Select Committee on Public Accounts Third Report


5 Combating fraud

Improving case handling procedures

29. Prior to the Audit Office examination, the Department's record in combating fraud was pretty abysmal. Case handling procedures were weak and many suspected frauds had not been pursued, without any explanations on file as to why not. In the three years prior to 2000, only 16 cases of suspected fraud were referred to the Department's Fraud Investigation Unit. In three of these cases, no further action could be taken because they had become statute barred as a result of delays in referral. Only one of the 16 cases led to a successful prosecution. In most of the remaining cases, the Department felt there was insufficient evidence to proceed with a prosecution.[28]

30. Given the widespread failings in the Department's administration of the scheme and the low level of prosecution, we questioned whether the Department had sufficient resources to operate the scheme. We were told, however, that resources were sufficient and that staff actually carry out more inspections than required by the EU. In our view, this makes the shortcomings in the scheme all the more damning.[29]

31. The Department accepted that there had been a shoddiness in its record-making but said that improved procedures now ensure that proper documentation standards are adhered to. It pointed out that the level of prosecutions had increased—in 2002, 20 cases had been referred to the Fraud Investigation Unit, of which 8 had resulted in successful prosecutions with 3 more cases pending. It also said that it was working more closely with the Director of Public Prosecutions to ensure that evidence stood up in court. In addition, after a successful prosecution, the Department now issues a press release naming the farmer, where he lives, the nature of his offence and the penalty imposed.[30]

Addressing conflicts of interest and the risk of collusion

32. One of the attempted fraud cases reported by the C&AG involved a farmer who had claimed premium for 71 sheep but was found not to have any at inspection. What we found particularly worrying was that the case involved an employee of the Department. Prosecution was not recommended because, as regards the evidence, it was only the inspector's word against that of the farmer. In a case where a civil servant deliberately seeks to exploit the rules within his own Department, we are of the firm opinion that the Department must make a particular effort and be seen to take strong action. This should always include disciplinary proceedings and, wherever possible, prosecution.[31]

33. In evidence provided after the hearing, the Department said that 231 of its 3,762 staff are currently claimants of agricultural subsidies. Although the proportion is relatively small, the numbers involved are substantial. The Department told us that all staff are now required to sign a statement in which they declare whether there is any conflict between their own activities, or those of their family, and the work which they do for the Department. This is a useful procedure. However, it will only be effective if the Department carefully considers each situation where a potential conflict of interest exists and ensures that the risks are fully addressed, for example, through separation of duties and effective supervisory checks.[32]

34. We asked whether there was an increased risk of collusion between those administering the scheme and farmers, because the scheme involves European money. The Department said that it takes steps to ensure that collusion does not occur, including the rotation of some 20% of inspectors each year. Given the many examples in this scheme where non-compliance with the rules was ignored by the Department, we are by no means convinced that collusion has not been an issue.[33]

Shortfall in sheep numbers discovered during the Foot and Mouth cull

35. The most serious fraud cases were those discovered during the Foot and Mouth emergency in 2001. Following the cull of animals which resulted from the outbreak, significant discrepancies in claims for sheep premium were found in three locations. Of 206 claims submitted from these areas, 108 (52%) had shortfalls between the number of sheep claimed and those counted during the cull. This included 17 cases where no sheep at all were found. The overall figures were later revised to 80 claims (39%) because, in 28 cases, the cull records compiled by the Department were insufficient to enable the discrepancies with the claims for sheep premium to be substantiated. Numbers recorded during the cull had not been confirmed in writing with the applicants, nor had there been a cross-check against flock records. It is extremely unsatisfactory that the 28 claims involved had to be paid in full.[34]

36. In one of the areas affected by Foot and Mouth, it is clear that at least 58% of claimants had been attempting to cheat the system. We found this a quite staggering state of affairs. As the attempted frauds only came to light through the Foot and Mouth outbreak, we were concerned as to how long this type of activity had been going on, how widespread it was and why the problem had not been flagged up by the Department's risk analysis. The Department said that it did not believe that the findings in this particular year were typical and suggested that it may have reflected an opportunistic response in the border area to a high market price for lambs in the Republic of Ireland in early 2001. We are not at all persuaded by this. In our view, the Department's risk analysis is fundamentally flawed and needs to be re-examined as a matter of urgency.[35]

Estimating the total level of fraud

37. We were particularly interested in how much fraud there was across the Department as a whole. Surprisingly, the Department was unable to provide us with figures, but did say that it was in the process of developing a methodology to estimate this. As regards the Sheep Annual Premium Scheme, it had estimated fraud at some £0.5 million each year, based on an average 3% annual shortfall in sheep numbers noted at inspections. In our view, this estimate is almost certainly understated. We saw plenty of evidence that many irregularities, including shortfalls of sheep, have gone unreported. The actual fraud figure for the scheme, therefore, would be substantially higher than £0.5 million. We would strongly urge the Department to look again at this matter.[36]

38. Data provided by the Department showed that, over the past 10 years, it has spent almost £1.4 billion on livestock subsidies, including £178 million in the current year. With such enormous sums of taxpayers' money at stake, the Department must make fraud a decidedly unattractive proposition, by ensuring that it has an effective inspection process, a rigorous prosecutions policy and penalties that far outweigh the potential gains from fraud.[37]

Discrepancies highlighted by the Livestock and Meat Commission

39. The Northern Ireland Livestock and Meat Commission recently highlighted the lack of reconciliation between the total number of marketed sheep, estimated by the Department from its annual census statistics, and the actual numbers recorded in various market outlets. For 2001, the Department estimated that 1,382,000 sheep were marketed, whereas the Commission's statistics for recorded marketings was 984,000, some 400,000 less. The Commission was concerned that these sheep may not have been there to begin with, thus casting doubt on the accuracy of census returns and claims for sheep annual premium. The Department said that, with improved controls introduced in the wake of Foot and Mouth, it can now reconcile the difference. Also, because it now has actual data on all marketings, it no longer needs to produce an estimate. It accepted, however, the inaccuracy of the methodology it had been using to estimate marketings. We commend the Livestock and Meat Commission for drawing attention to the substantial inaccuracies within the Department's statistics.[38]

General comments

40. Having carefully examined the evidence, we are convinced that many fraudsters would have regarded an attempt to cheat the scheme as a risk worth taking. The reasons are apparent—six out of seven claims not selected for inspection, poor risk analysis, a slack inspection regime, a very low prosecution record and relatively modest penalties. The low level of risk to fraudsters was most starkly illustrated in the Foot and Mouth cull cases, where the 17 farmers who were found to have no sheep at all could not be prosecuted because of a loophole in the Department's legislation.[39]

41. Our overall impression is that the Department has in the past been soft on fraud and this has contributed to unacceptably high levels of fraud within Northern Ireland agriculture. The Accounting Officer has assured us that his Department now operates a policy of zero tolerance to fraud. Among the initiatives introduced are a counter-fraud strategy and an anti-fraud forum. In addition, clear instructions have been now issued to all staff and training courses have been provided. We acknowledge that these are positive steps in the fight against fraud. However, the procedures will only work if they are properly applied. In our review of the Sheep Annual Premium Scheme, we noted that many of the weaknesses arose out of the Department's failure to properly apply existing control procedures.[40]


28   C&AG's Report, paras 4.15-4.19, 4.26-4.30  Back

29   Qq 75-76, 85 Back

30   Qq 14-15, 62, 86-87 Back

31   Q 83; C&AG's Report, para 4.27 and Prosecution Case B

 Back

32   Qq 103-107 Back

33   Qq 84-85; C&AG's Report, para 2.10 Back

34   Q 11; C&AG's Report, paras 2.38-2.43 and Case C, p 52 Back

35   Qq 11-13, 73, 90-102, 131-134, 142-146; C&AG's Report, para 2.47 Back

36   Qq 150-158 Back

37   Ev 17 Back

38   Qq 20-43, 135-141; Ev 15-17 Back

39   Qq 66-73, 108-116; C&AG's Report, paras 4.31-4.32 Back

40   Qq 14, 63-65, 168-171 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 2003
Prepared 18 December 2003