---

Social Security (Claims and Payments) Amendment (No. 2) Regulations 2010

My Lords, I thank my noble friend for tabling the Motion and for giving me the opportunity to explain how this voluntary debt recovery trial will be evaluated. I take the point about the numbers in the House for this debate—this is an exclusive group—but I can claim to be a member, albeit a new one, of the DWP regret club from earlier this year. I need, however, to apologise, albeit on behalf of the previous Administration, for the fact that the Merits Committee was not provided with sufficient information when the regulations came before it. I should say by way of background that the amendment to the Social Security (Claims and Payments) Regulations 2010 will enable voluntary deductions to be taken from a person’s social security benefit to repay certain HMRC liabilities of under £1,000. The types of debt included are those arising from overpayments of tax credits and those arising from personal tax liability. The transition from work to benefit can be difficult. Before the previous Government introduced tax credits, it was possible for overpayments of in-work benefit, then known as family credit, to be recovered from other social security benefits. Once tax credits were introduced, this option ceased, even though many customers found it an easy and convenient repayment method. The regulations reinstate this option for those who wish to use it. As well as providing a convenient method of repayment for individuals, this will help those unfortunate enough to have a debt with both the Department for Work and Pensions and the HMRC. Under current processes, both departments could seek repayment independently and both expect repayment. Under the trial process, the departments will work together and ensure that customers have only to repay one debt at a time. Participation in the trial is purely voluntary. There is therefore no question of excessive compulsory deductions being taken from a customer’s benefit. The option to use this form of repayment and the rate at which any deduction is made will be agreed with the customer before any repayment commences. This brings me on to the issues raised by the Merits Committee and by my noble friend in this Motion. It is suggested that the trial has been based on inadequate design which will be unable to produce reliable evidence for evaluation. The real issue is perhaps rather that the Explanatory Memorandum failed adequately to explain the design of the trial and specifically the evaluation protocols. This is regrettable and I fully accept that the Merits Committee should always be given sufficient information in order fully to assess the validity of regulations coming before it. In fact, even before this Motion was tabled, the department had written to the Merits Committee, giving a much fuller account of the proposed evaluation methodology than was provided originally. It may be helpful at this point if I give some details of how DWP and HMRC will work together to run and evaluate the trial. We intend to approach tax credit customers and self-assessment customers who are in receipt of a relevant social security benefit and have new debts to HMRC. My noble friend criticised the structure of the trial as being a case of garbage in and garbage out. I think that the issue is that it is rather more narrowly focused than a wider trial might be. The aim is to assess whether voluntary deduction of debts owed to HMRC from DWP benefits is attractive to debtors and a cost-effective means of recovery. Outcomes of the trial will be compared to outcomes achieved through existing HMRC debt recovery methods. The trial will provide evidence to inform any future decisions on use of deductions from benefits. If the trial evaluation shows that this is a cost-effective recovery method for government and that it is attractive to customers, it will be retained post trial. However, if that proves not to be the case or there proves to be no interest from customers, it will not be continued beyond the period of the trial. The objectives of the trial are, first, to test proof of concept. By that, I mean to ensure that the deductions process is operationally effective, that customer information is safeguarded, that customer records are updated and that additional cash recoveries are accounted correctly. The second objective is to assess levels of customer take-up and the factors which motivate customers to participate in the trial or to pay directly to HMRC.  The third is to assess cost-effectiveness; that is, the additional recoveries achieved net of departments’ costs. There will be three groups for evaluation purposes. The first group will consist of those customers who volunteer to take part in the trial. The evaluation will look at how much was recovered, the spread of weekly deduction rates and how long a person is in receipt of a relevant benefit from which deductions can be made. The DWP and HMRC will both analyse their costs in running the trial to permit an overall cost to Government to be calculated. In addition, the performance of new joined-up operational processes between HMRC and the DWP will be assessed. As part of the evaluation, HMRC will seek feedback from its customers on the effectiveness of the trial and their overall customer experience. The second group is those customers who choose not to take part in the trial. The previous material may have suggested that this group formed a control group. Apologies for that confusion, as this is clearly not a control group but a non-random sample of people who have chosen not to take part in the trial. Their data have been contaminated for control purposes. However, there are still some useful data to be obtained from this group. HMRC will therefore note these cases and then deal with them under its current recovery processes, and track whether they subsequently repay their debt or overpayment by other means. For these cases, HMRC will also assess the likely recovery over the period of the trial had the customers taken part in it, and will collect data on why the customer chose not to take part. Finally, there will be a true control group who will not be approached to volunteer for the trial. Instead, they will follow the current HMRC recovery process. The control group will consist of at least 500 tax credit overpayment customers and at least 500 self-assessment debtors. These cases will be selected using a stratified sampling method. This will be performed by arranging the cases into different debt bands from £200 to £1,000 and then randomly selecting 10 per cent of cases in each band. This will ensure that the distribution of the control group represents the overall distribution. Since the overall population of cases in scope in the first tranche is not very large and the take-up rate may be low, it is better to choose a minimum control group and leave the majority of cases to go through the trial. For further tranches where more cases are available, we can increase the control-group size to give a higher confidence level in the results. The trial design has been developed by HMRC statisticians, all of whom are graduate members of the Government Statistical Service. By having three customer groups, HMRC will be able to compare the recovery rates, costs and customer experience across the different groups. It will be able to ascertain how successful the recovery of the debt or overpayment from benefit was compared with current methods. HMRC will be able to compare how much debt is collected via this trial compared with current recovery routes. It will also look at the customers who declined to take part in the trial and whether being contacted regarding this additional payment option prompts any of them to repay their debt or overpayment by an alternative route, or to contact HMRC to discuss other options. If customers taking part in the trial come off benefit, they will be removed from the trial with the necessary explanations being provided. They will not be reapproached if they return to benefit. The business analytics team in HMRC’s Debt Management and Banking Directorate will select the control group for this trial and will analyse the data generated from it. The team is lead by a member of the Government Operational Research Service, and I am assured that the team has a good track record. A separate team in HMRC will provide quality assurance on the evaluation process to the joint DWP and HMRC project board. More generally, I stress my department’s commitment to assessing what is termed as a social return on investment. That means not just looking at costs, but understanding better what is most effective in improving outcomes that matter for people. Wherever possible and relevant, we will develop evaluation methodologies that measure the wider social costs and benefits of public spending. The noble Lord, Lord Lucas, asked whether the provision would cost more than it brings in. In practice, it will cost very little. It is collecting a debt by deduction from benefit, which is effectively the most economical way in which to recover this money. Another question raised by a number of noble Lords was on the pressure being put on the poorest in society to repay their debts. In reality, this trial does not change anything; the debt to HMRC would have had to be repaid anyway—or it would have been chased to be repaid. The regulations apply a maximum recovery rate, and the actual rate is agreed with the customer and can be reduced. On the substantive point, if we have a welfare system, it is important that it encourages responsibility among those people who receive benefits. If we do not we are in practice being unfair to those people who do not overclaim or build up debt and pay up for others’ expense. My noble friend Lord Lucas asked what a tax debt was. The kind of debt that we are talking about here is established debt to the HMRC; it is not just an in-year liability that might fluctuate. He also questioned whether the sample was too small and whether it could be done on a geographic or area basis. In practice, matching the data between HMRC and the DWP cannot be done easily on a geographic basis. My noble friend Lady Thomas of Winchester asked whether customers would understand the voluntary nature of the trial. There was a stakeholders’ event last week where the proposed letters to customers were shared with Citizens Advice, the TaxPayers’ Alliance and others who were apparently pleased and said that they were ““much better”” than other HMRC letters. The trial staff will emphasise the voluntary nature of this. My noble friend also asked whether HMRC would advise taking out commercial loans. The answer is not in these cases. My noble friend Lord Kirkwood asked about the number of people with self-assessment debts. I do not have the exact number with me, but it is around 4,000 people for the last tax year. We are talking about those who were in employment but are now unemployed. My noble friend talked about the problem being generated by the complexity of the benefit system, and I have to agree with that. I hope that we will be in a position to show the way ahead in a matter of weeks in tackling complexity, which has two effects. It disincentivises people from working and freezes people from doing anything, and they become very conservative, which is exactly what we do not want.