Payment by results: ends, means and need

The government’s wide-ranging ambitions for introducing payment by results (PbR) regimes across a swathe of policy areas (already being seen in re-offending, employment, homelessness, troubled families and keeping children out of care) will significantly affect the way that social policy outcomes are addressed.

There are big differences between schemes grouped under the broad PbR umbrella but one common feature is just how much of a punt they can represent. In particular ‘schemes that make a large part of the payment dependent on performance are still largely untested and their effectiveness is not yet proven’.

That’s a quote from one of three relevant reports which have arrived in quick succession. The first (from which that quote comes) is a very balanced account by the Audit Commission of the opportunities and risks associated with locally designed PbR schemes. The second, also from the commission, is a hard hitting account of PbR in relation to short stays in hospital. The third has been the focus of some controversy in London with allegations that aspects of a very critical evaluation of the young offenders ‘Daedalus’ scheme have been watered down by the GLA.

The net effect of looking at these reports is to reinforce just how problematic PbR could be for professionals grappling with some of the most intractable local economic and social policy issues and the degree to which policy is running well ahead of evidence, understanding and perhaps even more critically the necessary skills, particularly in commissioning.

The broad theory is clear: (nominally) we don’t do targets any more, we much prefer outcomes based commissioning which transfers risk and creates incentives to greater efficiency since payments increase the more clients are successfully helped. Even better we pay the providers from the savings but are still quids in. What’s not to like?

Well, potentially quite a lot as has been well described in many, many places including:

• the major disincentive for many smaller organisations which lack working capital when payments are almost entirely made at the end of the process
• whether outcomes are short term or long term and, if the latter, whether there are sensible interim stages (and if so doesn’t this inevitably move away from outcomes)
• potentially significant complexity in designing outcomes and allocating risk to limit gaming and to identify specific roles and impacts of individual organisations in complex systems
• difficult and therefore more expensive clients being ignored or shunted.
• third party investment (via social impact bonds) adds more complexity in assessing returns to funders.

So a lot of focus is on design questions, in particular to address gaming and structure incentives correctly as well as to identify the right outcomes and to deal with that very knotty problem of whether risk is actually transferred. Some might start muttering PFI at this point.

From a practical perspective one of the most serious issues is the degree to which PbR accentuates the risks of skewing front line practice rather than, as the ‘black box’ theory suggests, leaving professionals to make the best decisions in the interests of the clients concerned. In this context it is illuminating that one of the passages allegedly excised from the London report is as follows:

‘Resettlement brokers said they would at times feel conflicted in their working practice; there was an uncertainty as to whether their key concern at any one time should be meeting targets, or meeting the needs of young people.’

One of the criticisms of the target culture was, as we used to say in one of my previous organisations, you could ‘hit the target but miss the point’. How much more pressing this becomes when all or the vast majority of the resources are associated with hitting that target. Miss it and you’re bust, perhaps literally so (and indeed that is the right outcome according to the theory).

The risk of badly designed regimes driving professional practice away from need deserves proper examination.

Secondly, any PbR regime demands accurate measurement. On this the NHS experience makes depressing reading. The Audit Commission identifies widespread disagreements between those who commission NHS services and those who provide them about how to classify a patient who stays in hospital for under 24 hours, leading to poor data quality and incorrect payments. To the untutored eye this seems a remarkably straightforward and long standing metric compared to some of those being considered in relation to other social policy outcomes yet:

‘We have also found that NHS managers are spending valuable time debating how patient treatment is recorded or described, time that would be much better spent focusing on the treatment itself.’

At this point, some might start muttering RIRO.

Good commissioning has the potential to generate localised responses which do indeed bring in new resources and organisations and allow professionals to focus on doing their job well. However, getting it right is extremely hard and at the moment one does fear that focusing immediately on the more radical end of the payment spectrum means being ushered into some very choppy waters in a group of coracles looking to the stars rather than in a sleek flotilla with the latest GPS.