House of Commons - Public Accounts - Twenty-Seventh Report

Select Committee on Public Accounts Twenty-Seventh Report

MEASURES TO COMBAT HOUSING BENEFIT FRAUD

WHAT ARE THE DEPARTMENT DOING TO TACKLE FRAUD?

New Powers to Combat Fraud

23. The Social Security Administration (Fraud) Act, which received Royal Assent in March 1997, has implications for all social security benefits, but many of the key measures focused on Housing Benefit. These include: improved access to data, and the sharing of data between the Department and local authorities; better access by local authorities to information held by each other, by landlords and by the Post Office; new criminal offences and improved powers to recover money; and powers to examine and report on how well local authorities administer Housing Benefit, through the newly created Benefit Fraud Inspectorate.[24]

24. We asked the Department whether these new powers were sufficient to enable them to deal with Housing Benefit fraud. They said it was difficult to tell. They had fairly extensive powers over data matching, over landlord fraud, and over administrative penalties and offences, and had set up the Benefits Fraud Inspectorate. They told us that they needed to get these powers fully into use and then see whether there were other areas they needed to move into.[25]

25. As regards their influence over local authorities, the Department said that they had stronger powers in the Fraud Act. If, following an inspection by the Benefits Fraud Inspectorate, the performance of a local authority gave sufficient cause for concern, the Secretary of State could give directions to the authority to bring their performance up to standard. If it did not do so, there were powers to impose subsidy penalties or even to direct that the authority should contract out part of their Housing Benefit operation. However, the Department saw these as powers of last resort, and hoped that their existence would bring about the desired result.[26]

The Complexity of Housing Benefit

26. The complexity of the Housing Benefit regulations increases the risk of confusion and error on the part of claimants and on the part of benefit assessors when determining claimants' entitlement. For example, there are over 60 different allowances, premiums, deductions and disregards attached to the benefit and one standard text on the interpretation of the regulations runs to nearly 500 pages.[27]

27. We asked the Department what consideration had been given to simplifying the benefit to reduce the opportunities for fraud and error. They acknowledged that the scheme was complex, although they pointed out that when it was introduced in 1988 it was a simplification of what had gone before. One of the reasons it was complex was that it was a means-tested scheme, and was designed to give people in different categories, such as elderly people and disabled people, different amounts of benefit. The Department accepted that it would be possible to have a simpler scheme, and thereby reduce the scope for fraud, but it would not be so finely tuned to people's individual needs and resources and would be harsher in some respects, or more expensive.[28]

28. They added that Housing Benefit was being looked at as part of the Government's Comprehensive Spending Reviews, and that one thing they would be looking at was the proneness to fraud. The Review would be reporting in Spring 1998.[29]

The Impact of Fraud in Other Social Security Benefits

29. Some 66 per cent of claimants on Housing Benefit are also in receipt of Income Support or income-related Jobseeker's Allowance. When a claimant for Housing Benefit receives Income Support or Jobseeker's Allowance, local authorities do not have to calculate the claimant's applicable amount of income when assessing entitlement to Housing Benefit.[30]

30. In July 1995, the Department published their first fraud review of Income Support and Unemployment Benefit. It revealed that frauds againt Income Support cost the taxpayer £1.4 billion a year, and that 1 in 10 claims were fraudulent. More recently, they have estimated that this fraud has risen to £1.77 billion a year.[31] In their review of fraud in Housing Benefit, the Department estimated that the knock-on effects of fraudulent claims for Income Support on Housing Benefit could amount to £178 million, almost 20 per cent of the total estimate of fraud.[32]

31. The Department told us that over the last three or four years the Benefits Agency had been putting a lot of effort into tackling accuracy in Income Support, and they had a series of local plans to try to raise the level of accuracy. They believed that the pendulum had swung too far in the direction of claims clearance without enough attention being given to accuracy but there had been a big change of emphasis within the Agency and after a very bad period it was beginning to improve.[33]

Detecting Fraud

32. In 1996, the Audit Commission reported that the amount of Housing Benefit fraud detected had increased fourfold during the 1990s to £55 million in 1995-96.[34] Local authorities have reported to the Department increasing levels of savings from their anti-fraud activities, rising from £91 million in 1993-94 to £306 million in 1996-97.[35]

33. The Department's main established measure for combating Housing Benefit fraud has been an Incentives and Penalties scheme which rewards local authorities for detecting fraud, based on estimated weekly benefit savings on claims stopped. The scheme has led to significantly greater efforts in combating fraud, and in 1996-97 the Department paid local authorities £43.6 million in subsidy based on their achievements. But the scheme has a number of shortcomings. In particular it has not yet produced results that approach the estimated level of fraud; there is error and abuse of weekly benefits savings claimed which may artificially inflate the levels of fraud detected; and the scheme offers a few incentives to prevent or deter fraud.[36]

34. We asked about the effectiveness of the scheme. The Department told us that although the amount of fraud detected had increased to over £300 million in 1996-97, there was no evidence that the actual level of fraud itself had increased. Rather, detected fraud had increased because they had put in incentives to get local authorities to go and detect it.[37] The Department acknowledged that the scheme was not ideal and had some perverse incentives. But they felt that it had been successful in getting local authorities to detect a lot more fraud.[38]

35. The 32-week multiplier at the heart of the weekly benefit savings calculations may distort reported levels of fraud, detected, and it does not reflect the actual savings to the public purse achieved by fraud detection. The Department told us that the current multiplier had the advantage of being easy to operate, and was based on research on the average length of time which Income Support cases would have lasted if fraud had run on. They acknowledged that the same figure did not apply to all benefits, and told us that one aim of the local benefit reviews they were undertaking was to look at whether they could find a more accurate multiplier for Housing Benefit.[39]

36. There were doubts about the validity of weekly benefits savings claimed in 30 per cent of the sample of claims examined by the National Audit Office/Audit Commission joint study team at the 14 local authorities they visited. A parallel exercise by the Audit Commission and the Benefits Agency showed a higher rate of questionable claims.[40]

37. The Department told us that the figures for detected fraud should not be wildly out, but they accepted that some of them had been overstated. They would be surprised if this cast doubt on the figure of 190,000 cases of detected frauds reported in 1996-97, but they felt that it might cast doubt on their total value.[41] In their view something like weekly benefit savings would always be required to measure the effect of individual detection initiatives even if they could find other measures and incentives to put alongside it.[42] In consequence, they were putting a lot of effort into making sure that the weekly benefits savings claimed were right. For example, they had issued new guidance about calculating savings, and the Benefit Fraud Inspectorate would look in their inspections at how authorities calculated them.[43]

38. The Benefits Agency also use weekly benefit savings to measure their anti-fraud performance, and they can claim savings where their investigative work on Income Support identifies cases where Housing Benefit is also in payment and fraudulently obtained. In 1995-96, the Agency claimed £193 million Housing Benefit weekly benefit savings out of the total £716 million they claimed across all benefits.[44]

39. Reviews by Internal Audit identified significant problems in the accuracy of the weekly benefits savings claimed by the Agency, and concluded that this would have resulted in significant overstatement of their achievements. In response, in 1996-97 the Agency reduced their figures for weekly benefit savings achieved from £1,870 million to £1,524 million.[45]

40. In answer to our enquiries, the Department admitted that the accuracy of the savings claimed by the Agency had not been good in the past, but they had issued new guidance to the Agency and were now having claims checked both locally and at area level. They said that the overstatement rate was 18 per cent in 1996-97. The final reported performance figure of £1,509 million reflected the remaining 82 per cent.[46]

41. We asked what impact this overstatement of weekly benefits savings had on the performance related pay and bonuses paid to the Chief Executive and other staff. The Department told us that under his contract the Chief Executive could earn up to 15 per cent of his salary by meeting the objectives in his performance bonus scheme. Achievement of the security savings target contributed to his bonus award, which in 1996-97 was 10.5 per cent.[47]

42. The Department added that the Agency's Finance Director could also earn a bonus of up to 5 per cent of salary, of which 1 per cent related to the total fraud and security savings target for the Agency. For other staff, a proportion of their annual pay increase was determined by their annual performance appraisal rating which was determined by their performance against all key work objectives and demonstration of a number of competencies: for fraud investigators, these could include performance in meeting savings targets.[48]

43. We asked the Department what guidance they provided to local authorities on targeting anti-fraud investigations. They said that their fraud investigation manual issued in 1995 drew on figures available from the Benefit Review and therefore drew attention to the kinds of areas that they should be targeting. The guidance was being revised for reissue in 1998. However, the Department admitted that although some checking was done by the Audit Commission, they themselves had not been checking whether this guidance was being followed except during visits, and that these visits were not a real audit of performance. They recognised that they could do more and that was one of the reasons why the Benefit Fraud Inspectorate had been introduced to carry out a much closer check than occurred at present.[49]

44. The use of visiting officers to check claimants' circumstances can be very cost effective, offering returns of between 6 and 20 times their costs of £15,000 to £20,000 a year.[50] We asked whether the number of visiting officers was increasing and whether it would be cost-effective to have more. The Department confirmed that the number of full- time designated fraud officers in local authorities and visiting officers in the Benefits Agency had gone up considerably in recent years. Returns from local authorities for 1996-97 showed that 1,900 full-time staff had been specifically designated as fraud officers and that authorities also employed visiting officers to undertake residency checks and customer care duties.

45. The Department added that on quite a lot of occasions in recent years they had been given extra money by the Treasury to produce specific extra results. Of the £16 million Challenge Fund money this year, £8 million was being given to local authorities to employ people for visits to claimants. In this packaged way both the Department and the Treasury could see the outcome of putting in more resources.[51]

46. In briefing to Parliament during the passage of the Social Security Administration (Fraud) Act the National Association of Citizen's Advice Bureaux drew attention to evidence which suggested that innocent people who had been overpaid benefits might be suffering as a result of over-zealous anti-fraud strategies. None of the authorities visited by the National Audit Office had a code of conduct for their fraud investigators.[52]

47. We asked whether compensation would be paid to people who had suffered trauma as a result of gross errors in fraud investigations carried out. The Department said that, as local authorities were responsible for the administration of Housing Benefit it would be for them to decide what they should do. If the authorities did not settle the matter themselves it would be for the local authority ombudsman to deal with it. The Department had its own redress schemes and there would be compensation if such cases happened in the Benefits Agency.[53]

Preventing Fraud

48. The Department's anti-fraud measures on Housing Benefit have focused on encouraging local authorities to detect fraud.[54] The Audit Commission reported in 1993 that perhaps only a third of authorities largely administer benefits properly, efficiently and effectively and that some authorities were reluctant to invest the necessary resources to prevent fraud.[55] In his report, the Comptroller and Auditor General noted that there continued to be highly variable standards of benefit administration by local authorities, and that to a degree the Incentives and Penalties scheme encourages, or at least leads to the toleration of, weak administration. This is because authorities can get subsidy for detecting fraudulent claims which enter the system.[56]

49. In response to our enquiries, the Department accepted that they needed measures and incentives to prevent fraud in the first place.[57]

50. The Department told us that since the Committee's previous report in 1990 they had done a great deal to bring the standards of administration in all authorities up to the level of the best. They had progressively issued more good practice guides, and in 1993, they had introduced the incentive scheme whereby authorities could earn subsidies by detecting fraud. They had introduced monitoring of local authority performance, and they had made visits to authorities to try to help them with their performance. They had also produced service level agreements to facilitate working with the Benefits Agency, and they had issued a fraud investigation manual.[58] They felt that their work to put in incentives and provide guidance had begun to pay off,[59] and that the standards of administration had improved considerably in recent years.[60]

51. The Department added that they were now putting in a verification framework which would give authorities guidance about exactly what information they should be getting at the start of a claim to ensure it was correct, and how they should carry out checks and reviews later.[61] At the end of the year there would be an auditor's certificate that they had complied with the framework.[62]

52. We asked the Department what they were doing to pinpoint those authorities which were performing really badly. They told us that the new Benefit Fraud Inspectorate would be their main tool on that and they would get a series of reports on the performance of individual authorities. The Inspectorate's first full inspection was currently underway and the Department expected the Inspectorate's reports to be published once cleared with the local authorities concerned.[63]

53. The most frequently expressed local authority grievance is that the Department do not meet their costs of administering Housing Benefit in full, and that local authorities bear the full marginal cost of any increases in expenditure to improve administration. [64]The Department told us that they used to pay authorities their actual costs but they found this to be not a very good system because the costs shot up year after year.[65] They now paid around £300 million as a basic administration subsidy, which fell short of the £400 million authorities said they spend; this sum was the Government's view of what local authorities needed to administer Housing Benefit and it was then up to each authority to decide how it spent its money. The Department pointed out that there was an enormous range of cost per case in the figures authorities provided, and that they found it extremely difficult to assess what an efficient administration cost was.[66]

54. The Department added that in addition to the separate incentive schemes for detecting fraud and recovering overpayments, local authorities which signed up to implement the verification framework would also get funding, so that there would be a straight incentive to apply that system in their Housing Benefit administration. In these ways, they were trying to ensure that extra money could be geared to things which directly improved administration rather than just going into a general pool. They hoped to start that soon: the framework was ready, they just had to reach agreement with local authorities.[67]

Deterring Fraud

55. There are limited incentives to deter fraud, and there has been a low level of prosecutions by local authorities. Of the 380 local authorities now administering Housing Benefit 182 authorities undertook no prosecutions at all in 1996-97.[68]

56. We asked the Department about the low level of prosecutions. The Department told us that the evidence they had was that fear of detection, and a very tight policing system, was actually a stronger disincentive than fear of prosecution, because the penalties applied by the courts did not tend to be very high. And they pointed out that something like a third of those convicted when the Benefits Agency prosecuted just got a conditional discharge.[69]

57. The Department added that local authorities have been reluctant to prosecute. Some were reluctant in principle, despite Audit Commission advice that they should and guidance from the Commission and the Department on how to go about it. Prosecutions were very low; particuarly compared to the Benefits Agency which prosecuted some 12,000 to 13,000 cases a year, and the Department told us that they wanted to encourage local authorities to prosecute more.[70]

58. We asked the Department what they could do to increase the level of prosecutions, and whether they had any contact with the courts on sentencing policy to discover whether there were regional variations in sentencing. They told us that they and the Audit Commission had given authorities advice, and that they were looking to see whether they could offer the use of the Department's solicitors. In addition, the Benefit Fraud Inspectorate would be looking at authorities' prosecution policies. However, the Department was reluctant to offer financial incentives because that might lead authorities to prosecute people who ought not to be prosecuted. They said that they did not know whether they had any information about regional variations.[71]

59. Of the 385 claims for weekly benefit savings examined by the National Audit Office and the Audit Commission, overpayments had been identified in 164 cases, but recovery had been attempted in only 38 per cent of these. And only one of the 14 authorities visited maintained a central record of the status of Housing Benefit overpayments.[72] The Department told us that local authorities could recover such overpayments and they believed that they did so in about 60 per cent of cases. They added that, under the 1997 Fraud Act, the Department and local authorities would be able to offer administrative penalties as an alternative to prosecution where there was sufficient evidence to prosecute. The penalty would be related to the amount of the fraudulent overpayment.[73]

Specific Measures to Address Landord Fraud

60. Nine of the local authorities visited by the National Audit Office said that, although they had indications of landlord fraud, they did not pursue them because of the opportunity costs involved. Diverting investigators to complex fraud cases was not economically viable, and the Incentives and Penalties scheme encouraged the pursuit of "easy pickings".[74]

61. The Department told us that initiatives already introduced, such as the Fraud Hotline and data matching, would help tackle landlord fraud. There was also a number of new measures in the Fraud Act. These included provisions for local authorities to enter business premises, to find details of other properties owned by the same landlord and to stop direct payments to landlords under suspicion.[75]

62. The Department had also set up and financed the London Organised Fraud Investigation Team in 1996. They had established this first pilot in London because Benefits Agency evidence suggested that 75 per cent of all detected organised fraud occur there. The Team had a commitment to supply the Department with monitoring and evaluation data to enable a report of their results to be made available by the end of January 1998.[76]

63. As regards prosecutions, the Department told us that they did not know how many landlords had been prosecuted. They agreed that in cases where landlords had committed a substanial fraud there ought to be prosecution. This would normally be for the local authority to decide,[77] though where the Department was investigating fraud against Income Support and in the same case there was also fraud against Housing Benefit, Departmental lawyers could prosecute those accused of Housing Benefit Fraud including landlords.[78] The Department added that the London Organised Fraud Investigation Team had had one or two successful investigations that had led to arrests and the likelihood of prosecution.[79]

Conclusions

64. We are worried at the Department's uncertainty about whether the new powers contained in the Social Security Administration (Fraud) Act 1997 will be enough to enable them to deal effectively with Housing Benefit fraud. This uncertainty reinforces our concern that the Department do not have sufficient grip of what is needed. We expect them to monitor closely the impact of these new measures, and to take decisive action should further action be necessary.

65. Those designing benefit systems have to strike a balance between addressing different needs equitably and fairly on the one hand and simplicity and security on the other. However the present system does not find the right balance. There are 500 pages of guidelines as well as over 60 assorted deductions and allowances to understand. This complexity exposes genuine claimants to confusion, inadvertent error by omission, and the risk of investigation; confuses administrators leading to high levels of errors in payment; and provides a breeding ground for fraud. In these circumstances, the system can be neither fair nor secure, and the Department need to take a fundamental look at the scheme during their current spending review.

66. A significant level of fraud in Housing Benefit arises because of the failure by the Department and the Benefits Agency properly to control claims for Income Support and Jobseeker's Allowance. In effect, this failure provides a gateway to fraud in Housing Benefit. The Committee of Public Accounts have previously expressed serious concerns about fraud and error on Income Support, and recently we have examined this issue again and will report separately on what more needs be done.

67. The Department's main measure for increasing fraud detection is an incentives and penalties scheme which cost £43.6 million in subsidies to local authorities in 1996-97. This scheme has increased the resources devoted to fraud detection, but contains perverse incentives which encourage inflated claims of savings achieved by authorities and do little to encourage authorities to prevent or deter fraud. There has also been overstatement of savings achieved by the Benefits Agency. The Department are looking at how the system could be improved, and are searching for a more accurate way of assessing savings from anti-fraud work on Housing Benefit. It is worrying to note their lack of confidence that they will find a solution.

68. Accurate reporting of savings achieved is essential in assessing the success of, and returns on investment in, anti-fraud work, and the rewards to be paid to local authorities and to individuals in the Benefits Agency. We therefore look to the Department to secure more robust measurement, and independent validation, of savings reported. Investment in anti-fraud work can be very cost effective, and the Department have secured extra funding for specific initiatives on a `spend to save basis'. In our view, increased investment may be justified, but there must be greater assurance over the accuracy of savings achieved.

69. It is disturbing that innocent people may be suffering as a result of over zealous anti-fraud strategies. Those who commit fraud should be dealt with effectively, but the conduct of fraud investigations should also be properly controlled to inspire public confidence that suspects are treated fairly and that innocent people do not suffer. We expect the Department to take the lead in drawing up a code of conduct for fraud investigations.

70. The measures introduced by the Department have been effective in increasing the amount of fraud detected. But they have not helped to prevent fraud getting into the system in the first place. Poor administration in many local authorities is a contributory factor in the amount of fraud, and the Department must bear a large share of the responsibility for letting this happen. The Department are now seeking to improve Housing Benefit administration by introducing the verification framework, which will give authorities guidance on the checks needed to ensure that claims are correct. This is a step in the right direction. We are surprised, however, that implementation of the framework is to be on a voluntary basis. If this remains the case, the Department will have lost the opportunity to drive up standards and ensure consistency across the country. In our view, compliance with the new framework should be mandatory.

71. The Department are placing considerable emphasis on the newly established Benefits Fraud Inspectorate to drive up standards of administration and anti-fraud work. Inspections of individual local authorities' performance will also cover, for example, authorities' targeting of anti-fraud work, prosecution policies, and weekly benefit savings claimed. The Inspectorate's reports to the Secretary of State will also be published. We welcome this initiative.

72. Fear of detection can be an effective deterrent against committing fraud, but only if it is backed up by sufficient prosecutions and effective penalties. Yet the level of prosecutions by local authorities is incredibly low, at under one per cent of detected frauds, and fraudsters have a 99 per cent chance of getting off scot-free. The Department need to provide a stronger lead in identifying and resolving the obstacles faced by local authorities, in helping them to implement cost-effective prosecution policies, and in looking at regional variations in prosecution policy.

73. It is unacceptable that authorities often do not seek repayments from people who have obtained benefit fraudulently. We remain unconvinced that local authority performance in recovering overpayments is as good as it should be. The Department have introduced administrative penalties as an alternative to prosecution. But the effectiveness of the new penalties as a deterrent will be diminished unless authorities make more concerted and consistent efforts to recover overpayments. The Department should examine ways of making this happen.

74. The Department have introduced a number of measures to combat organised and landlord fraud, including the London Organised Fraud Investigation Team. We expect them to review the success of these initiatives to ensure they are securing the desired impact, and to assess the benefits of extending the specialist investigation of landlord and organised fraud to other parts of the country.

24 C&AG's Report (HC 164 of Session 1997-98) para 3.2 Back

25 Q 72 Back

26 Q 150 to 151 Back

27 C&AG's Report, para 1.16 and 1.23 Back

28 Qs 2-3, 50-51 Back

29 Q 51 Back

30 C&AG's Report (HC 164 of Session 1997-98) para 1.20 Back

31 C&AG's Report (HC 251-Appropriation Accounts 1996-97, Volume 12, Class XII) Back

32 C&AG's Report (HC 164 of Session 1997-98) para 1.32 and Figure 6 Back

33 Q 7 Back

34 C&AG's Report (HC 164 of Session 1997-98) para 1.30 Back

35 C&AG's Report (HC 164 of Session 1997-98) para 2.16 and Figure 6 Back