Fraud and Error Stocktake Contents

Conclusions and recommendations

1. HMRC’s recent reduction in tax credits fraud and error is encouraging, but it does not know what further reductions are possible. HMRC has managed to bring down fraud and error from 8.1% in 2010–11 to 4.4% in 2013–14, through a combination of focusing on high-risk areas, increasing the number of interventions and introducing new requirements for claimants. HMRC acknowledged that its recent reductions reflected the scope for improvement created by a historically high rate of fraud and error in tax credits. Indeed, its current rate remains higher than many DWP benefits. While HMRC has sought to evaluate the effectiveness of its initiatives, it does not fully understand the reasons for reductions in fraud and error. We were disappointed with HMRC’s failure to estimate what level for fraud and error it might achieve, and its unwillingness to set targets for reducing fraud and error further during this Parliament.

Recommendation: HMRC should set regular targets for reducing fraud and error in tax credits during the transition to Universal Credit, based on an assessment of how recent reductions were achieved for each major risk area and the level of further reductions that are achievable.

2. DWP has not met its overall target for reducing fraud and error, despite being helped by changes in the mix of benefits. Since 2010, changes such as the increased benefit spending on State Pension, which has a low rate of overpayments, have helped bring down DWP’s aggregate rate of overpayments by 0.2%. Despite this, DWP’s preliminary estimate of the overall fraud and error rate for 2014–15 is 1.9% of benefit spending, compared to a target of 1.7%. However, more detailed assessment of performance is difficult because DWP’s aggregate measure masks the different effects on fraud and error of various factors, both positive and negative. For example, DWP has achieved reductions in some benefits, but these improvements have been offset by delays to Universal Credit, increases in Housing Benefit fraud and error, and the cancellation of some planned initiatives to prevent fraud and error. DWP agreed that it had not met its target and that tackling fraud and error needs a more detailed benefit-by-benefit approach, and it has recently started to develop strategies for individual benefits.

Recommendation: DWP should build on its development of individual strategies by publishing targets for reducing fraud and error for each major benefit, having assessed what level of further reductions is achievable, and set out clear operational plans to deliver this.

3. The likely impact of welfare reforms on fraud and error is promising, but the reforms will not solve all the problems of tackling erroneous benefit payments. Both departments stressed that reforms, such as the roll-out of Universal Credit and introduction of the New State Pension, would reduce the complexity of benefits and help tackle some causes of fraud and error. However, we were surprised that DWP still expects fraud and error to increase from £4.6 billion in 2013–14 to £5.8 billion in 2020–21, and that Universal Credit will only save £0.5 billion each year in fraud and error overpayments when it is fully rolled-out. Significant areas of fraud and error will still remain after the introduction of Universal Credit. Real-time information has helped the departments reduce, but not eradicate, fraud and error resulting from a claimant’s income being misstated. HMRC has been more successful than DWP in tackling overpayments caused by uncertainty over claimants’ living arrangements, but both departments continue to struggle with this risk. We were not impressed by the departments’ paucity of ambition to make the real improvements needed in tackling fraud and error. We recognise that Universal Credit will reduce overall government expenditure by removing in-year tax credits overpayments. However, as these overpayments are neither fraud nor error, we were unconvinced by DWP’s rationale for including them alongside fraud and error in its forecast baseline. Furthermore, we note that this forecast was prepared for the 2014 Autumn Statement, and has not been updated to reflect HMRC’s lower-than-expected latest fraud and error statistics.

Recommendation: DWP must set out how it will target the causes of fraud and error that will remain after the introduction of welfare reforms, and update the Committee each year with clear forward projections for fraud and error, based on the latest information available, so that we can assess its performance. The Departments should have a strategy in place to identify and minimise the key risks of fraud and error arising from implementing and operating major reforms, including setting targets for what levels of fraud and error will arise.

4. The departments have made little progress in preventing fraud and error over- and underpayments occurring. In recent years both departments have concentrated on detecting and correcting fraud and error. Although this emphasis has enabled them to achieve savings from correcting erroneous claims, we are concerned that their reliance on such activities puts a burden on claimants which could be avoided if more errors were prevented in the first place. Between 2010 and March 2015, DWP spent just £27 million of its planned £192 million budget for initiatives to prevent fraud and error, and over the same period reduced its forecast savings from prevention by almost £1 billion. A large part of the under-investment in prevention was due to DWP closing its IRIS programme, which should have identified risky claims; but the department was unable to explain clearly why it had stopped a programme which potentially would have saved more than it cost. We recognise the very real challenges in preventing fraud and error, particularly when it is caused by changes in people’s circumstances and living arrangements, but it is important that both departments set out clearer plans and make progress in this area. In doing so, the departments should jointly explore how to exploit third party data to identify high risk claims. Neither department has set targets for preventing underpayments, but they gave no good reasons why they had not.

Recommendation: Both departments should:

5. HMRC has not given sufficient consideration to how its activities to tackle tax credits fraud and error might affect people, including more vulnerable claimants. HMRC has made changes to the way it interacts with claimants, for example introducing new reporting requirements and using digital accounts. We recognise that it must strike a balance between helping claimants maintain up-to-date information, and being firm and forceful when necessary. But we are concerned that HMRC fails to support claimants adequately, is quick to blame claimants when errors occur, and does not accept responsibility for contributing to mistakes itself. HMRC has employed a private sector partner to increase the number of tax credits claims that are checked, but we are concerned that the contractor’s approach has been excessively threatening. Requiring people to provide large amounts of information in less than a month, and cutting off benefits from those who fail to do so, can cause people enormous difficulties.

Recommendation: HMRC should work with the government-wide Fraud, Error and Debt Steering Group to commission an independent review of claimants’ experience of the tax credits process. The review should include the impact of using its private sector contractor and identify ways to reduce unnecessary burdens on people.

6. DWP does not understand the deterrent effect of the penalties it applies. DWP has strengthened its measures against fraudsters and negligent claimants. Since 2012 it has imposed nearly 70,000 penalties for fraud, including prosecutions, cautions and stopping benefits, and more than 150,000 fines for claimant error. It has also run two communication campaigns to raise claimants’ awareness of its activities to detect and punish fraud. DWP’s initiatives to punish wrongdoing have lower financial returns than its other activities for tackling fraud and error. However, DWP has no evidence regarding the impact of penalties on the behaviour of general claimants or fraud offenders.

Recommendation: DWP should assess the impact of its enforcement approach, including modelling and reviewing evidence on the deterrence effects of its penalty regime, to establish how effectiveness could be improved.

© Parliamentary copyright 2015

Prepared 26 October 2015