My right honourable friend the Minister for Employment and Welfare Reform (Tony McNulty) has made the following Written Ministerial Statement.
I wish to update the House on the progress made in piloting voice risk analysis technology.
In its 2005 strategy document Reducing fraud in the benefit system: achievements and ambitions, the Government made a public commitment to test voice risk analysis technology, as one method of preventing fraudulent claims to social security benefits over the telephone.
The department is conducting its own research into the efficacy of voice risk analysis based on trials by local authorities and Jobcentre Plus. The department's research is based on field tests of the technology that assess its capacity to differentiate successfully between high- and low-risk claims. The department is committed to ensuring that decisions to use or not use the technology are supported by evidence. The department’s evaluation is being conducted by departmental statisticians who are bound by a national code of practice.
Departmental statisticians have completed an evaluation of the data from six trials in local authorities and one by Jobcentre Plus. The pilots included reviews of existing housing benefit claims in local authorities, or consortia of local authorities, and new claims to income support and jobseeker’s allowance in Jobcentre Plus. One additional local authority trial returned too few calls for analysis.
In each trial customers provided information to support their claim over the telephone and staff used voice risk analysis technology to help classify the call as high- or low-risk. A selection of customers from both groups was followed up with a face-to-face review that aimed to corroborate the operator’s judgment by seeing whether this further investigation revealed changes that would have affected the benefit in payment.
The results are shown below.
“Low risk” cases with benefit change following further investigation “High risk” cases with benefit change following further investigation Difference between benefit changes in high and low risk cases New Claims Jobcentre Plus 29% 37% 8% Reviews Birmingham 5% 40% 35% Derwentside 4% 35% 31% Edinburgh 12% 43% 31% Harrow 22% 32% 10% Lambeth 16% 21% 5% Wealden 10% 12% 2%
“Low risk” cases with benefit change following further investigation
“High risk” cases with benefit change following further investigation
Difference between benefit changes in high and low risk cases
These findings indicate that the results from the application of VRA varied from pilot to pilot.
Earlier this year, the department announced a second phase of local authority pilots (24 in total) that are now under way and will run for 12 months. These will provide a broader evidence base and a clearer indication of the technology’s capacity to distinguish reliably between high- and low-risk cases. Results will be available in spring 2010.
All of the councils involved have reported a very high level of customer satisfaction with the new process. Customers find it more convenient and faster than the previous process based on filling in forms or visits.
I will update the House as to the findings when they are available.