Can Payment by Results really trigger innovation?

Fiona Sheil was responsible for co-ordinating NCVO’s programme of seminars, training and advice work on public service commissioning and procurement. Fiona left NCVO in October 2013 but we have retained her blog posts for reference.

The first point to concede with payment by results (PBR), is that despite the rhetoric it doesn’t guarantee any kind of ‘win’ by assuring the taxpayer only pays for results. This is a myth. The logic sounds plausible however there are at least five possible failings:

1. Overpayment for outcomes a cheaper procurement approach could have better purchased – you have to pay a premium to transfer risk to providers via PBR;
2. Where a scheme underpays and subsequent disappointing outcomes follow, the taxpayer picks up the tab for failure;
3. Paying for what we know as ‘deadweight’; that is, ‘results’ which would have happened regardless of service intervention;
4. Paying for the wrong outcomes and missing the important outcome by focusing, by necessity, on what is measurable and therefore auditable.
5. The cost of setting up PBR is significant enough that many experts currently argue it isn’t sustainable.

Commissioning Payment by Results: in what way will it achieve innovation?

PBR has two fundamental aspects with are intended to trigger innovation and subsequent improvement in public services:

1. The use of payments at risk – or payments as reward – to trigger change in behaviour;
2. Professional freedom to innovate the nature of service delivery outputs, and to focus on and end outcome

Ways in which the former can be made to operate are the subject of this blog.

The first question to ask in PBR is: are you trying to achieve a new innovation (ie something without a strong evidence base) or something that is proven. If it’s the latter, and you want to purchase a service that is acknowledged as best practice, then why use PBR to pay for it rather than a normal (‘fee for service’ contract)? PBR is expensive and complex to establish, potentially destabilising and divisive in the market, and expensive to pay out as the commissioner must reward providers sufficiently for the risks they take. If there is a clear service model, then it’s clear PBR simply isn’t an ideal tool for purchasing.

PBR is most appropriate when no one has a clue what works – because then it enables innovation through a focus on outcomes (as long as the risk balance incentives providers to trial new ideas and not just worry about safeguarding their business by sticking to the status quo). In such instances where the commissioner is seeking a step change in the quality of outcomes achieved (such as child protection services and other services around complex needs where significant improvements are made – identifying areas that need a step change in quality is how Social Finance target areas for social impact bond development) PBR can be used to create evidence around new innovation. It enables the provider to focus on end outcomes – as long as the system above and surrounding the provider is integrated, responsive and intelligent enough to enable the provider to implement effective ways of achieving that outcome.

Yet areas of greatest complexity, like child protection or parts of probation services, which are perhaps those areas most in need of change, are also those areas where it is hardest to achieve. In these areas where law or clinical rulings stipulate certain practice must take place, PBR is less possible and less relevant, as much activity by the provider is pre-determined. Without flexibility, the use of a risky innovation-stimulating procurement isn’t appropriate. In these cases the relevance of PBR’s intentions is diminished, as is the appropriateness of injecting risk into a part of the system that shouldn’t be upset by entering a competing driver (payment) in a world driven by other necessities.

If innovation is the main driver, is PBR really the best way to achieve it? Couldn’t a separate innovation fund be established, or investment made into research and development by more traditional mechanisms? This would surely be cheaper that established a complex PBR model as a research tool.

So how can PBR be used to trigger change in the right place?

How can PBR then try to tackle those most stifling parts of the system? It cannot change law or clinical guidelines, so where can it influence change? The answer is: only in those parts of the system that it can put at risk, and therefore stimulate by risk. In short, this means us: the market, not the commissioner.

Let’s take an example then of a situation in which it is supply chains that need to improve. We can argue that commissioners can stipulate in contracts where the risk can sit. They can specify that the payment risks stays with the prime or supply chain management body (consortia special purpose vehicle or lead) and that they contract down to their supply chain using standard upfront ‘fees for service’ as per usual. This way the prime / lead is encouraged to create better results through better management, partnership, performance management and integrated learning, and the frontline, presuming it isn’t the focus of need to innovate and improve, is safeguarded.

All well and good, but I would argue the failings of public services are mostly systemic and exist at the top of the tree in commissioning structures that aren’t integrated. Quality of innovation in provision isn’t really the problem: its innovation and risk taking in commissioning which we don’t see. (The evidence as to whether PBR’s focus on outcomes will encourage commissioners innovate by pooling budgets, or whether it’s inherent risks will push them further apart, is yet to be seen. But we’ll be reporting back on those PBR pilots in drug and alcohol and ‘troubled families’ that are testing this question.)

Using PBR to step change

If PBR is to create innovation, some further essentials are required, not least the rare beast that is an inquisitive commissioner, unbounded by the risk-adverse strictures of public sector procurement practice. Imagining you have both that and a market confident, keen, and capital-funded to try something new, you also need:

• To be ok with failure. Things may not work.
• Being a bit less casual about it: how able is the system and the user to cope with any failure? How predictable and manageable might such failure be?
• A solid system for gathering evidence. It needs to be evidence about what works in commissioning terms, commercial terms, and delivery terms (for the end user). Too often these pilots aren’t capturing what the market or user experience – something which requires effort and resource.
• To know that cashable savings are unlikely in a system already over-pressured. Instead, savings come over the long term through better outcomes, and outcomes that we hope are sustainable to halting recurrent problems.
• Have the confidence and clarity to decommission. The purpose of innovation is to replace the old and failed. A system needs to be prepared to manage the change.

If all this is fair, proven innovations achieved by PBR can be set into new ‘fee for service’ contracts – the standard model of payment. This stepped process proposed to transfer innovation from PBR into widespread ‘fee for service’ contracting cuts a role for PBR as a stimulus for innovation in the short-term and in specified circumstances only.

This blog is written following the early January meeting of the NCVO’s Payment by Results Working Group. The discussion fixed mostly on the suitability of payment by results (PBR) to solve a problem of needed innovation and improvement in public services. In what way, if any, is PBR a suitable tool for triggering innovation?

This blog captures only half our conversations from this meeting. In the next blog I’ll continue the discussion and ask: does the payment-risk-reward model squeeze out most voluntary sector providers?