I have detailed my responses to the action points raised at the Treasury Select Committee hearing on 27 October in the attached annexes.

Transcription

1 @) HM Revenue &Customs Jon Thompson Chief Executive and Permanent Secretary Andrew Tyrie MP Chair of the Treasury Committee November / Parliament Street London SW1A2BQ - Concentrix and Tax Credits Dear Mr Tyrie, I have detailed my responses to the action points raised at the Treasury Select Committee hearing on 27 October in the attached annexes. HMRC's focus has been on resolving the situation for claimants as quickly as possible and we have made good progress on this. As I mentioned at the hearing, HMRC took back 181,000 incomplete cases from Concentrix. I am pleased to inform you that all these cases have now been completed. As of 3 November, HMRC had received approximately 31,700 req uests for reviews of Concentrix decisions and concluded around 25,000 of these (79% of those received to date). At this point in time 90% of these Mandatory Reconsiderations of Concentrix decisions have been overturned in the customer's favour. On the 3rd November, we were reviewing requests for Mandatory Reconsiderations that we received on 24 October. I hope you find this information useful. Should you have any further queries, please do not hesitate to contact me. Yours sincerely, Text Relay service number

2 JON THOMPSON CHIEF EXECUTIVE AND PERMANENT SECRETARY

3 ANNEX A 1. NL to double check whether HMRC s specific guidance on dealing with domestic violence cases was shared with Concentrix. HMRC provided specific guidance to Concentrix on how to deal with claimants who told Concentrix that they were in a domestically violent relationship. This guidance also instructed Concentrix to refer these sensitive cases back to HMRC, as soon as they were aware of the situation. 2. NL to double check whether the data given by HMRC to Concentrix had sufficient detail for them to know that a case involved a child, or an adult, with a disability. The data provided to Concentrix by HMRC contained sufficient detail for them to know that a case involved a child or an adult with a disability, where that information had been previously disclosed by the customer to HMRC. 3. NL to clarify numbers and outcomes of MRs for cases involving queries over hours worked and whether childcare was accurate including the proportion which had been overturned in the customer s favour. The 947,000 cases mentioned by Helen Goodman MP were the figures for the total number of cases opened by Concentrix from November 2014 to August The 324,000 cases discussed at the Work & Pensions Selection Committee hearing refer to the total number of cases opened by Concentrix for the 2016 High Risk Renewals Campaign (from April to September 2016). The cases opened in High Risk Renewals (HRR) 2016 break down as follows (as at 30 October 2016): Category Opened Amended MRs received * MRs closed MRs overturned UP 138,492 38,035 19,826 16,458 15,528 (94%) W&H 47,024 10,069 4,155 2,448 2,202 (90%) Childcare 138,936 10,955 5,480 3,184 2,402 (75%) Total 324,452 59,059 29,461 22,090 20,132 (91%) *These figures do not include HRR 2016 MRs worked by Concentrix prior to 14/09/16. Concentrix actioned 961 HRR MRs prior to HMRC taking control of this work, of which 823 were overturned in the customers favour. The breakdown by risk type is not available. 4. JT to provide approximate figures for the cash flow benefit to HMRC from the practice of making repayments in instalments rather than up front lump sums. Claimants who have been paid less money than they are entitled to, and are therefore owed arrears of tax credits, will receive a lump sum repayment. Those who don t are those who aren t owed any more tax credits. Our repayment processes are not designed to provide a cash flow benefit to HMRC. However, where payments are stopped or reduced for any reason and then reinstated, with or without a lump sum payment, there will clearly have been a period where the Government would otherwise have been making payments. The gain in cash flow terms is very small and the system is

4 not designed for this purpose. On the basis of very broad assumptions, if 30,000 awards are interrupted for 6 weeks, and based on average annual award amounts of 6500, the cash flow gain at an assumed interest rate of 0.25% amounts to around 56,250. This assumes that payments were stopped completely when many were amended to a reduced payment. 5. NL to provide the names of the third party data sources (e.g. credit reference agencies) used by Concentrix. This commercially sensitive information was provided in a separate private conversation. 6. JT to provide an estimate of the difference in costs for HMRC between taking the Concentrix staff onto HMRC s payroll and paying Concentrix to do the work. I am unable to provide a precise figure for the amount HMRC would have paid Concentrix if the contract had continued until May Concentrix were paid on a payments by results basis where they were paid for the savings they brought the taxpayer. Payment would also have been affected by Concentrix s performance against a number of performance indicators, with payment reduced if Concentrix failed to meet performance and customer service standards. The internal budget set aside for the period in question was up to 9.9 million. Taking on around 250 Concentrix staff for the period November 2016 to May 2017 will cost HMRC approximately 2.04 million in paybill. HMRC had in any case been planning to recruit in Belfast and so had funding allocated already. 7. NL to set out the proportion of fraud and error cases (by number) over a run of years, that are fraud on the one hand and error on the other hand. Estimates of the proportion of error and fraud in finalised awards are based on a stratified random sample of cases which are selected to be representative of the Tax Credit population. Tax Credit error and fraud statistics are publicly available at the links below: data/file/431964/child an d Working Tax Credits - Error and Fraud Statistics 2012 to 2013 Revised.pdf data/file/527393/ Child and Working Tax Credit Error and Fraud Statistics.pdf data/file/527392/ Child and Working Tax Credit Error and Fraud Statistics.pdf A summary of these statistics is as follows. These are central estimates which have been estimated using the methodology described above, they are not actual numbers of error and fraud cases. The historical estimates of error and fraud breakdown by number of awards and value are:

5 Number of awards ( 000) Error 1,190 1, Fraud Totals 1,380 1, Value ( m) Error 1,600 1, Fraud Totals 2,270 2,090 1,530 1,360 1,370 Value as a percentage of tax credits entitlement Error 5.7% 4.2% 2.9% 2.5% 3.1% Fraud 2.4% 3.0% 2.4% 2.2% 1.7% Totals 8.1% 7.2% 5.3% 4.7% 4.8% 8. JT to provide further reassurance about how certain we can be about the rightness of the decision in cases where the claimant did not request a Mandatory Reconsideration in what percentage of these cases would the decision have been overturned if there had been a Mandatory Reconsideration? HMRC regularly and frequently checked the quality of Concentrix s decisions. To do this, HMRC randomly selected 500 Concentrix decisions. HMRC then checked the evidence that Concentrix had used to make its decision to see whether the decision to amend the award was correct and whether the amendment was processed correctly, resulting in a financially accurate amended award. The commission amount payable to Concentrix was reduced, if its quality performance level for this check fell below 97%. Concentrix s performance was regularly over 90% and when performance fell below target, HMRC worked with Concentrix to put a correction plan in place. As a normal part of the tax credits process, where a claimant subsequently supplied new evidence as part of a Mandatory Reconsideration and this helped demonstrate that the initial decision should be reconsidered, HMRC and Concentrix made the appropriate amendments to the award.

6 We take steps to ensure that customers are aware they can request a Mandatory Reconsideration if they believe their tax credits have been incorrectly varied. When a tax credits award is varied, customers are sent a letter informing them of the decision and an amended award notice. These both include guidance on the Mandatory Reconsideration process. There is also information on the gov.uk website. We have recently re-emphasised the Mandatory Reconsideration process in the guidance used by our contact centres, so that they can provide advice to customers calling the Helpline, as well as introducing an facility for our customers to make their request. 9. JT to confirm whether there had been breaches of the Data Protection Act and what HMRC were doing about them. We take the security of customer information very seriously and we have very robust processes in place to ensure that personal data is properly protected. Where a data breach was reported, Concentrix were expected to report it to HMRC and to investigate. Around 40 cases of data breach were reported to us in this way. Each one is regrettable and should not happen but in a large scale postal process occasionally mistakes like this can occur. The numbers were relatively small in comparison with the overall volumes of post received and sent out and we saw no evidence of a systemic or widespread issue. A review of the cases recently identified has not led us to change this view and has confirmed that they were all handled in accordance with our procedures. 10. JT to explain how HMRC can make financial sense of a contract where they give as many as a quarter (or even a third) of all tax credit cases to Concentrix. The contract with Concentrix provided additional capacity to check tax credits claims on top of the checks HMRC were undertaking. Over the course of the contract from (November 2014 to September 2016), Concentrix opened 947,000 enquiries. HMRC did not have the capacity to open the additional enquiries and Ministers were clear that a contract with the private sector provided a cost-effective way to reach a large number of tax credits claimants with claims that needed to be checked. The contract with Concentrix provided a good return on investment and produced considerable savings for the taxpayer that would not otherwise have been achieved. Payments to Concentrix up to August 2016 will be in the region of 30 million and, as stated at the select committee hearing, our estimate of savings has been around 270 million (although we expect this to reduce as we work through the Mandatory Reconsiderations of Concentrix decisions). 11. NL to provide detail of the additional analytical work that Concentrix did on top of what HMRC was doing - including the use of credit reference agencies. HMRC sent Concentrix cases to review that we thought were worth checking because we had an indication that the tax credits claim may be incorrect (using predictive modelling to identify cases of risk of error and fraud based on the characteristics of the claim).

7 Concentrix were responsible for using their internal analytics and data matching to identify cases they wished to work from the risk ranked cases HMRC sent them. Concentrix performed an initial risk assessment to screen out low risk cases, including cases that had previously been worked or, for undeclared partner enquires, cases with strong empirical evidence that there was no undeclared partner. Concentrix then applied analytics and data matching to third party data, including credit reference agencies. For Undeclared Partner cases, this involved investigating if there was another adult linked to the claimant s address such as through identifying linked financial arrangements. These processes enabled Concentrix to de-select approximately three quarters of the cases that they were sent for review. Concentrix and HMRC will never be able to screen out all cases without error and fraud through data analytics alone. This is why we and Concentrix write to customers to ask for more evidence to enable an informed decision to be made. 12. JT to confirm if there are any other cases (apart from the Government-wide contract for debt collection) where HMRC use private companies for customerfacing activities and the value of the work that they perform for HMRC. HMRC engages with various companies to provide goods and services to support our work and a small number of these will have some interaction with customers such as interpreters, postal services and dog handlers. The main contracts with the private sector for front-of-house services relate to debt management. HMRC have a call-off agreement from the Debt Market Integrator (DMI), a cross government initiative. The Cabinet Office on behalf of HM Government, appointed a private sector partner, Indesser to establish a new joint venture capable of delivering a suite of debt management services to a range of UK government departments and other UK Contracting Authorities. The new company, is jointly owned by Government and TDX Group (an Equifax company). HMRC represent circa 90% of Debt and spend flowing through the framework call-off contracts. HMRC currently have circa 1.2 billion of debt under management via the call-off contract, which is for a 5-year period. The estimated value of the DMI contract over this period is 35m. At the hearing, you mentioned the VAT Retail Export Scheme. Border Force have an arrangement in place with Travelex at nil cost to provide a checking service for passengers at London Heathrow and Manchester. This is not an HMRC contract. 13. JT to confirm his thoughts on whether to conduct an internal inquiry (in addition to the NAO investigation). As announced on 7 October, the National Audit Office will be conducting an investigation into HMRC s contract with Concentrix. An explicit objective of this review is to highlight lessons that HMRC and the rest of government can learn for the future. We therefore believe that this is the best way to both perform a review and ensure that lessons are learned within HMRC and across government.

8 14. JT to confirm the number of people who have received the payments of up to 100 as compensation for the distress caused (in addition to restitution of the money due) and the total value of such payments. HMRC will pay redress where it is evident that Concentrix have made a mistake or that their actions have caused undue worry or distress. Customers should write to HMRC to complain and HMRC will decide on a case-by-case basis whether redress is appropriate. The amount of compensation will depend on the exact circumstances of each case. From 1 April to 28 October 2016, HMRC has paid redress in 634 cases relating to Concentrix, totalling 34,985.