The strategy to deliver personalisation through personal budgets using a process conceived by the charity In Control was ushered in with expectations few would argue against:

• • A new way to allocate resources that would be both fair and transparent. This would be delivered through a Resource Allocation System (RAS) to replace the capricious ‘professional gift’ system (Duffy, 1996).

• • Savings through a reduced level of bureaucracy. With service users carrying out their own support planning, there would be less need for assessment and care management (Leadbeater, 2008).

• • Greater control for service users. This was the overriding vision, and would end the domination of support planning by the menu of pre-purchased services.

These expectations form three tests that, between them, can be used to evaluate the degree of success of the strategy. Five years of implementation has provided a fund of evidence to do so.

Has it changed the way resources are allocated?

The up-front allocation is meant to be ‘indicative’ of how much the person is entitled to. This is to allow any minor adjustment that might be needed to ensure eligible needs can be met. As such, the indicative and actual allocations should be ‘as close as possible’ (TLAP, 2011) to each other. Measuring the difference between the indicative and actual allocations therefore offers a test of whether the RAS is in reality a new way to allocate resources.

There is no routine collection of either of these sets of data. However, Freedom of Information requests can produce the data. The largest of these was carried out by the London Self-Directed Support Forum (2013). All 33 London councils were asked for the indicative and actual allocations for all new service users for the year 2011/12, resulting in the following findings:

• • 21 of the 33 councils could not provide the data as they did not collect it. This was surprising as there is clear advice from the Association of Directors of Adult Social Services (2010) to collate the data in order to continuously monitor and adjust the RAS. It suggests that these authorities may not be regarding the up-front allocation as seriously as expected.

• • For those that did provide data, the difference was measured as a ratio, whereby if they were the same, the ratio was 1. It was found that the average ratio was 2.6.