(2) what research his Department has (a) undertaken, (b) commissioned and (c) evaluated on the effects of the use of voice risk analysis technology in benefit claim procedures on rates of error and fraud in such procedures;
(3) how many people have been prosecuted for benefit fraud as a result of information originally obtained using voice risk analysis technology since such technology was introduced; and if he will make a statement;
(4) how many people have had the payment of a benefit terminated as a result of investigations triggered by indications of risk from voice risk analysis technology since the introduction of such technology; and if he will make a statement;
(5) how many investigations of benefits claims have been instigated as a result of voice risk analysis technology providing an indication of risk since the introduction of such technology; and if he will make a statement;
(6) in how many telephone conversations between benefits staff and claimants voice risk analysis technology has been deployed since the pilot introduction of such technology.
In the last six months, there have been twice monthly telephone conversations between officials and the company selling the software to help ensure that the operation of the pilots runs smoothly. These are supplemented by face to face visits between officials and the company selling the software, and attendance at workshops as and when required.
Previous Ministers attended presentations and events with representatives of the company selling the software on three occasions.
There have been no meetings with independent experts in the field of voice risk analysis at this time.
The Department’s assessment of its voice risk analysis trials during 2007-08 in Jobcentre Plus and local authorities, and details of its evaluation methodology, has been made available. I refer the hon. Member to the written ministerial statement of 11 March 2009, Official Report, columns 19-20WS.
The Department has not commissioned external research bodies at this time. Departmental statisticians are conducting the research and they are required to offer independent advice to support the policy formulation.
At this stage, the Department has not estimated the impact of the rates of fraud and error detected or prevented via voice risk analysis on the national stock of fraud and error. Such estimates will be calculated should the voice risk analysis process prove effective at discriminating between high and low risk customers.
The information from voice risk analysis calls was not used to prosecute customers for benefit fraud or terminate customer’s benefits. In the pilots, voice risk analysis technology was used to trigger subsequent reviews of both high and low risk cases. As a result of these reviews, information could come to light that could lead to a prosecution or benefit termination but that decision would not rely on the results from the voice risk analysis call but on the information provided in the subsequent review.
During the pilots, there were no prosecutions from any of the reviews exercises.
The proportion of case changes are available across the pilots and have been previously reported. I refer the hon. Member to the written ministerial statement of 11 March 2009, Official Report, columns 19-20WS. However, the proportion of these case changes that were terminations is not reliably available across the first phase of pilots.
There have been no benefit claim investigations instigated as a result of voice risk analysis. The information from voice risk analysis calls has not been used of itself to determine investigations of customers benefit. Voice risk analysis technology has been used to trigger subsequent reviews of both high and low risk cases and there were a total of 787 reviews in Jobcentre Plus and a further 1,998 reviews in local authorities.
There were 1,676 telephone conversations between benefit staff and customers in Jobcentre Plus, of which a sample of 787 were reviewed and included in the final evaluation. From the first phase of local authority trials there were 6,816 telephone conversation between staff and customers, of which 1,998 were reviewed and included in the final evaluation.