| Previous Section | Index | Home Page |
9 Jan 2006 : Column 64W—continued
Benefits Administration Errors
Danny Alexander: To ask the Secretary of State for Work and Pensions what estimate he has made of losses due to error in the administration of benefits for each of the last eight years, broken down by the type of benefit. [28773]
Mr. Plaskitt: The Department produces estimates for the amount overpaid through fraud and error across the benefits system. Previously published results are included in the table.
| Monetary value (£ billion) | |
|---|---|
| From April 1997-March 1998 through to April 2002-March 2003 | 1 |
| April 2003-March 2004 | 1.5 |
Notes:
All figures rounded to the nearest £500 million.
9 Jan 2006 : Column 65W
The methodology used for estimating this was changed in 2004–05 to improve the quality of the estimate. We have also estimated fraud and error in
9 Jan 2006 : Column 66W
2003–04 using this new methodology for the purposes of making comparisons. The estimates for official error for 2003–04 and 2004–05 are in the table.
(16)Individual estimates have been uprated in line with 2004–05 expenditure and may differ from published values.
(17)Figures expressed as percentages (percentage) give the overpayments as a percentage of the benefit paid out in the year.
(18)The housing benefit estimates in this table include approximations for the purposes of producing an overall fraud and error figure. The estimates in the table will therefore be different from those published previously as National Statistics.
(19)The 2004–05 DLA National Benefit Review identified cases where the change in customer's needs have been so gradual that it would be unreasonable to expect them to know at which point their entitlement to DLA might have changed. These cases do not result in a recoverable overpayment as we cannot quantify or define when the customer's change occurred. Because legislation requires the Secretary of State to prove that entitlement to DLA is incorrect, rather than requiring the customer to inform us that their needs have changed, cases in this subcategory are legally correct. The difference between what claimants in these cases are receiving in DLA and related premiums in other benefits and what they would receive if their benefit was reassessed is estimated to be around £0.7 billion (+/- £0.2 billion). This figure is not included in the total above, but a similar figure was included in the global fraud and error figure of £3 billion reported in the resource account for 2003–04.
(20)Official error rates have been updated with estimates from the 2003–04 measurement exercise.
(21)Overpayments due to knock-on effects between loss of DLA component entitlement and premiums on income-related benefits.
(22)The estimated rate of fraud and error on unreviewed benefits, for which we have no reliable review information, has been set at 2.7 per cent. and divided equally among fraud, customer error and official error.
Notes:
1.Fraud and error on CTB has not been measured before and is assumed here to be equal to fraud and error rates on HB.
2.Although quoted to the nearest £0.1 billion, there are considerable uncertainties around the total estimates (95 per cent. confidence intervals of between ±£0.2 billion and ±£0.3 billion).
3.Individual figures may not add across to total fraud and error figures due to rounding.
The Department publishes National Statistics reports including estimates of overpayments through official error. Estimates for income support (IS), jobseeker's allowance (JSA), pension credit and housing benefit (HB) are published every year so these are the benefits where we have most information on changes over time. These figures are presented in the tables.
Note:
The figures for April 2003-March 2004 and October 2003-September 2004 in the above table have been adjusted for data quality problems found. Previous figures have not been adjusted in this way as there is not enough information available to do so. The tables therefore are not all suitable for making reliable comparisons over time but give a reference of our best estimate of these overpayment figures in each year. Confidence intervals are not provided in these tables although are typically around ±20–30 per cent. of the central estimates.
9 Jan 2006 : Column 67W
(23)Latest published estimates
Notes:
1.These figures will not include all overpayments in HB through official error. They were published in the latest National Statistics report for HB fraud and error, and only cover the 85 per cent. of HB expenditure that is within the scope of the HBR sample. In these reports, there was an extra category of error labelled non-residence" error, an unknown amount of which will be due to mistakes made by officials—these overpayments are not included in the figures in this table.
2.The Second National Housing Benefit Accuracy Review, in 1997–98 gave an estimate of £60 million for official error, but was based on a different methodology, so the results are not considered comparable with more recent estimates.
The Department also publishes one-off snapshot reviews. The most recent report covered disability living allowance (DLA) and the official error figures from these are in the table.
| Amount of money overpaid (£ million) | Percentage benefit overpaid | |
|---|---|---|
| Official error | 60 | 0.8 |
Annual estimates of official error only are made for short-term benefit overpayments (incapacity benefit and severe disablement allowance). The recent results are shown in the following table. Previous results were carried out using a different methodology and are not comparable.
1.These estimates are subject to wide margins of error. Our best estimate of STB overpayments for 2004–05 includes a further amount of around £20 million due to incorrectness types not captured under methodology in earlier years.
9 Jan 2006 : Column 68W
Annual estimates of official error only are made for long-term benefit overpayments (retirement pension, widow's benefit and bereavement benefit). The recent results are shown in the following table:
| Monetary value (£ million) | |
|---|---|
| April 2001 to March 2002 | 43 |
| April 2002 to March 2003 | 14 |
| April 2003 to March 2004 | 26 |
| April 2004 to March 2005 | 27 |
Notes:
1.The sample for long-term benefits is a small one and excludes some types of cases. It provides reassurance that official error on these benefits is a very low proportion of expenditure, but the estimate is only a broad indicator.
2.As a result of the wide margins of error, the changes over time are not statistically significant.
| Next Section | Index | Home Page |