---

Danny Alexander (Inverness, Nairn, Badenoch and Strathspey) (LD): In each of the last 15 years the National Audit Office has qualified its assessment of the Department's accounts because of problems relating to benefit fraud, and because of errors in the benefit system. What discussions has the Minister had with the NAO about the level to which fraud and error would need to fall before the NAO could finally give an unqualified assessment of the accounts?

Mr. Plaskitt: It is true that the NAO has not signed off the accounts for 15 years because of the amount of fraud and error that has been reported. It should be noted, however, that the NAO is now complimenting the Department on what it is doing to improve the measurement systems. As I told the hon. Member for Wycombe, as long as we continue to refine those systems and are clear about where fraud occurs and how we should deal with it, we shall be more successful in dealing with it, and can expect fraud levels to fall.

Mr. Paul Goodman: I thank the Minister for being so generous in giving way. He says that the level of fraud has fallen. What has happened to error, which is dealt with on page 26 of his document?

Mr. Plaskitt: The hon. Gentleman is right to point out that we still face a considerable challenge in that regard.
 
13 Oct 2005 : Column 487
 
If he will bear with me, I shall deal with error at length in a moment. If what I say does not satisfy him, he is more than welcome to return to the subject.

The fourth PAC report on fraud and error in benefit expenditure was published on Tuesday. I acknowledge the existence of the issues that it raises. We are working to simplify benefits wherever possible, and the Green Paper on welfare reform, which will be published in the autumn, will outline our vision for a reformed and simplified benefit system.

The report published on Tuesday refers to the 2003–04 accounts. While we welcome the report, it is important to recognise that things have moved on over the last 18 months, and that we have published more up-to-date figures for fraud and error loss than is mentioned in the PAC report. The Department has also made progress in several areas that the report does mention.

We have now improved our estimates of fraud and error, which are accurate to the nearest £100 million rather than the nearest £500 million—which the PAC report mentions and, rightly, criticises. We have begun to benchmark our performance against that of other organisations, as the report suggests. We are, however, proud of our reputation for being at the forefront in the development of estimates of welfare fraud loss, which has been recognised by the NAO. We now use the private sector in debt recovery, as the PAC recommended. Initial performance since April 2004 has been promising, with more than £5 million of benefit debt recovered by the contractors. We accept that there are areas in which we should consider simplifying the benefit system, and we intend to build on that in the forthcoming welfare reform Green Paper.

Miss Anne Begg (Aberdeen, South) (Lab): Will the Minister take this opportunity to re-emphasise that when someone on incapacity benefit, for instance, is invited for an interview to establish what job opportunities may exist, there is no implication that that person is guilty of benefit fraud?

Mr. Plaskitt: I am happy to give my hon. Friend that reassurance. I am sure she will find that it is reiterated in the documents that are due to appear in the not too distant future.

Vera Baird (Redcar) (Lab): I apologise for not being here at the beginning of my hon. Friend's speech.

My hon. Friend mentioned the use of private contractors to recover debt. I am anxious to know whether they are used to recover debt from people who are still on benefit, or only to recover it from people who have been deeply criminal, are not on benefit, and have other resources from which recovery can be made.

Mr. Plaskitt: In most cases, we are using private debt collectors to go after the debt that has proved most difficult for officials in our offices around the country to collect. It is just one part of the armoury against fraud and people acknowledge how important it is to do our best to collect debt.
 
13 Oct 2005 : Column 488
 

Understanding the extent of the problem was one part of our strategy from the outset. Our measurement systems are able to tell us where the main areas of fraud occur. We can also publish estimates for the total fraud loss in the benefits system and I have already said that we currently estimate fraud loss to be less than £1 billion and equivalent to 0.8 per cent. of total benefit expenditure. We used our understanding of the problem to develop our strategic direction.

The strategies that we are pursuing were first articulated in the 1998 Command Paper, "Beating Fraud is Everyone's Business", and further described in the 1999 Command Paper, "A New Contract for Welfare: Safeguarding Social Security". Those documents provided the foundation of the development of our policies and much of what they said still holds true today.

Also in 1998, the Government set themselves targets in the form of public service agreements to reduce the proportion of fraud and error that led to overpayments in income support and jobseeker's allowance. Evidence suggested that those benefits were the most vulnerable to loss, so we have measured the levels of fraud and error in income support and jobseeker's allowance continuously since 1997, which was not done before. Our measurement systems produce results, which are published as national statistics.

In 2002, we also set ourselves a public service agreement to reduce loss in housing benefit—another vulnerable benefit—and a service delivery agreement to reduce loss in pension credit. All our most vulnerable benefits are measured continuously and can provide our best estimate for showing changes in levels of fraud over time.

Mr. Nigel Dodds (Belfast, North) (DUP): To put levels of fraud and error in context, can the Minister provide an estimate of the amount of non-take-up of benefits in the system? Many who are eligible do not actually claim the benefits to which they are entitled and it would be helpful to have the figures on that.

Mr. Plaskitt: It is entirely right and proper to have a debate on fraud in the system, which greatly concerns our constituents, but the hon. Gentleman has provided me with an opportunity to stress that my Department is just as concerned about constituents who are entitled to benefits but do not receive them. I am particularly concerned about pensioner take-up of council tax benefits and we are piloting a number of different schemes around the country in order to improve that take-up. Those pilots appear to be proving quite effective. It is just as important that our constituents who have a benefit entitlement should receive it: that is what the system is there for—to provide the help. It is right to think of both aspects together, as everyone in the benefit system and all recipients also expect us to do the best we can to remove fraud. The hon. Gentleman is quite right to raise that issue. We also publish the results of one-off exercises that report levels of fraud and error in other benefits that we administer. Again, those are published to National Statistics standards.

All those exercises contribute to the overall figure of fraud loss in the benefit system. As I said, the NAO has complimented us on our measurement systems, but we
 
13 Oct 2005 : Column 489
 
are continually looking at ways to improve estimates and provide results more quickly: we work closely with the NAO to achieve that. For example, we are refining the way we measure income support and jobseeker's allowance and will be able to produce a revised baseline, which reflects how we administer our benefits more closely, for the next public service agreement. We have a good record to date on reducing levels of fraud in IS and JSA. Since 1997, we have reduced levels of fraud in those two benefits by two thirds.

Our initial public service agreement target for income support and jobseeker's allowance was to reduce levels of fraud and error by 33 per cent. by March 2004, from the initial baseline measured in 1997. Our achieved reduction of 37 per cent. means that we exceeded that target, and we have set ourselves an even more ambitious one: we now aim to reduce fraud and error by 50 per cent. from the initial baseline by March 2006. Our latest figures, published for the period up to September 2004, show that we have already reduced fraud and error by 41 per cent. Achieving the remaining 9 per cent. reduction remains challenging, but I am optimistic that we will do it.

I shall now set out what we did on the ground to achieve that reduction, how we persuaded our customers and the general public that benefit fraud is a crime, and how we enabled our staff to prevent and detect it.

In summary, we did that by adopting an end-to-end approach—by getting the benefit right at the outset and then keeping it right once it was in payment. We also put benefits right when there were signs of clear abuse or error. We supported local authorities in their fight against fraud and error, and reduced our vulnerability by pursuing a welfare-to-work agenda.

We also used the measurement exercises outlined earlier to identify the areas that we needed to address first, and where we should deploy the majority of our resources. For example, with income support and jobseeker's allowance, the practice of "working and claiming"—that is, concealing paid employment while drawing benefit—accounts for a large proportion of fraud loss. By targeting our resources in this area, our estimate of fraud in working and claiming has been reduced by around 70 per cent. between 2000 and 2004.

Getting the benefit right at the outset is of great importance—both to us, as administrators of large amounts of Government money, and to our customers, who need to know that they are being paid the correct amount. We made more rigorous the checking process in the Department on new claims, especially for income support, and we did the same in local authorities, which administer housing benefit. We introduced additional checks on claims that carry a higher degree of risk—a theme to which I shall return later—including making visits to our customers' homes to check their circumstances at first hand.

Most importantly, we have ensured that the principles of programme protection are embedded in staff roles throughout Jobcentre Plus. For instance, in Jobcentre Plus, we now have over 2,500 financial assessors, who make sure that the customer is paid the correct amount of benefit, and 1,300 national insurance number allocation officers, who help prevent identity fraud.
 
13 Oct 2005 : Column 490
 
Those staff are in addition to our fraud investigation officers, and are also solely responsible for safeguarding benefit expenditure.

In addition, of course, we are maintaining the numbers of front-line fraud investigators. We are building on our successes by creating a new, nationally organised service of 2,000 investigators, who will concentrate on criminal law investigations. Also, 1,000 front-line compliance officers will bear down on lower level fraud and non-compliance, making a strong link between detection and prevention. We are not reducing the levels of front-line staff, but refocusing to maintain the number of sanctions while gaining efficiencies in the process.

We have also put more effort into keeping benefit right once it is in payment. We employ data matching, using our information and that of other Departments, to detect claims that have become incorrect. We have worked closely with Her Majesty's Revenue and Customs in this matter. Since 2000–01, the amounts of overpayments identified through data matching with HMRC have nearly doubled, to just under £40 million per year.

By means of data matching we identify nearly 130,000 incorrect claims within the Department a year, and local authorities identify an additional 65,000. We are also using increasingly sophisticated technology and analytical techniques to risk-profile the types of cases that are more likely to become incorrect, and we correct them when they do. We also want to make use of additional data sources, and I shall outline our plans in that respect in a moment.

We invested in professional training for our investigators, developing an intelligence-led approach to investigations, and introduced new powers to combat fraud in the Social Security Fraud Act 2001. Our investigators can now access suspected fraudsters' bank accounts and other financial institutions, and there were nearly 40,000 requests for such information last year. We updated our policy on sanctions, including the commencement of cautions and administrative penalties. That ensures that we maximise the deterrence effect as there are a variety of penalties that we can impose for even small offences. In 2004–05, we penalised nearly 43,000 people for benefit fraud—a fourfold increase from 1997. We also now use the Proceeds of Crime Act 2002 to confiscate the assets of major benefit fraudsters, and we recovered more than £2 million in 2004–05.

We support local authorities in their fight against fraud and are pleased with their success to date. Reducing fraud and error remains a considerable challenge in housing benefit, and we will continue to work with them to reduce it further. We have also introduced the national anti-benefit fraud advertising campaign to deter people from committing benefit fraud, and increased the capacity of the national benefit fraud hotline to receive information from the public about benefit fraudsters. The hotline has proved to be very cost effective. For every £1 spent on running it, we identify more than £9 of defrauded benefit. We welcome the support from the public—our constituents—and want that to continue. We are working to increase the capacity of the hotline still further as sometimes callers are unable to get through. That is not acceptable, and we are taking steps to address that. Soon the hotline will be
 
13 Oct 2005 : Column 491
 
part of our national contact centre network, which will increase the number of people able to take fraud referrals from the public.

---