Select Committee on Public Accounts Fifty-Fifth Report



THE SCALE OF FRAUD AND ERROR IN BENEFIT PAYMENTS, PARTICULARLY INCOME SUPPORT AND JOBSEEKER'S ALLOWANCE

  1. Fraud and error affects all benefits to some degree. The Department estimate that in total there may be 2 billion of fraud and 1 billion of official and customer error each year (Figures 1 and 2).[6]
  2.  

    Figure 1: Total Estimated Fraud Losses Each Year

     

    Continuously measured benefits

       

    Income Support

    573m

     

    Jobseeker's Allowance

    201m

    774m

    National Benefit Review in 2000

    Incapacity Benefit

     

    19m

    Other benefits - adjusted figures derived from older National and pilot benefit reviews

    (Disability Living Allowance, Attendance Allowance, Retirement Pensions, Invalid Care Allowance, Child Benefit, Housing Benefit

       

    Total (rounded)

     

    1,100m

    Estimate for unmeasured benefits

    Total (rounded)

     

    190m

    Total estimated fraud losses (rounded)

     

    2,000m

    Source: Department for Work and Pensions[7]

     

     

     

     

     

    Figure 2: Estimated Levels of Customer and Official Error Each Year

    Continuously measured benefits

       

    Customer Error

       

    Income Support

    154m

     

    Jobseeker's Allowance

    11m

    165m

         

    Official Error

       

    Income Support

    173m

     

    Jobseeker's Allowance

    78m

    251m

         

    National Benefit Review of Incapacity Benefit

       

    Customer Error

     

    16m

    Official Error

     

    22m

         

     

    Other benefits - adjusted figures derived from older national and pilot benefit reviews

     

    Customer Error (rounded)

     

    400m

    Official Error (rounded)

     

    100m

         

    Estimate for unmeasured benefits

       

    Customer Error (rounded)

     

    60m

    Official Error (rounded)

     

    40m

         

    Total estimated customer and official error (rounded)

     

    1,000m

    Source: Department for Work and Pensions [8]

  3. These figures are not comparable with the estimate of 7 billion of fraud used in the 1998 Green Paper, Beating Fraud is Everyone's Business: Securing the Future, [9] because that broad estimate included 5 billion for cases where there was high or low suspicion of fraud, but no evidence. Since then, the Department have based their estimates on what is known about confirmed fraud, and in addition there have been changes in benefits, for example Family Credit has ceased to be a benefit.[10]
  4. The figures for fraud and error in Income Support and Jobseeker's Allowance, 1.2 billion in 2000-01, are based on an in-depth rolling programme of reviews of some 30,000 cases. In addition, cases where there may be suspicion of fraud, but where there is an absence of evidence, totalled some 412 million on Income Support and 108 million on Jobseeker's Allowance, both significantly down on comparable figures for 1999-2000. Figures 3 and 4 break down estimates of fraud and error by benefit, between fraud and error and by type of error, and show the estimates for cases where there is high or low suspicion of fraud. [11]
  5. Figure 3: Income Support - Losses from fraud and error

     

    Type of Loss

    1997-1998

    1998-1999

    1999-2000

    2000-2001

     

     

     

    Amount

    As % of spending

    Amount

    As % of spending

    Amount

    As % of spending

    Amount

    As % of spending

    Fraud

    596m

    5.10%

    549m

    4.70%

    559m

    4.40%

    573m

    4.30%

    No. of cases

     

     

     

     

     

     

     

     

    216,000

     

     

    204,000

     

     

    Customer Error

    131m

    1.10%

    140m

    1.20%

    147m

    1.20%

    154m

    1.10%

    No. of cases

    Not available

     

     

    Not available

     

     

    301,000

     

     

    296,000

     

     

    Additional Official Error

    173m

    1.50%

    171m

    1.30%

    184m

    1.50%

    173m

    1.30%

    No. of cases

    Not available

     

     

    Not available

     

     

    493,000

     

     

    476,000

     

     

    Fraud and Error

    900m

    7.7%

    860m

    7.20%

    890m

    7.10%

    900m

    6.70%

     

    In addition, there are cases where there is suspicion of fraud, but no evidence

     

    High Suspicion of fraud

    Not available

    Not available

    236m

    2.00%

    206m

    1.60%

    125m

    0.90%

    Low Suspicion of fraud

    Not available

    Not available

    365m

    3.10%

    363m

    2.90%

    287m

    2.10%

    Source: Department for Work and Pensions[12]

     

    Figure 4: Jobseeker's Allowance (JSA) - Loss from fraud and error

     

    Type of Loss

    1997-1998

    1998-1999

    1999-2000

    2000-2001

     

     

     

    Amount

    As % of spend

    Amount

    As % of spend

    Amount

    As % of spend

    Amount

    As % of spent

    Fraud

    302m

    8.40%

    291m

    8.20%

    266m

    8.20%

    201m

    7.00%

    No. of cases

     

     

     

     

     

     

     

     

    120,000

     

     

    86,000

     

     

    Customer Error

    24m

    0.70%

    22m

    0.60%

    21m

    0.70%

    11m

    0.40%

    No. of cases

     

     

     

     

     

     

     

     

    32,000

     

     

    20,000

     

     

    Additional Official Error

    154m

    4.20%

    197m

    5.60%

    133m

    4.00%

    78m

    2.80%

    No. of cases

     

     

     

     

     

     

     

     

    125,000

     

     

    86,000

     

     

    Fraud and Error

    480m

    13.20%

    510m

    14.4%

    420m

    12.90%

    290m

    10.20%

     

    In addition, there are cases where there is suspicion of fraud, but no evidence

     

    High Suspicion of fraud

    Not available

    Not available

    111m

    3.10%

    77m

    2.40%

    28m

    1.00%

    Low Suspicion of fraud

    Not available

    Not available

    137m

    3.90%

    135m

    4.20%

    80m

    2.80%

    Source: Department for Work and Pensions [13]

  6. These figures exclude underpayments, which for Income Support totalled an estimated 180 million in 2000-01 and represented people not receiving benefits to which they are entitled. The Department are looking at issues that cause these underpayments, which arise from customers not notifying changes in circumstance or errors by staff, and addressing them through their performance improvement teams and in their training programme, which is being rolled out to all staff.[14]
  7. Of the overall estimate of 2 billion in fraud, the Department consider that organised fraud totalled around 100 million in 2000-01. It falls under two main headings:

  • Instrument of payment fraud (around 77 million), for example thefts of order books from within the Consignia delivery systems, counterfeiting of instruments of payment and manipulation of the payment instrument to gain higher amounts. This had fallen significantly, after the introduction of the order-book control service in 2000-01.

  • Identity fraud (under 20 million), including hijacking of legitimate identities, and using false documentation to establish numerous false identities after entry to the United Kingdom. It can reflect the activities of larger criminal networks engaged in people and drug smuggling.[15]

  1. The Government has set the Department a target of reducing the percentage loss on Income Support and Jobseeker's Allowance to 8.1 per cent of expenditure on these two benefits by March 2002, from a baseline of 9 per cent (1.38 billion) in 1997-98. By reducing estimated losses in 2000-01 to 7.4 per cent (1.2 billion), the Department had already met their first milestone by 31 March 2001.[16]
  2. Further milestones are to reduce losses to 6.75 per cent by March 2004 (a 25 per reduction on the 1997-98 baseline) and 4.5 per cent by 2006 (a 50 per cent reduction).[17] The Department see this as very demanding. They have in place a counter fraud strategy and have paid serious management attention to improving performance across the board on the administration of Income Support. In addition, the development of Jobcentre Plus should provide more face to face interaction with customers and new ways of working, including a more active case management approach. They recognise the risk that Jobcentre Plus might divert attention from tackling fraud and error and operational performance, and are managing that change in stages.[18]
  3. For Income Support, the Government has also set the Department a target to achieve 87 per cent accuracy in Income Support awards. This relates to the percentage of individual cases where errors are made by the Department's staff, rather than customer error and excludes small errors below 50 pence a week. In 2000-01, the Department exceeded for the first time the accuracy target. Moreover, in 2001-02, by the time of our hearing, accuracy had further improved to 91.5 per cent, on a comparable basis, which was a further step in the right direction.[19]
  4. The Department have rolling reviews to continuously assess the levels of fraud and error in Income Support and Jobseeker's Allowance, and targets to reduce it, but not for other benefits. The Department have focused much of their effort on these two benefits, which together with Housing Benefit they view as the most vulnerable to fraud and error. For other benefits, the estimates in Figure 1 above are based on snapshots—national benefit reviews, using smaller samples—taken at various times or pilot exercises, for example on Attendance Allowance. For Housing Benefit, however, the Department have embarked on work with local authorities to review 10,000 cases, to get a robust measure and a baseline of fraud.[20]
  5. One result of the scale of fraud and error is that the Comptroller and Auditor General has qualified his opinion on the Department's Accounts for the last 12 years. For every factor leading to the qualification, the Department have improvement targets, and achieving the Secretary of State's targets for reducing fraud and error will help reduce both the level of error and the scale of irregular expenditure in the accounts. However, removing the qualification completely will take some time.[21] The Comptroller and Auditor General has indicated to the Department that, to avoid qualification, it would need to reduce the level of fraud and error to below one per cent of gross expenditure.
  6. REDUCING FRAUD AND ERROR IN INCOME SUPPORT AND JOBSEEKER'S ALLOWANCE

  7. To further reduce losses from fraud and error in Income Support and Jobseeker's Allowance by 25 per cent by March 2004 and 50 per cent by 2006, the Department have set in train a range of initiatives.
  8. In their counter fraud strategy (April 1999), they introduced area benefit reviews to improve estimates of error and fraud; a programme of more rigorous checking of evidence supplied by claimants and new powers to obtain information from third parties; and the establishment of a new National Fraud Strategy Unit.[22]
  9. Under their Performance Improvement Strategy (February 2000), they gave early priority to improvement in poorer performing offices focusing on error and fraud, accuracy and clearance times. Key actions include establishing a performance improvement framework with Performance Improvement Action Teams working within each Area Directorate focusing on the resolution of local problems and the spreading of good practice.[23]
  10. Drawing on our predecessors' examinations and the Report by the Comptroller and Auditor General, we looked at the:
      • impact of complexity on error, and the scope to simplify Income Support and Jobseeker's Allowance;
      • the development of new information technology;
      • effectiveness of sanctions to deter fraud;
      • progress in reducing performance variations across the Department's offices.

    (a) The impact of complexity on error, and the scope to simplify Income Support and Jobseeker's Allowance

  11. In their reports on the Department's Appropriation Accounts for 1997-98 and 1998-99,[24] our predecessor Committee concluded that a key cause of error and fraud in Income Support and Jobseeker's Allowance was complexity of the regulations. In his latest Report, the Comptroller and Auditor General noted that the regulations are complex because they cater for the wide range of circumstances in which individual claimants find themselves. In turn this adds to the complexity of administration. The many millions of changes in claimants' circumstances each year affecting the rate of benefit in payment, some of which are not reported by the claimant when they should be, give rise to risks of errors by staff, and errors or fraud by claimants.[25]
  12. Despite the Department's efforts to simplify benefit rules and their administration, amendments and additions to already complex rules over time have exacerbated the problem. Significant simplification would require major changes in policy and legislation. Payments would be less closely related to individual needs, which could potentially increase costs. In the short term at least, it is probable that transitional payment arrangements to protect claimants' existing entitlement would have to be implemented. [26]
  13. The Department assured us that every time they considered an area of new policy they looked at the administrative consequences and the implications for fraud and error of making the change. Complex benefits are difficult to administer tautly and Income Support depends on self-declaration of means and circumstances. However, in some cases where more money was available, steps had been taken to reduce complexity. For example, when the Pension Credit is introduced, with a more stable caseload, payment of benefits will be for longer periods, thereby not putting people in the position of having to report changes in circumstances so frequently.[27]
  14. The Department have changed their approach to focus on client groups, for example in creating new agencies for pensioners and the working age. They had already streamlined the claim process for pensioners by simplifying the Minimum Income Guarantee claim form. A helpline, a teleclaims service where staff could talk people through applications, had delivered a 40 per cent improvement in accuracy, and electronic claiming offers another way of directing people to answer questions accurately. They plan to work with bodies like Age Concern and Citizens' Advice Bureaux to help elderly pensioners complete the forms. For people of working age, the Jobcentre Plus model involved personal advisors and financial assessors checking the information.[28]
  15. (b) The development of new information technology

  16. The Department's inadequate Information Technology systems continue to constrain improved performance. As part of their Information Technology strategy, the Department aim to replace existing systems for calculating benefit entitlement on Income Support by 2006. The Department expect the new system to provide common access to benefit claimant data currently held in over 20 separate systems, and to maximise the automation of benefit calculations and reduce the need for clerical interventions. The major challenge facing the Department is to design and implement a system that can cope with the complexity of the rules governing the benefit, and the linkages to other benefits. Until this happens the Department will continue to rely heavily on clerical intervention.[29]
  17. The Department are breaking down the process of modernising their systems into modules, to avoid previous problems with large, over-ambitious, rushed and badly managed systems, starting with the Child Support system. They had a welfare modernisation fund of some 2 billion over three years, but it would take some years to replace the back office legacy systems, which were huge.[30]
  18. (c) The effectiveness of sanctions to deter fraud

  19. The Department investigate some 450,000 cases of possible fraud in Income Support and Jobseeker's Allowance each year, and in 2000-01around 182,000 of these led to a change in the benefit rate. In some 24,000 of these cases, the Department had found sufficient evidence to apply a sanction, an administrative penalty of some kind. They took around 10,000 cases to court, and obtained convictions in 98 per cent of cases.[31]
  20. The Department's strategy is to prevent fraud occurring in the first place through deterrence and tight checks at the gateway to benefit. Once in payment, they seek to reduce exposure to loss by additional checks in cases judged most at risk and intervening quickly where there are early indications of fraud or error. Then, in cases where there is good reason to suspect that significant fraud is occurring, they investigate in depth with a view to prosecution or other sanctions.[32]
  21. As a general rule, first time frauds of less than 400 are offered as cautions rather than prosecutions. Above that mark, but below 1500, the Department would usually offer an administrative penalty in a first time case. Cases where fraud has been proved before, or the loss is over 1500, are normally sent for prosecution. In 2000-01, the Department had a 98 per cent success rate with prosecutions.[33]
  22. The Department have strengthened the sanctions they can apply to those previously convicted. Sanctions for a first or second offence involve stopping benefit for a week. But under the Social Security Fraud Act 2001 anyone who is convicted twice can have their benefit suspended for 13 weeks. The Department estimate that of the number of people fraudulently claiming Income Support and Jobseeker's Allowance, 26,000 have been caught before representing 0.5 per cent of the customer caseload. These cases involve mainly instrument of payment fraud. Following the changes introduced in the Social Security Fraud Act 2001, the Department are collecting additional data on repeat offences from April 2002.[34] All law enforcement agencies rely heavily on the support of members of the public and on information received from them. In a future report we intend to examine carefully the extent of information on social security fraud supplied to the Department by the public, and how effective the Department is in putting this information to best use.
  23. (d) Progress in reducing performance variations across the Department's offices

  24. There remains wide variation in the levels of error and fraud across the Department's 13 Area Directorates and 118 districts. For example, the Department estimate that error and fraud as a percentage of programme expenditure in the poorest performing Area Directorate (Chilterns, encompassing part of London) is double the level in the best performing (East Scotland). At district level, payment accuracy[35] rates range from 79 per cent to 97 per cent. As a result of the Department's performance improvement strategy, the number of Districts below the minimum benchmark standard of 80 per cent accuracy fell from 10 to 2. In addition, the lowest level of accuracy achieved by a district increased from 72 per cent in April 2000 to 79 per cent in April 2001.[36] Evidence was submitted to the Committee on the Department's prosecution policy. Given the level of fraud, the Committee does not believe the Government has achieved the right balance between prosecution and administrative sanctions against claimants suspected of fraud.
  25. The most important causes of variation in the performance of districts relate to the experience of the administrative staff and the organisation of their workload. External factors such as labour market competition for new and existing staff affect performance but to varying degrees. These hit districts hardest in London, the South East of England and the major cities, though even there some local offices have secured performance improvement in recent years.[37]
  26. The Department have put in place a range of strategies to tackle the underlying causes of variations in performance. Area Directorates have submitted new performance improvement plans with significant performance improvement targets supported by an additional fund of 26 million, which local offices could bid against for innovative ideas to improve performance. New technical and performance improvement training packages and programmes are being rolled out to frontline staff and managers drawing on the practices adopted by the better performing offices. Work continues on mapping out and generating new management information on how Income Support work can be organised and staffed to aid the identification of best practice, as does the initiative to twin good and bad performing offices. The aim is to achieve greater sharing of good practice and consistency across offices and target resources on those areas most in need.[38]
  27. The Department have recognised the need to support performance improvement by tackling key human resource problems. The Department have identified those offices where they face acute staff recruitment and retention difficulties. A number of short-term measures to assist these offices, such as diverting work away to other parts of the country and the use of funding to support local staff recruitment and retention initiatives, have helped ease some of the most immediate problems. The Department have reduced the time it takes staff to reach the maximum on pay scales, and are in the process of looking more widely at how the organisation of the work, better support for newly recruited staff and differential pay can help achieve longer-term improvements in performance.[39]

 


6   Qs 130-145, 247; Ev 25-26 Back

7   Ev 25 Back

8   Ev 25-26 Back

9   Cm 4012 Back

10   Qs 130-145, 247; Ev 26 Back

11   Qs 44-52, 63-66, 72-81, 129-145, 247; Ev 23-24, Appendix 2. Back

12   Ev 23, Appendix 2 (ref. to Q1) Back

13   Ev 24, Appendix 2 (ref. to Q2) Back

14   Qs 16-21, 34-36, 39-41 Back

15   C&AG's Report, para 2.7; Qs 229-239, 242-243; Ev 21-22 (ref. to Q239) Back

16   C&AG's Report, para 2.4  Back

17   ibid Back

18   Qs 2, 9 Back

19   C&AG's Report, para 3.11; Qs 22 -24, 39, 43 Back

20   Qs 53-55; Ev 24 (point 1, Losses in Income Support and Jobseeker's AllowanceBack

21   Qs 102-114, 122-128, 248-252 Back

22   C&AG's Report, para 3.4  Back

23   ibid Back

24   Committee of Public Accounts: 3rd Report, Department of Social Security Appropriation Accounts 1997-98, Class XII, Vote I, Central Government Administered Social Security Benefits and other Payments (HC 103, Session 1999-2000); 32nd Report, Department of Social Security Appropriation Accounts 1998-99, Class XII, Vote I, Central Government Administered Social Security Benefits and other Payments (HC 521, Session 1999-2000). Back

25   C&AG's Report, para 3.7  Back

26   ibid, para 3.8  Back

27   Qs 3-5, 146 Back

28   Qs 42, 147-152, 211 Back

29   C&AG's Report, paras 3.9-3.10  Back

30   Qs 6-8, 214-220 Back

31   Qs 63-71, 196-199 Back

32   Qs 191-193; Ev 20-21, Appendix 1 Back

33   Qs 179, 191-193, 199; Ev 20-21, Appendix 1 Back

34   Qs 94-98, 168-178, 201-203; Ev 21 (ref. to Q239) Back

35   The percentage of individual cases where errors made by the Department's staff are identified as part of internal quality assurance work. Back

36   C&AG's Report, paras 3.5, 3.11  Back

37   C&AG's Report, para 3.12  Back

38   Qs 12-14, 213 Back

39   ibid Back

 
previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2002
Prepared 11 September 2002