Audit Committee,

Meeting date: Tuesday, May 30, 2006

Official Report 303KB pdf

---

National Fraud Initiative

For agenda item 2, we will hear a briefing from Audit Scotland on the recent publication "No hiding place: the National Fraud Initiative in Scotland". I invite the Auditor General for Scotland to introduce the report.

It might be helpful if I take a moment to explain the nature of the report. "No hiding place: the National Fraud Initiative in Scotland" is not a formal report from the Auditor General to Parliament but a paper that Audit Scotland has prepared that arises out of its auditing of public bodies in Scotland. The report captures the results of a major data matching exercise in some key audited bodies that was facilitated by Audit Scotland with the help of the Audit Commission, which has run similar exercises for a few years now in England and Wales.

Russell Frith, who is director of audit strategy for Audit Scotland, led the project for us. I invite him to give a quick summary of the project and its key findings.

I briefed the committee about 18 months ago, before we started the exercise, so this is a follow-up to report on the results of the exercise.

All public bodies have a duty to minimise fraud and overpayments. As part of our audit of public bodies, we undertook the exercise with the aim of helping them to improve their efficiency and effectiveness in doing that. The exercise involved all 32 councils, the joint police and fire boards, the Scottish Public Pensions Agency and the Student Awards Agency for Scotland. The data that were matched included payrolls, the local government, national health service and teachers pension schemes, information about students and the Department for Work and Pensions register of deceased persons. The exercise was carried out with the assistance of the Audit Commission, which has carried out similar exercises in England and Wales. The Audit Commission will publish the report on its latest such exercise this morning.

As a result of the exercise, we found £15 million of overpayments, fraud and forward savings. That compares to £96 million in England and Wales. That total splits roughly into £5 million in housing benefit, £5 million in the local authority pension scheme and £5 million in the NHS and teachers pension schemes. The total can also be split between overpayments of about £6 million and forward savings of about £9 million.

The results included 270 cases of occupational pensions that were paid to deceased persons, 215 cases of housing benefit overpayments to students and 564 cases of housing benefit that was paid wrongly to public sector employees or pensioners. Of those, 53 cases have either had prosecutions instigated or been referred to the procurator fiscal for further consideration.

Those bodies in which very little fraud or overpayment was found can take positive assurance from the exercise about the integrity of their staff and the quality of their systems. We believe that the exercise was worth while whatever the outcome for particular bodies.

We believe that the exercise was successful and we intend to run it again from October this year, when we hope to widen the range of bodies that take part. In that respect, we have had constructive discussions with the Scottish Executive Health Department with a view to using NHS data and we hope to have similarly constructive discussions with the Scottish Executive shortly. We also intend to widen the data sets to include things such as blue badges and care home payments.

That is all that I will say by way of this short briefing, but I am happy to answer questions.

Members now have the opportunity to ask questions.

If my arithmetic is right, it sounds as if Scotland has a proportionately greater problem of overpayment than England and Wales do. Is that right? Is there a reason for that?

It is difficult to make a conclusion at this stage, as this was the first such exercise that we have run in full across Scotland, although we piloted some elements of it two years ago. It might well be that, in this case, we are catching up with a number of longer-standing overpayments or frauds that would have been picked up in England and Wales in previous exercises. We are probably covering cases over a longer time period this time. If the results are still higher next time, the conclusion that you draw would be reasonable.

Fifty-three prosecutions out of a £15 million overspend does not seem to be that many. Does that suggest that it is more a case of systems not working than deliberate fraud, or is it just that there were only 53 cases in which the allegations would stick?

We do not have a great deal of information about that, but we know that all the local authorities consider carefully the circumstances of the cases. For example, if a very elderly pensioner claims housing benefit without declaring some or all of their occupational pensions, the local authorities tend to decide not to prosecute. Filters are applied.

I would like further clarification on the cases that were reported to the fiscal. During the past few weeks, some of those cases have come to court and they involve quite substantial figures. Do you have any information that would provide us with an average figure for those 53 cases?

No, we do not have the information at that level on each of the cases.

Would it be possible to get that information in the future so that we could see it at a glance, as well as some of the case studies? I have seen reports of some of the cases in the local paper. One involved an employee and housing benefit; the sum in that case was somewhere in the region of £7,000. At the weekend, there was a story about a family and the DWP and the figure was somewhere of the order of £40,000 because the family were continuing to claim. It would be helpful if we could have a breakdown of the sums involved in those prosecutions.

We will certainly try and do that for the future.

We need to be careful about data protection, particularly if there are current court proceedings. We are talking about a small number of cases.

Yes, but it would be helpful if we could see a range of the figures involved.

Are the police and the fire and rescue services included in the areas that you have looked at?

Yes, they are.

What, if any, were the common themes that you found in the difficulties that arose?

Some of the matches that we are doing are ones that the local authorities could not have done for themselves. We have been helping the authorities to do things that it would be difficult for them to do. I would not say that the matches that we have seen show particular weaknesses in the authorities' systems, because we have been able to cross-match data that cannot usually be matched.

In some of the pension cases, there is a lack of clarity of understanding of what a pensioner can and cannot do. There are commonly held beliefs—unfortunately untrue—about being entitled to continue claiming pensions when someone has died. One of the things that could help in future is greater clarity in the information that is given to pensioners about whether they can continue to claim the full pension or not if they claim other benefits or go back to work, which might be an increasingly common thing. Different schemes have different rules.

Were the issues that arose the same as the ones that you mentioned two years ago? You have done pilot research in the same area; did the same issues come up?

Yes.

How should public bodies use the information that we have in front of us to ensure that when the exercise is carried out again in 12 months, the same problems do not arise?

We will find some of the same problems, because it is difficult for some bodies to check such matters for themselves, particularly if people do not declare income when they claim housing benefit.

Why would it be difficult for organisations to check that?

Because the local authorities and the pension schemes do not have the power to match the data. For example, if somebody claims housing benefit and does not declare that they are a pensioner, it is difficult for a local authority to demonstrate that they are a pensioner. They are looking for something that they do not know exists, which is always difficult.

Are you saying that we will only ever be able to find out about such matters in retrospect?

Yes, in some cases. As long as the data protection rules remain on a United Kingdom-wide basis, it will be difficult for public bodies to be sure that they are making the correct payments.

One benefit of the exercise that we are convinced exists but which is impossible to quantify is the deterrent effect. Perhaps Russell Frith can say whether the Audit Commission has found evidence of a deterrent effect.

It is difficult to say, because each time the Audit Commission has carried out the exercise, it has found greater levels of overpayment and fraud, but it has constantly expanded the data sets and the number of bodies that are involved. There is no clear trend.

Mr Frith talked about the UK data protection rules. I presume that local authorities south of the border have the same difficulty with finding out information.

Yes.

Might the way in which registration officers will now work—they will be organised according to the 32 local authority areas—have an impact? When a death is registered, a special form is given to the DWP. That may impact on payments through occupational pension schemes and housing benefit. Have you considered that?

Not specifically. The data set that we use is the DWP's UK register.

So that does not fit with the computerised system that the registration officers in the 32 council areas will operate.

If that is the case, we will need to consider the impact.

I would like to follow up on the £15 million of savings that the initiative has made as a result of identifying fraud and overpayments. The document explains that £5 million of the savings relates to housing and council tax benefit overpayments; £5 million relates to pension-related overpayments by, and forward savings for, councils; and £5 million relates to pension-related overpayments by, and forward savings for, the Scottish Public Pensions Agency. However, the total saving in overpayments is £6 million, which suggests that the majority of the savings—£9 million—are forward savings. Is that correct?

Yes, particularly with pensions.

You mentioned the Student Loans Company, which obviously does not deal with pensions. Therefore, is the degree of fraud minimal in relation to that?

I think that you mean the Student Awards Agency. The category of students who are eligible to claim housing benefit is very small. In 215 cases, it was found that students who should not have claimed housing benefit were doing so.

Thank you for correcting me—of course it is the Student Awards Agency.

The success of the national fraud initiative in bringing errors and overpayments out of the pension system may mean that you will suffer from diminishing returns in future years. Is that a possibility? Are you having to enlarge the data set to ensure that future returns are significant?

There is the possibility of diminishing returns. However, if the exercise was being carried out properly and there were diminishing returns, I would see that as successful. Unfortunately, the evidence so far from the Audit Commission's work in England and Wales is that the diminishing returns have not shown up.

Finally, the report findings show housing benefits claimed fraudulently or in error by 564 public sector employees or pensioners and 53 cases being reported to the procurator fiscal. Is that what one would normally expect as a proportion—I think that it is roughly 10 per cent—or would it be lower or higher in other cases?

I am sorry, I do not know.

Okay.

If I may, I will offer one final thought in response to your question on diminishing returns. I remind the committee that we contributed £71,000 to the Audit Commission for the data matching exercise. We managed the rest of the exercise within our existing resources. The return was made on a very small financial commitment from Audit Scotland; the exercise is more about smarter working on the part of the audited bodies with the good data that they are getting, which allows them to improve their prevention and detection of fraud. For a very small financial outlay, we are getting a better assurance on the way in which public money is managed.

A point well made. I am sure that committee members are pleased to see such a worthwhile exercise, which is making an impact both financially and, hopefully, in the lessons and message that it is sending out.

I have a point of clarification. The convener asked about 53 housing benefit cases being reported to the procurator fiscal, but was the figure of 53 solely for housing benefit cases or was it the global figure?

It was the global figure.

Thank you.

Thank you for the briefing.