Skip to main content

Social Security Benefits: Fraud

Volume 491: debated on Wednesday 22 April 2009

To ask the Secretary of State for Work and Pensions what assessment he has made of the level of fraud and error losses in working age income support and jobseeker's allowance in each (a) region and (b) Jobcentre Plus district (i) in each year since Jobcentre Plus's fraud and error targets were launched and (ii) in each of the last 24 months. (264236)

The level of fraud and error in income support and jobseeker's allowance is estimated using results from a sample survey of benefit claims, which is designed to produce a robust estimate at the national level.

The first Monetary Value of Fraud and Error (MVFE) target for Income Support/Jobseeker's Allowance (IS/JSA) in SR02 ran from a baseline year of 1997-98 through to 2005-06. During this period as well as the national estimates used to evaluate progress against the target, tables were also published that showed estimates of MVFE for Government Office Regions, which were equivalent to JCP regions from 2002.

With the introduction of a new SR04 target in 2005-06 DWP statisticians analysed the historical regional estimates produced during SR02 to assess whether they had been a robust indicator of MVFE at regional level. Analysis showed that for many regions the estimates of fraud and error fluctuated substantially year on year. These fluctuations were judged to be due to the relatively small sample of claims investigated in any one region. Since the sample survey is designed to produce the best estimate at a national level, estimates of fraud and error at a lower geographical level are based on a small sample of claims and have a high risk of not accurately representing the amount of fraud and error for all claims in a region.

Following this analysis DWP have not published regional estimates from 2006-07 onwards as they were judged to be not robust enough to meet the standards required of national statistics and there was a high risk that fluctuations in the estimates would be misinterpreted as indicating real change in fraud and error in the regions.