England & Wales

Public finance reform and the limitations of evidence


Away from the main stages and media drama a consistent fringe theme at all three party conferences was public service reform and in particular their future financing.

Although it was framed in different language, there was a recognition amongst all three parties that we needed to find new ways of designing public services to cope with long term demands that did not rely on a never ending, infinitely expandable stream of money from the state.

One of the most talked about ways of doing this was various forms of payment by results. This is something the government is very keen on, it figures prominently in the work programme and councils as different as Westminster and Havering are using it to fund local services.

Some of the attractions of payment by results are obvious: the complete (or apparently complete) transfer of risk to an external  provider. Other are subtler but no less real: a focus on outcomes, better sharing of data and an improved understanding of cause and effect amongst them.

There are also several areas of concern that I heard expressed repeatedly. The first of these concerned the willingness of central government to put its money where its mouth is. Many of the savings created by preventative programmes are realised from someone else’s budget.

So local government might invest in public health on a payment by results basis,  through smoking cessation or healthy eating programmes for example, but the savings generated by these initiatives accrue to the NHS budget and local government realises no benefit from its investment. So while we might achieve limited results by operating on a payment by results basis within the purview of the local authority, to achieve any sort of transformative scale we need a mechanism for pooling budgets and sharing savings across the public sector more generally. Community budgets are key to this but at present interest in them at government level seems confined to DCLG.

The second area of concern arises from this point.  Many participants in conversations about payment by results were anxious about the level and quality of available evidence to support this approach. A real evidence base around the cashable savings achieved was seen as key to getting support for such initiatives particularly from the hard headed mandarins of Her Majesty’s Treasury.

There are some particular problems here. Many of the real savings achieved by payment by results programmes are complex to measure and extremely long term: children not going into care, people not developing heart disease. Unless you have a financial model which can operate on very long term time frames (and it’s not clear that anyone does) you are likely to want to make payments not on the basis of results per se but on proxies for those results which introduces another level of complexity and possible error.

So evidence matters of course, but we should be realistic about its limitations. People bemoan a lack of proper evidence. But evidence doesn’t just happen. You have to go out and find it. And evidence isn’t simple. As philosophers of science know, it’s rarely impartial, normally it’s gathered to support or refute a particular thesis. It’s hedged about with hardwired cognitive biases whereby we see what we expect and want to see or discern patterns that are not really there.

So we need to manage evidence and in part that’s about having the political courage to undertake the sort of initiative that will drive the evidence forward. It’s easy to sneer about how quickly evidence policy making becomes policy based evidence making. In reality, however, that may be the only way to stimulate progress.