Voice Stress

Councils using controversial lie detector tests to catch benefit fraudsters

‘Voice risk analysis’ being used by 24 English authorities at a cost of millions – despite scientists’ claims that it ‘does nothing’

Local councils are using voice risk analysis software against benefit fraudsters, even though the Department for Work and Pensions has dropped the technology. Photograph: Andreas Rentz/Getty Images

More than 20 councils have used or plan to use controversial lie detector tests to catch fraudulent benefits claimants, even though the government has dropped the technology having found it unreliable.

Responding to freedom of information (FOI) requests, 24 local authorities confirmed they had employed or were considering the use of “voice risk analysis” (VRA) software, which its makers say can pick out fraudulent claimants by listening in on calls and identifying signs of stress.

Although in 2010 the Department for Work and Pensions (DWP) announced it had given up using VRA software, the FOI responses show some councils have been spending millions of pounds on it.

Local authorities have continued to use the system to check whether people are honestly claiming the single person council tax discount, which allows single adults to pay 75% of the amount levied on a family.

Tory-controlled Derbyshire Dales said it had taken part in a county-wide review of council tax in 2011 that had used the technology – a contract worth £280,000 to Capita.

The same company was hired by Labour-run Southwark in south London and was paid £2.5m over three years. The council says VRA technology “was used as one tool to assist in determining the customers’ eligibility for the discount”.

The council said it did not record how effective the scheme had been, but did say that its real worth was in making the public aware that it would crack down on benefit cheats. A council minute last year records: “Although [VRA was] used in a minority of cases, a significant amount of publicity was received that assisted in communicating to residents the council’s intention to remove discounts if property occupancy could not be evidenced.”

VRA is supposed to detect signs of stress in a caller’s voice by analysing short snippets of speech, and is still used in the insurance industry to catch fraudsters. But critics say the system is not powerful enough to distinguish cheats from honest callers.

A number of councils – Redcar, Middlesbrough, West Dorset and Wycombe – said they were convinced of VRA’s merits and were considering using it in the future.

False Economy, the trade union-funded campaign group that put in the FOI requests to more than 200 local authorities, told the Guardian: “It says a lot about council outsourcing – and the benefits-bashing agenda – that this pseudo-scientific gimmick is now making its way in through the back door. Capita is a firm with a long rap sheet of expensive failure. Neither they nor their technological snake oil should be trusted.”

There have been complaints from claimants who were assessed using the technique. In South Oxfordshire two people formally protested after having their voices tested in 2013. The council says that Capita’s system helped reduce the number of people claiming the single person discount by 3%, and it would consider using it again.

Voice risk analysis has been mired in controversy since scientists raised doubts over the technology soon after it reached the market. In 2007 two Swedish researchers, Anders Eriksson and Francisco Lacerda, published their own analysis of VRA in the International Journal of Speech, Language and Law. They found no scientific evidence to support claims for the device made by the manufacturer.

Lacerda, head of linguistics at Stockholm University, told the Guardian that VRA “does nothing. That is the short answer. There’s no scientific basis for this method. From the output it generates this analysis is closer to astrology than science. There was very good work done by the DWP in the UK showing it did not work, so I am surprised.”

However, the Local Government Association, which represents English and Welsh councils, said the tool was used to help identify possible fraud. Peter Fleming, chair of the improvement board, said: “Councils detect almost £200m-worth of benefit fraud committed every year.

” No one is going to be prosecuted for benefit fraud on the result of voice analysis tests alone. But, in a small number of areas, councils use this technology as part of a wider range of methods to identify cases which may need closer scrutiny.”

The DWP told the Guardian: “Local authorities are free to design their own approaches to preventing benefit fraud.”

In a statement Capita said that, when it “undertakes a council tax single person discount review, councils can choose to use voice risk analysis technology as part of the process. The technology was never used in isolation. It is only used in cases which are deemed ‘high risk’, when earlier stages of the review have indicated that more than one person may be living at the property.”

Capita added: “The selective use of VRA technology is a useful additional tool in the validation process of identifying potentially fraudulent claims for single person discount.