear and accurate data that will enable the outcome to be precisely defined and measured.
- The amount of the contract price that is at risk.
If more than the provider’s margin is at risk, and the provider is therefore being asked to fund all or part of its operating costs, the risk profile shifts and may have a significant impact on which organisations can bid and the ultimate cost of the service.
The risk also increases with an increase in time between service commencement and the date on which it is clear that the outcomes have been achieved and the payment for the achieved outcome is made. This may be why many projects that rely on a PbR mechanism in this sector are relatively short term, for example, the Troubled Families programme expects to see results within a year.
The key financial challenges include:
- Cash flow: Many providers in this sector will not be able to finance the whole or a significant part of a service themselves. They may therefore be prevented from bidding for it, or will be forced to access finance externally. The cost of that finance will be built into the contract price.
- Price-setting: It can be difficult to determine the appropriate price for a PbR contract. The price could reflect the cost of providing the service, plus margin (plus the cost of finance), or the savings that will be generated if outcomes are achieved. For example, it may be relatively easy to quantify the immediate costs of a child being taken into care but what is the cost (financial and in terms of well-being) of avoiding that outcome? The long-term impact may not be known for many years; and the savings may accrue to a different part of the public sector from the commissioner’s.
- Who commissions for individuals with complex needs and who pays for their outcomes, for example, individuals coming out of prison claiming benefit and with drug & alcohol problems. The commissioning organisation that is responsible for meeting the enhanced payments may not be the one that benefits the most. This could mean more multi-agency working.
- Balancing incentives against the pressure on the public purse. If a bonus scheme is used, over-performance can mean over-payment. Bonus schemes should be modelled appropriately and capped.
- Financial incentives are not the only way of incentivising the delivery of quality services. For example, where the provider is gaining enough profit from the arrangement already, any bonus may be icing on the cake that it is too difficult to pursue. Furthermore, much of what works in health and social care is due to dedication of employees who may not be incentivised by (or receive any of the) financial reward. How is that drive maximised/ rewarded?
- Future costs must be reflected in long-term planning. In other words, the commissioner may be deferring payment now for a bigger payment later.
- Risk of provider failure may be higher due to cash flow risk. The commissioners must plan for contingencies.
How are outcomes defined and measured?
Defining the outcomes can be tricky. There is a risk that poorly defined outcomes will drive the wrong behaviour, for example by incentivising activity by paying for numbers of service users on a particular programme rather than looking at what benefits participation in that programme brought to the service user. Or the provider may focus more of their attention on those service users with whom they are most likely to achieve the outcome than on others who may require more intensive interventions, a concept known as “creaming and parking”.
Lack of quality data makes defining outcome tricky. The problem is particularly acute in the social care sector in which it is more difficult to articulate what success looks like. Many outcomes, for example around well-being, seem to defy objective measurement. This may lead to arrangements for which the first years of the contract are about establishing a baseline against which outcomes achieved in later years are measured, though this brings contractual, and potentially procurement, risks.
Some of the other challenges include:
- Causation: The outcomes may have occurred anyway, without the service, for example, a scheme funded to encourage people back into work may achieve its outcome, that is, a percentage of long-term unemployed back in employment, but this could be due to wider economic factors rather than anything the provider did. Similarly, a smoking cessation service may achieve results due to a national rise in cigarette prices. In some cases, these variables can be taken into account by benchmarking the provider’s results against national averages or local averages, but as the metrics have to be agreed in advance, it will be difficult to foresee what external influences might affect the outcomes in the future.
- Attribution: What if multiple providers are involved? Who gets the payment? Whose intervention made the difference?
- Administrative pressures: PbR contracts can be burdensome. For example, reporting requirements might be onerous, and the relationship between the outcome and payment may be overly complex. Measurement must be integrated into an organisation’s own governance and processes, not be an add-on, providing an extra layer of bureaucracy. Ideally, the reporting draws on information that is already generated, for example for statutory or audit purposes.
- Part of the outcomes may not be in control of provider- providers may have to take some risk for matters outside their direct control. The more they do this, the riskier and, logically, the more expensive the service.
- A better outcome might not be a cheaper outcome, for example, more people might want to use a service.
When is PbR appropriate?
Using some form of financial incentive or deduction will be appropriate in most contracts. But where the amount at risk eats into the provider’s operating costs, special care is necessary. PbR arrangements in which more than the provider’s margin is at risk are often appropriate where:
- Outcomes from the service are clearly articulated and understood and delivery is, or is largely, within the provider’s control.
- Outcomes can be expressed so as to incentivise only positive behaviour.
- Accurate data is available to establish the “as-is” position and will also be available to determine how far the provider has improved on the “as-is” at the end of the measurement period.
- The providers in the market can finance the risk themselves, or secure cheap finance to do so.
- The value of the service is sufficient to justify a complex structure or financing arrangements, if they are required.