Public Accounts - Minutes of EvidenceHC 666

Back to Report

Oral Evidence

Taken before the Committee of Public Accounts

on Wednesday 16 October 2013

Members present:

Margaret Hodge (Chair)

Mr Richard Bacon

Stephen Barclay

Jackie Doyle-Price

Meg Hillier

Mr Stewart Jackson

Fiona Mactaggart

Austin Mitchell

Nick Smith

Ian Swales

Justin Tomlinson

________________

Amyas Morse, Comptroller and Auditor General, Gabrielle Cohen, Assistant Auditor General, Paul Keane, Director, National Audit Office, and Marius Gallaher, Alternate Treasury Officer of Accounts, were in attendance.

REPORT BY THE COMPTROLLER AND AUDITOR GENERAL

HM Revenue & Customs 2012-13 Accounts

Examination of Witnesses

Witnesses: Lin Homer, Permanent Secretary and Chief Executive, HM Revenue & Customs, Ruth Owen, Director General Personal Tax, HM Revenue & Customs, Simon Bowles, Chief Finance Officer, HM Revenue and Customs, and Nick Lodge, Director General Benefits and Credits, HM Revenue & Customs, gave evidence.

Q1 Chair: Welcome. You have had a bit of respite from us for a bit, and we will now see a lot of you.

I will just say how we will handle the two hearings. We will talk today about real time information, tax credits, some general PAYE things, some customer service and perhaps a little bit on child benefit. So we will cover those sorts of issues today, and then we will come back on the 28th and do the whole lot around the tax gap, tax avoidance and the work that you are doing to get the money in. Is that all right?

Lin Homer: That’s fine. Chair, I hope you got the message that for the next meeting I am away.

Q2 Chair: Yes.

Lin Homer: But we are expecting Edward and the team to come down, so I think you will be well served.

Q3 Chair: Okay. Good. Let us start with the real time information, if we can. Having just come out of a session on universal credit, this looks like a much better story. What lessons have you learned from the pilots you have done so far?

Lin Homer: I will hand over to Ruth, because I think she deserves the credit for where we have got to. We would just caution that it is too early to call it "done" yet. We think we are part-way through a significant introduction. We think it has gone well so far and we think we know some of the challenges we still face, but our view is that the whole of this first period needs to be seen as a learning period and a transition. I think it will be some while before we would see this as something that you mark as "done". With that, over to Ruth to talk about-as you say-the lessons learned and where we think we are.

Ruth Owen: We have taken a gradual approach to implementation, learning as we go. RTI was dreamt up in 2009 and we did a pilot, starting last year. Between those two times, we did a full public consultation, at which point we learned from and listened to what businesses said, and changed the design, the solution and the length of the pilot off the back of that immediate feedback, which is what those businesses recommended. Then we gradually built up from a handful of schemes coming in during April 2012 through to 60,000 schemes by the end of the period.

What we learned along the way is exactly how PAYE works out in the real world, so the assumption about how PAYE works changed from an end-of-year reconciliation process to what companies actually do every week, every month. There was the feedback we got from the pilots, about how that was actually operationalised, and the feedback we got from pilot employers about what they thought went well, the trouble that they had and therefore the things that we could put right during the pilot. So we were much more confident by the time we went national in April this year that we knew roughly how businesses were able to respond.

Q4 Chair: You knew how big businesses were going to respond. You have had much less experience of the small and medium-sized enterprise sector.

Ruth Owen: 73% of the pilot schemes were under nine employees, so we had a really good set.

Lin Homer: There were a lot of micros. You are absolutely right. That was the biggest transition. A lot of our pilot were small, but there were many more to be brought in. They are very varied.

Q5 Chair: You have given the SME sector much longer. You have given it until April 2014 to come on board nationally-am I reading that right?-so you have allowed another six months.

Ruth Owen: Yes. We mandated everybody to join this year, and almost everybody has now. The point about real time and what we mean by real-time reporting-the phrase "on or before the time at which you pay your employees"-was one of the pieces of feedback we had from the pilot; some small businesses were finding it tricky. We agreed that for this year they would have a relaxation. As a minimum they would have to report monthly, but not necessarily on the day of payment. As you say, that has been extended until the end of this operational tax year.

Q6 Chair: Do you think you are giving sufficient support to the SME sector to enable it to adjust and do its business with you in a different way?

Ruth Owen: I think generally we have given good support to businesses. The feedback generally has been good. The pilot employers said they felt well supported. Most businesses say, "I was quite worried about going in; I wasn’t really sure how this was going to work in practice." Generally, once they get going, if they have got the right software, it becomes an integral part of how they run their pay-as-you-earn system and payroll system.

Of course, for some people it has been difficult. We are not ignoring the fact that some people have had difficulties. For some people it has meant some significant changes to the way they run their payroll system-small businesses in particular. We have constantly supported businesses through our helplines, websites and things like that. As soon as we pick out where issues arise we update our guidance to help them through that.

Q7 Chair: What about the zero-hour contracts? A growing number of people now have those contracts. They are going to change very often.

Ruth Owen: It is not the contracts themselves; it is when people get paid. One of the pieces of feedback is that many businesses have a high turnover of people who get irregular payment, such as people who come in on a Friday night to help in the pub but don’t have regular employment. We concentrated on those areas to try to understand what the impact of the changes will be for those businesses. Over the summer we put out a survey to ask businesses exactly what burdens and difficulties they face, and we have had 24,000 responses. I have not got the conclusions yet-it closed only at the end of last month-but we will certainly listen to what that survey tells us.

Q8 Chair: Is that a good response rate?

Ruth Owen: Yes, it is.

Q9 Chair: And is that because they are worried?

Ruth Owen: We pushed it out as much as we could because we genuinely want to hear which businesses are finding this difficult, how they are coping and what we can do as a consequence.

Q10 Chair: Let me ask about an issue about which there is a little bit more concern, which is the resilience that you have not built into the system because you felt you could not afford to do it within your financial constraints. If there is a breakdown you do not have the resilience there. Is that a wise decision?

Lin Homer: I think there are two forms of resilience here. I might ask Simon to talk about financial accreditation. We went live with partial accreditation, but that was still better than what it replaced, so it was a move forward. Simon can update you on our plans to bring that up to full. In addition, this is another new system and we have been making sure it has got the normal business continuity resilience built in. If one part of the system-as with all of our technology systems-goes down, how will that impact on employers, employees and the Department for Work and Pensions? We are confident that we are in a good place on that, but clearly as universal credit stands up we will have to keep testing that.

Q11 Chair: What does that mean: "in a good place"? I read somewhere that it broke down. Clearly, while piloting, it will break down. So it has broken down but I do not know how often.

Ruth Owen: Not a lot.

Q12 Chair: How many times since you have been running it as a pilot?

Lin Homer: It is not really one system.

Q13 Chair: What do you mean, it is not one system? My understanding was that people logged their information in and there have been times when that information has not been shoved through into your internal systems so you can then use it and make sure that the person pays the right amount of tax at that time.

Lin Homer: I do not think that we have had a situation where we have had a long-term or ongoing problem, but with all our systems we have moments, including recently with an enterprise release.

Q14 Chair: That is the point at which you need a back-up. You did not invest in one, though.

Lin Homer: We did. We have business continuity arrangements in place.

Q15 Chair: You go back to the old PAYE system, don’t you?

Lin Homer: It depends what the issue is. Sometimes we might just give a signal that says: "Do it again tomorrow". Sometimes we might revert to the old system or to more traditional systems of updating. It depends a bit on what the problem is. We do not have a situation where if something goes wrong the whole system goes down. There are switches and resilience built into what we have done and so far that has proven to be-

Q16 Chair: Give me some examples if I log in and tell you what my employee has earned.

Ruth Owen: If we had a catastrophic failure and one of our data centres went down, for example, we can queue for up to a week employers sending us things. Generally that is plenty of time for our well-trialled business continuity, to bring it back up again and get the information flowing. So it is built to give us at least a week’s recovery.

Q17 Chair: Have you the staff to handle that backlog?

Ruth Owen: It is automated. You do not need the staff. You just need to automatically open the doors again.

Paul Keane: The point that we raised in the report, in 2.26 and 2.27, is the choices that you made at the planning stage-

Lin Homer: About financial accreditation?

Paul Keane: Sorry, around technical resilience and that you did not have full technical resilience. What the report tries to pick out is that it is fine up to a point, but it is whether then you have a dependency with other Government Departments that might need the information urgently or may need it now, and whether you have that facility to service them. The answer that we are trying to pick out is an issue in the report in 2.26 and 2.27.

Ruth Owen: We chose not to have what I would call 24/7, 365-day availability, because pay-as-you-earn does not need that and the discussions we have had with other Departments have not identified that spending tens of millions of pounds on that is a good choice right now.

Lin Homer: That was my point about universal credit. As it develops, we are in constant debate with universal credit. They are going to have a chance to test their system with us incrementally. If we reached a point, then these systems are then capable of being changed and of adding extra resilience. So I do not think that this is a decision for all time.

Q18 Chair: Do you want to say a few words on financial accreditation?

Simon Bowles: May I start by reassuring the Committee that financial accreditation has no impact on what the customer sees or on our ability to collect taxes. Financial accreditation is essentially our internal gold standard, which looks at 15 different characteristics of a financial system. When we launched RTI we focused on what the customer sees and getting a good landing for customers. We are now following on with work to ensure that we move to resolve the issues which NAO identified. These are mainly about tracing through from the tax that is paid, through to this document, the published accounts. You will be aware that our accounts have not been qualified in respect of that, and I can assure the Committee that we have work in hand using the existing data feeds that will ensure that we can sign off the accounts. I believe NAO can approve them at the end of this financial year.

Q19 Chair: Having lived through the PAYE fiasco, and particularly because universal credit has now not become as time-demanding as it might have been originally-i.e. it is going to be jolly late-we think that you should be getting this right, both on the financial underpinning and on the resilience. You should prioritise that, rather than speed of change. That does not mean that speed is not important. If you can get the new system in, I can see that it saves you right the way through the organisation. But given your history with PAYE there is just a question mark here as to whether you are moving a bit too fast and you have not got the resilience in. There are quotes around the place. The Student Loans Company has talked about losing the individual because of a change of circumstances. What is it called? There is a word for it, but I cannot remember. Duplicating.

Lin Homer: Duplicate records, yes.

Q20 Chair: So somebody reports to the Student Loans Company that they have lost their job when they haven’t, it is just that the information has not been translated on the same pay and tax record. So it does impact on individuals. Basically, you are looking at the whole picture.

Lin Homer: Of course. Our whole system impacts on individuals, and I think your point is well made. The fact that we are standing up the system before full pressure is applied actually gives us as long as we need to check and develop. As I said earlier, we have used the whole of this first year to test and develop. We are planning to add financial accreditation, and we will keep the resilience point under review. You may not have picked up on it, but we have recently appointed a new CIO. One of the things we are doing with him is looking across the piece at our technological resilience, if I can put it like that. So I think it is very well made.

All I would say is that we should not think that our old system did not inconvenience individuals. Our experience of RTI so far is that, when people get used to it, it is working well and accurately. It is bringing a bit more tax in sooner than we used to see, and it is making people more aware of what they are doing as they are doing it, and not shunting problems to year end. We will keep considering impact. I think Ruth was being modest. The reason why we have had a good response to the consultation is that people have observed her to be listening to what they say. So we will absolutely keep tuning in, and if that requires us to add some more layers of resilience, of course that would be a high priority.

Q21 Chair: The Sunday Times said that you had issued incorrect tax codes to 40,000 people as a result of duplication. Is that right?

Lin Homer: Yes. I think we have identified a number of duplicate records at the start, but that was a feature of our old system, too.

Q22 Chair: Okay, but 40,000 is a lot.

Lin Homer: No. That is since we started.

Q23 Chair: Since you started nationally?

Ruth Owen: It is 40,000 schemes in which a duplicate record occurred, so something came in that didn’t look quite like your normal PAYE record.

Q24 Chair: So you started a new record?

Ruth Owen: We did in some circumstances. What we have learned from the previous PAYE changes is that things that change within our systems can very quickly generate incorrect output to customers. Our learning from that was that we tried to quarantine new changes coming into the system, so of those 40,000 duplicates that could have been created, only 10,000 incorrect tax codes actually reached customers, and we have now corrected 98%. We have tried to spot where information looks like it is coming in incorrectly and stop it generating a new code for a customer, and where a new code has been generated, because we did not spot it or because the matching was not quite tight enough, we have been able to spot and correct it.

Lin Homer: One of the main changes with this system, which, again, has been quite challenging for employers but has been really important, is that they have had to recognise that they have to have clean and accurate information about their employees. The old system, in a way, could allow them to ignore that for a very long time. We could end a year with an employment record of someone who might have been paid throughout the year on data that was basically just fill-in data-we would sometimes get "A.N. Other, born 1 January 1892." This has required effort from the employer, but that up-front effort makes the whole system much more reliable. That is really good practice for universal credit, because if we get that right, the ability to change payments for the individual as their income fluctuates is based on much better and much more accurate information. These are cleansing but deeply important changes, and again, I think we are giving employers help and assistance to get through that.

Q25 Chair: I accept that this is not as many as 40,000 and that it has already been done nationally, but in the instances where it is wrong for the customer and there will be an underpayment-

Lin Homer: Under or over.

Q26 Chair: They are all underpayments?

Lin Homer: No, some can be over.

Q27 Chair: There is both; okay. What happens when there is an underpayment? Will the customer-like with tax credits, which we will come on to-be liable for the tax that they did not realise they would have to pay?

Ruth Owen: We think we spotted most of them. Where customers have identified an incorrect tax code having been received by them, they have contacted us and we have put them right straight away. So, again, it is not stacking up debt for the future. We believe that we are on top of this and where customers are identifying it, we are correcting it there and then on the call with them, so we can tell them, "Your tax code is now put back to x, and we have made sure that the duplicate is now removed."

Lin Homer: Under the old system, they might not have noticed that until three months after the last tax year. That is the NPS experience, where we then go back quite a long time later and say, "Actually, three years ago you underpaid tax," and they say, "I didn’t know that." In this system, it is a much more rapid response that hopefully narrows the risk time when they are under or over and gives us a chance to keep them much more straight.

Q28 Chair: Okay. So what you are really telling us is that you can identify overpayments and underpayments earlier-

Ruth Owen: Yes, more quickly.

Q29 Chair: But that the individual, if it is an underpayment-even though it is not their fault-will still be liable for that.

Lin Homer: The same rules apply as have always applied. There is a responsibility on any individual taxpayer to pay the right tax, but if there is clearly a mistake on the part of their employer or us, we have rules to deal with that.

Q30 Chair: What does that mean?

Lin Homer: You probably know this from your constituency postbags-

Q31 Chair: You sometimes write it off.

Lin Homer: Exactly. If we have clearly made a mistake, we have rules-we are quite rule-bound-that allow us to accept that. We also have some rules that allow us to pursue the employer. But if an individual knows that they have two employments and, for instance, they are getting their basic rate of tax relief on both, we will not necessarily accept that, because we will be saying, "You should have known that you don’t get your tax relief twice over." But in this system, at least, we might be having that conversation after a few months, not a few years. That would be the difference.

Q32 Ian Swales: Just testing this a bit further, the essence of universal credit is very close connection between what you are doing and what DWP are doing. Can you say a bit more about the progress of that and whether the required system integration is going according to plan?

Ruth Owen: Yes, we completed the link over 12 months ago. It is working and we have got 2,197 universal credit claimants logged on our system and we have been feeding back information on them, with just over 5,000 payments so far. In the scale of the pathfinder so far, it is being tested. It clearly has not been scaled up yet, until we are ready to extend it.

Q33 Ian Swales: What is happening to people whose circumstances change frequently? One of the problems with the old systems was that people would not take a job that lasted for a few weeks because of the battle they would have in informing all the authorities that they had got work, followed by the huge battle in informing them that they no longer worked. Have you any experience of cases where people have gone in and out of various parts of the system through their circumstances? How effectively has it been working with that?

Ruth Owen: It is probably too early to say whether we have had lots of people going in and out, but we have certainly had people going into work and having-

Q34 Chair: Will you repeat those figures on universal credit claimants, Ruth?

Ruth Owen: The number of claimants was 2,197.

Q35 Chair: Out of those, you had some other figures underneath that.

Ruth Owen: Just over 5,000 payments have been made as a result of the information that we provided to DWP.

Q36 Chair: 5,000 payments. Are those the old tax credit payments? What kind of payments are you talking about?

Lin Homer: No, those are benefit payments.

Ruth Owen: These are people moving into work.

Q37 Chair: 5,000 payments of tax?

Ruth Owen: No, payments of universal credit.1

Lin Homer: At the DWP end-

Q38 Ian Swales: At what speed are those 5,000 payments to the 2,000 people in the pilots taking place? Is it in real time?

Ruth Owen: Yes, it does live up to its name. It goes over daily to DWP. I do not have a lot of data on lots of people moving in and out of work, but I have got anecdotal evidence from people who work in the jobcentres that they are able to use it to say to people, "This will support your move into work. You will have confidence that universal credit will match up to your real-time earnings."

Q39 Ian Swales: What about the management and the staff interface in this now? Universal credit, by definition, is one system but you have two Departments involved.

Lin Homer indicated assent.

Q40 Ian Swales: Are you satisfied that you have put the right procedures in place outside the system to make this work effectively? I ask that because one of the things anecdotally we hear about universal credit is that there is something of a turf war going on between the two Departments about who does what and where the responsibility lies. By definition universal credit is meant to be one system. How are you making it look and feel like one system as far as customers are concerned?

Lin Homer: I don’t think from the customer perspective that they should see any evidence of a turf war. So if you have picked that up, Robert and I would want to hear about it. In terms of the day-to-day working, both Ruth’s and Nick’s people are heavily engaged in the programme that is universal credit. The decisions we take at our end, both in RTI and tax credits that either have an impact or will in the future on universal credit, we sight DWP on. Robert and I have regular conversations. We have plans for the transfer of staff so it does not feel like a turf war. Indeed, when the Secretary of State went before the Employment Committee, Suzanne Newton, who is one of Ruth’s key people on RTI, went with him. So it feels joined up to me but there is a lot of work still to do. I am sure we will have our moments.

Q41 Ian Swales: To clarify, it is more to do with the systems work behind the scenes rather than the point you were just making.

Lin Homer: That comes into it.

Q42 Stephen Barclay: On that point, the reason for the Committee having that sense was that in response to Q197 at our previous hearing, Sharon White of the Treasury said: "The relationships-certainly the ones across Whitehall-are not perfect but they are much more solid than they were previously". I think she was alluding to the fact that there have been difficulties in the relationships.

Lin Homer: She may have been speaking for herself.

Q43 Ian Swales: Is it true to say that the expression "tax credit" should be disappearing from your lexicon and that you should no longer be talking about tax credits in future in the world of universal credit?

Lin Homer: Yes. In due course, tax credit ends and migrates. We have come up with an agreement. Nick can talk about this if you like. We have come up with an agreement which is for a transition over of tax credits in a way that is agreed with DWP. We are not doing this as a traditional machinery of government: one day they are mine, the next day they are all Robert’s. We are moving at a rate and in a way that suits universal credit as the dominant project but at some point tax credits will not exist any more.

Q44 Ian Swales: Are you talking about the people who are already in the system and you are still dividing the cake in two ways? Does the expression "tax credit" still apply to the 2,197 people in the pilot?

Lin Homer: Tax credits-

Ruth Owen: They are on universal credit.

Lin Homer: Tax credits are still being paid to people who are not on universal credit.

Q45 Ian Swales: Okay, I understand that. So it is joined up now for the people who are on universal credit? They only see one system facing them. Is that true?

Lin Homer: Yes

Q46 Ian Swales: And it is joined up behind the scenes?

Lin Homer: Yes.

Q47 Chair: May I just delve a little into that number: 2,197 is tiny. That 2,197 is from when?

Ruth Owen: The start of the pathfinder.

Q48 Chair: Which was when?

Ruth Owen: Fairly recently. April.

Q49 Chair: April? So these are people who are in work but they are applying for universal credit because they are low-paid and to make work pay? Is that right?

Lin Homer: No. Universal credit covers all benefits.

Q50 Chair: So it might be child benefit or it might be-

Lin Homer: No. Child benefit is with us. But housing benefit, jobseeker’s allowance-

Q51 Ian Swales: Is it true to say that the 2,197 are the proverbial single man, no children or whatever? What is the range of people that you are now covering?

Chair: On JSA.

Nick Lodge: The pathfinder is currently testing the roll-out of universal credit in a quite a narrow range of potential claimants.

Chair: We know.

Q52 Mr Bacon: Define "quite narrow", please, just for the record.

Nick Lodge: It tends to be the customer base that Mr Swales just mentioned, so it is the people who are perhaps coming into work. There is no connection with tax credits. They were not on tax credits.

Q53 Chair: Single and with no children?

Nick Lodge: Exactly. That is my understanding.

Mr Bacon: Living in Ashton-under-Lyne.

Q54 Ian Swales: Let us clarify this: the people going into the pilot were not the people in receipt of tax credits.

Nick Lodge: By and large, so far, that is absolutely the case.

Lin Homer: Not so far.

Q55 Ian Swales: So you have given us confidence that very simple cases are now operating smoothly.

Chair: Not really. It is only 2,000, Ian. It is hardly testing anything.

Q56 Ian Swales: Well, as I often say in this Committee, if you can make a system work for one person, you can make it work for a million. The problem is systems that do not work for even one person. In this case, what comfort can you give the Committee that you have processes in place to deal with people who are currently in receipt of tax credit and people with more complicated family situations? In other words, what is the pace at which you will be able to cover the population that you have to deal with?

Lin Homer: Those are decisions for the Secretary of State and the DWP. We plan to work with them as they come into the area that has traditionally been ours. We will be involved in the design and test of the system; we will be transferring staff who have experience in our system, and we have agreed to transfer the staff that DWP need. So this is my point about our very close partnership. I am not just saying, "Have ’em all." We are saying, "These are the folk you need." We will agree the pace at which they go, and we are determined to try to share our experience with them and to have our business in as good a shape as possible so that, as they are transferred over, the customer impact is as good as it can be.

Q57 Ian Swales: Can you think of any example where it would actually be HMRC that would be on the critical path of that implementation, or are you saying that DWP will always be on the critical path of rolling this out? In other words, the moment they say, "OK, now it’s man, woman and two children," are you ready to respond as fast as they can move? Or are you on the critical path on any systems or organisational issues?

Lin Homer: RTI is a critical part of the system, as we discussed. Timewise, it might not be so critical now. We are key to their project at a number of points. With the planning that we have in place, we will always know the lead-in time. We are being very clear about that to DWP. They are understanding. We are docking all the way up, so we are in all their key programmes. We have got people working closely. Indeed, Robert and I sit down and discuss pinch points regularly, but your point is well made. From the customer perspective, they should not expect Departments to make life difficult for them because we cannot get on, so we are doing our best to culturally and organisationally make that a low risk.

Q58 Chair: May I ask one more question on RTI? You are introducing automatic filing penalties in April 2014.

Lin Homer: Yes.

Q59 Chair: So if an employer puts the data in and for some reason it does not go through, because your systems do not work properly-hence the question about resilience-how do you ensure that they do not face a penalty because of a fault at your end?

Ruth Owen: It is what we call reasonable excuse. If a penalty is applied to a business, for which they have a reasonable excuse and they could not use our system or the system failed-

Q60 Chair: They may not know. As I understand it, with some of these duplicates, you do not know that it has not gone through

Ruth Owen: No, we should, and they always get a confirmation. If you did not get a confirmation, you should check to make sure that we received it.

Q61 Chair: And they know that, do they?

Ruth Owen: That has been happening, so everyone should be getting used to it now. Every time you send us something, something pings back to say, "Got it". That is standard electronic data exchange. If, in a major disaster, people were sending us things and nothing was coming back, they would have a record to say that they had been trying to send us something and nothing came back. If by chance we had not spotted that, which would be very unusual, they would be able to apply for any penalty to be discharged, but generally, we would have been able to see if we had had a major outage, and that would be the reason why the submissions had not come through.

Lin Homer: In addition, we are contacting people this year when we are not applying penalties and saying, "If you had done that when we were applying penalties, you would have left yourself exposed to a penalty." So we are not penalising people, but we are using it to illustrate to them what they did not do or where they went wrong. Again, we are trying to use this first year as a chance to familiarise people with the system before it is for real. I think that that is proving helpful and should also help us to learn if there are circumstances we have not spotted where people can do something wrong without knowing.

Q62 Stephen Barclay: You said a little earlier that most employers do not have a problem if they have the right software. What proportion of employers do not have the right software?

Ruth Owen: I didn’t quite say that. I said that most employers are in-we have 90% of employers in. The feedback that we have had from businesses so far is that they have found it quite straightforward and that it has been less of a difficulty than expected for most employers. The point about the software is that if you have it-including the software that we give away free-and run your payroll on it, generally that means that when you run your payroll the final task says, "Press the button to make sure that this gets reported to HMRC." People who have software like that have said, "That has been quite straightforward. I know what I have to do to run my payroll and that is one additional step."

Q63 Stephen Barclay: So you don’t foresee any business cost to firms for software?

Ruth Owen: Well, there could be if people choose either to upgrade their software or change supplier. All the major payroll software suppliers and payroll bureaux have been working with us on redesigning for at least two years. Of course, they refresh every year according to tax rates and things like that. They have all been integrating it into their systems.

Q64 Stephen Barclay: So it is not that the free software that you are giving away will be fully fit for purpose? They might need something bespoke.

Ruth Owen: Ours is very basic. It is called the basic payroll tool for a reason, so if you need to do additional things such as pay pensions or make deductions for loans-things like that-you need something greater than that to run your payroll. Our software is specifically to help you to run PAYE.

Q65 Stephen Barclay: Are the businesses that are most likely to need a more bespoke system the pubs and some of those firms you have identified that had particular problems in the pilot and for whom you have de-scoped the requirements on an interim basis?

Ruth Owen: I don’t think that that is a software issue but, again, I will look at the research and see whether that tells me anything. The people who I understand are having most problems are those who run their payroll monthly-using software or whatever records they keep-but pay their employees weekly and have always caught up at the end of the month. Our requirement is for them to tell us every time they make a payment. If they did that and were paying people weekly, that would multiply the number of times that they were running a payroll. I assume that in the pub at the time they were probably just keeping a record.

Q66 Stephen Barclay: Sure, but you are allowing it monthly at the moment. If you attach fines in April at the point where you remove the exemption, you are raising the bar on something that has already proven difficult, at the same point that you apply a fine.

Ruth Owen: indicated assent.

Lin Homer: Well, the requirement was always there. I think that RTI has shown up organisations and small businesses that may not have been doing it. I think we had this debate on one of the previous occasions when I was here. We would encourage owners of small businesses to think about how they are keeping good records if they are paying out weekly but-truthfully-reconciling monthly, because there is a risk that they are inadvertently making errors in that system.

It has always been a requirement of our system that you deduct tax and give it to us when you pay. We will look at the impact of the changes. I know that one of the consultation responses we have had is about whether we will run a penalty-free period for people who are coming in later. That is going to be one thing on Ruth’s list, but I do not think that we have reached a conclusion about it.

Q67 Stephen Barclay: Thank you; that was very helpful. On the 2,190 so far, could you tell us what the next two interim milestones are?

Lin Homer: Sorry-are you going back to universal credit?

Q68 Stephen Barclay: Yes, I am asking about the join-up with universal credit. We know that universal credit is behind the pallet-from memory I think that it was 140,000 by April 2014, and I think they are on around 8,000, so I assume they are not going to deliver that. In terms of your feeding into them, what are your next two milestones?

Lin Homer: In terms of our positioning, we stand ready to move those numbers up as they come forward.

Q69 Stephen Barclay: So when will it be 100,000 then?

Lin Homer: I haven’t come with those details. Those are for the Secretary of State and DWP to decide. Obviously, we think we can cope with transition either that is slow and steady or a bit steeper. We would encourage them to give us time to learn and adapt, but those are not figures that I have to hand.

Q70 Stephen Barclay: May I just ask about the increase in cost? I think the original cost was around £240 million, and it has gone up £115 million since then, which seems quite a significant increase.

Lin Homer: It is.

Q71 Stephen Barclay: Why were there so many unforeseen costs?

Lin Homer: I will get Ruth to give you the detail. I think we did not put all the elements of cost in to start with. So it is not a situation in which individual things we estimated turned out to be significantly wrong. In some areas we didn’t recognise the additional costs in other parts of our system. For instance, this allows us to keep an earlier and closer eye on debt. That will probably be good for us in the long term, but in the short term it can make collecting debt more expensive. I don’t think we had anything in the original business case for that. Perhaps Ruth could add a bit more detail.

Ruth Owen: There were two main reasons. The first one was that having a year-long pilot gave us lots more things to identify that we wanted to build in as we learned through the pilot. The IT costs increased as we saw things that we could improve based on the feedback. The costs that increased more significantly were the business change costs, where quite frankly I think it was underestimation of the changes that we need to make within our own HMRC business at the beginning of the outline business case.

Q72 Stephen Barclay: So, why was that not covered in contingency costs allocated at the start of the programme?

Lin Homer: Because we are a big Department and we run a significant number of projects, we don’t tend to drop contingency into every project. We deal with that in our business planning. I think we have made the point to you before that the increasing costs for RTI HMRC has absorbed, so we run our programmes as a portfolio.

The reason Simon’s beard is grey is that he holds that space in the middle. Within year and year to year we adjust our programmes effectively to provide that contingency. That is a portfolio management approach as opposed to a project approach. That is an approach that has been thoroughly tested with MPA and is something that NAO has observed. We think that is a better way to get our capital spend quite close to the amount. You would rightly criticise us if we significantly underspent, but that is a bigger risk if you attach contingency to every project.

Q73 Stephen Barclay: You have given us very reassuring answers. Is the project still rated amber by the Major Projects Authority?

Ruth Owen: Yes.

Q74 Stephen Barclay: In light of the reassurance today, why is it still rated amber?

Simon Bowles: I think any large programme of this scale is going to be amber until we are pretty close to delivery. We started by saying that, although progress is good, we have still got a way to go.

Q75 Chair: But you’ve got no contingency left.

Stephen Barclay: There wasn’t any to start with.

Simon Bowles: Chair, I think we have shown that we have managed to absorb those pressures successfully.

Q76 Chair: Are you expecting more pressures?

Simon Bowles: I think we have reached a high-water mark, and indeed recent iterations of the business case have shown costs coming down.

Q77 Stephen Barclay: As part of getting to where we are today, around £1 billion of tax was written off.

Lin Homer: Yes. Again, I think we have talked with you about that.

Q78 Chair: That was PAYE. We’ll come to that at the end. We will come back to that as it is a very important point.

May I raise a final thing, and then I think we can move on out of RTI? I got hold of a survey done by HW Fisher and Co, an accountancy firm, of 2,000 SMEs. Have you seen that survey?

Ruth Owen: No.

Q79 Chair: It is less glowing than the report you have given us today, which is on the SME sector. Almost a quarter described RTI as frustrating. Less than a third said they had a positive view about it. Nearly half said they had encountered hitches. A third described the transition as difficult, and 39% felt it was a cost burden to their business.

Lin Homer: I hope we have not painted too rosy a glow. You started by saying that, at least relative to others, it had gone quite well. What both Ruth and I have tried to say is that we think this is work in progress. We would wholly accept that the transition has been challenging for some people, and I am not surprised at the use of the word "frustration". We would hope that once people have settled into this way of working, they will not continue to find it frustrating, and indeed that it will really help them run their business well. Part of the consultation is to pick up any messages we are missing, and we are trying to do a retrospective evaluation of business administrative burden to challenge our original assumptions.

Q80 Chair: Are you going to do your own customer service testing? That would be helpful.

Lin Homer: That is what our survey is.

Q81 Chair: So it will be a sort of customer service survey, will it? Your current thing is that you put out a consultation and 26,000 responded, which is good. If you got somebody to do a survey for you, you might get a slightly different picture-a truer picture-if it is done properly.

Ruth Owen: I will see what our original survey shows us. We continue to want to understand the voice of the customer, and to know which segments of the market are struggling and what we can do to support them.

Q82 Stephen Barclay: Your public statement on that original one seems at odds with that survey. The Department’s position was that RTI was "easy" for most employers. The majority reported no problems with RTI. There was "no change in payroll running costs" for most. That is the experience of the majority. The longer-term expectation is that the majority anticipate that the burden will decrease.

Ruth Owen: Yes. That is from the pilot.

Lin Homer: Those words are from the pilot, and that was true.

Q83 Stephen Barclay: Do you think that that is consistent with that survey, though?

Lin Homer: I think we still believe that overall this will be a significant admin reduction for business. What we want to test is whether that is evenly distributed, or whether it is better for some-medium to big-and tougher for small. We are open-minded about how it falls, and that is part of the reason why we want to do a retrospective on the admin burdens. I think I have said before that businesses that were not running their administration particularly well are going to find this something of a challenge.

Q84 Chair: But that is a lot.

Lin Homer: We would be very interested in seeing that survey.

Chair: H.W. Fisher and Company, in September-2,000 SMEs.

Lin Homer: Yes. There are 4.6 million2, so what we need to do is check that we can hear their voices. We have got to keep testing.

Q85 Fiona Mactaggart: You have said that you think that companies that are not well run are struggling with this-fair enough. But I got the sense from what Ruth said that you think that some sectors of business are finding this harder than others. Can you tell us about that, please?

Lin Homer: That is why we put some easements in.

Ruth Owen: So it is sectors that have particular labour forces or payroll practices. I think pubs were the example, where you might have people who you have just hauled in on Friday night and you have not kept records, necessarily. In the farming industry, we have worked very closely with the National Farmers Union, because they have harvest casual workers and things like that where it is very hard to keep records and keep track of people who are-if you like, itinerant workers. You do not generally keep in-depth records of those people because they turn up for one shoot for a weekend and you never see them again. We have tried to work with the NFU to come up with compromises on how we can address some of those issues without being burdensome. We continue to work with them, because that is the key question of our surveys: which sectors are struggling and for what purpose, and what can we do about that?

Q86 Fiona Mactaggart: I am just struck by the fact that on the screen behind you is the title of the debate that is happening now in Parliament, which is about zero-hours contracts. In effect, what you are saying is that this modern contracting system could cause real problems for some businesses in doing this.

Lin Homer: To be clear, zero-hour contracts are often used by big employers, such as big retail. We expect them to be able to run systems as complex as their operating model. We would not necessarily expect to make as many easements for that sector as we would for some of the people that Ruth has described. In fact, RTI ought to help with that, because you can make different payments each week and our system should bring them all together. Zero-hours contracts are probably less of a challenge for us in an RTI world than they were before. I am not saying that they are only used by big employers, but they are used significantly more by some of the very big employers with variable work loads than they are by others.

Q87 Fiona Mactaggart: I accept that the big employers are capable of doing this, but small employers are beginning to follow their big leaders.

Lin Homer: That is more like the agricultural worker and the pub worker. We were already looking at them.

Q88 Chair: I am going to move on to tax credits, because that is the other biggie. It is a less happy story. Paragraph 4.26 on page 47 of the Report tells us that you are going to write off nearly 70% of the tax credits debt. Isn’t that really letting the taxpayer down?

Lin Homer: Do you mean Amyas’s Report or the annual accounts?

Chair: The NAO Report. Paragraph 4.26 on page 47 states that, basically, 69% of tax credits debt is going to be written off.

Lin Homer: I am going to invite Nick to explain that to you.

Chair: It is not a very happy story and it is not very good for the taxpayer.

Nick Lodge: This is the impairment we have made to the tax credits debt balance, which was £4.8 billion as at the end of 2012-13. We ran an exercise to look at the collection rates. That came up with the figures that you mentioned. According to that exercise, the recoverability of tax credits debt was, I think, a shade over 30%-on average, 31%-implying that 70% would not be collected.

Q89 Chair: So 69% won’t be collected. That is pretty shocking.

Nick Lodge: It is a large figure. As we have discussed in this Committee several times before, tax credits debt is very difficult to collect. If we look at the historical figures, going back to 2003, and look at the total amounts that have been created in terms of overpayments, which is about £16 billion, the total amounts collected and remitted, and the total amounts on the debt balance, actually, we have collected about £7 billion overall and remitted about £4 billion. So generally we do a bit better with regard to that.

Q90 Chair: I have to stop you there. I asked for a little bit of work on this. If you look at the value of the debt arising out of 2003-04, done on lower bound, central estimate and high bound, it came to £3.5 billion. That was the first year, when you were bedding down and expecting things to be going wrong-I accept it is a complex system, and all that. But then look at 2011-12: it has doubled, in the paper I have, to £6.25 billion. So, despite the fact that one would have thought that you are getting smarter at detecting and stopping problems over the years, the amount you have overpaid or people have overclaimed-whatever it is due to-is £6.25 billion. It has gone up. That is a really sorry story.

Mr Bacon: Can I just add something? You are adding a further £1.8 billion-this is in paragraph 4.27 on page 47 of the Report-of new personal tax credits debt in 2012-13. The Report says: "HMRC estimates that it is likely to produce £1.9 billion of new personal tax credits debt in 2013-14 and, based on current business processes, total debt could increase to £5.5 billion by 2014-15." That is on top of what we have now, isn’t it? It is new extra debt.

Lin Homer: It is important to explain the general context of tax credits and our approach to them. After that I will get Nick to explain some of what he is doing. The nature of the system is that we make a payment based on what we are told at the time. At the end of a year we reassess that and we adjust that payment retrospectively: so if someone has ended up earning more than they expected to or they have told us, by the end of the year, that they owe us money, there are then very strict rules about-

Q91 Chair: Lin, I am going to stop you. I want you to withdraw a little bit there. When we had Ruth in looking at 0845 numbers, she told us that when people were ringing up on that last day of the year to discuss their change in circumstances on the tax credit hotline-

Lin Homer: They couldn’t get through.

Q92 Chair: If I remember correctly, only 16% got through. It is not all about people not telling you. If you have a system where people cannot even tell you or ask you about their change of circumstances, it is bound to be wrong and these are all poor people. I accept that it is a hideously complex system. We all understand that.

Lin Homer: That was my point. It was not the individuals-

Chair: We understand that, but no progress has been made. When there are simple things such as people not being able to get through on the phone-

Lin Homer: I am sure we will come to that. I know you have already had Ruth here earlier in the term. The point that I was trying to make was the system point. We then seek to collect any overpayments where we can from ongoing payments that we are making. That means we will hold a debt arrangement with an individual in tax credit much longer than you would hold a debt arrangement with a business or in a normal business environment. We might take 10 years to take the sum back. We will let the debt sit on our books for a long time while we try to get it back.

Q93 Chair: We are not talking about that; we are talking about writing off 69%.

Lin Homer: No. The figures you-

Chair: It’s writing off.

Lin Homer: The figures you were talking about when mentioning our debt growing are about us being prepared to wait longer to get-

Chair: I started on the write-off.

Lin Homer: You did, but you moved on to debt. I think our debt will go up because we wish to take as much opportunity as we can to allow the families in this very complex system-I will be completely frank with you. Were it not transiting to universal credit, it would be at the top of my priorities to change, because it is complex for us and for the country and difficult for the people who are in it to run, but we are in transition, so we are making the best of it. Debt is likely to go up a little bit as we actually pursue more of the debt, rather than write it off, but pursue it over quite a slow payment process. During this year, we think we will have touched all the debt that we are owed, but in some of those circumstances that will be to enter into arrangements to take very small sums back over a very long period. That leaves the debt on your books and I think that is the right thing to do.

Q94 Chair: 69% is being written off. Perhaps Nick can answer that question.

Lin Homer: Only in a given year.

Nick Lodge: Well, 69% is the book figure. That is not necessarily the amount that will end up being written off, because, as Lin said, we will attempt to collect all of that debt. We will put it through a full collection cycle this year and will continue to recover it-

Q95 Chair: Can we read this paragraph again? "This led it to reduce estimated recovery rates by 12 per cent". You have reduced your estimated recovery rates-I accept that you are going to do a little bit every week over a longer period of time-by 12% "from 43.3 per cent to 31.4 per cent". You will collect 31.4%, which means that 68.6%, you don’t.

Lin Homer: That is of the historic debt. That does not count the amounts which, as overpayments, are turned into ongoing collection-

Q96 Chair: What is "historic"?

Lin Homer: Where there isn’t an ongoing payment-

Q97 Chair: What years?

Lin Homer: If there is not a continuing tax credit payment or an agreement to code it out, we will turn that into traditional debt and we will seek other ways to receive it. We are not suggesting that we only get back 30% of all overpayments in the tax credit system.

Q98 Chair: Define "historic debt" for me. What is it?

Lin Homer: That is debt that is accumulated over many years that might still be owing at a point, for instance, when a tax credit payment finishes.

Chair: It has only been in since 2002-03.

Lin Homer: Yes.

Chair: So it’s not that many years.

Lin Homer: Quite a lot-

Chair: Where is your cut-off point? What is your definition of historic?

Lin Homer: In terms of when we move debt over, it will be if we are in a position where we no longer have an arrangement that we can attach the debt to, an example of which is if someone has fallen out of tax credits. At the moment, we-

Chair: They will still be earning.

Lin Homer: Yes, but then we have to pursue it as a normal debt. We cannot offset, so then you-

Q99 Chair: This is writing off. I will read it again. This is a reduced estimated recovery rate to 31.4%.

Lin Homer: On debt.

Nick Lodge: That is applied to the tax credits debt balance for the accounts. Those figures are absolutely right and they were calculated with the NAO. It does not mean in practice that we necessarily will write off that amount. Historically, we have done better than that. We are making improvements to our debt collection to try to do better than that for the future.

Q100 Chair: Well, you haven’t done better-sorry about that-because you have reduced it. Your performance was such that you reduced the estimated recovery rate by 12% from 43.3% to 31.4%. You can’t say you were doing better so you assumed you would get less in. That doesn’t make sense.

Lin Homer: That paragraph is referring to our judgment about how much we can recover.

Q101 Chair: I understand that. I understand that it’s an estimate of recovery, and I understand that you might do a little better. I also understand that the amount you estimate you are going to get in has gone down, so you can’t say your performance has got better. I also understand that there is a heck of a lot you estimate you won’t get in. You might do a little better than that, but it is still a heck of a lot.

Simon Bowles: I think it is worth saying, Chair, that this is a provision in the accounting records. In fact, there is a lot of work under way that Nick probably wants to talk about to ultimately prove that that is a conservative view.

Q102 Mr Bacon: Can I just be clear about this? Are you saying we should be clear that this is a provision in the accounting records? You make it sound technical, and not merely boring but unimportant.

Simon Bowles: That is not my intention, Mr Bacon.

Q103 Mr Bacon: Good. I read again from the paragraph the Chair was reading from: "and increase the provision for irrecoverable debts by £985 million"-in other words, nearly £1 billion. Can I just be clear about one thing? This £985 million we are talking about is money that was paid out that should not have been paid out because it was an overpayment. That is correct, isn’t it?3

Simon Bowles: That is correct.

Q104 Mr Bacon: You are now saying it is irrecoverable.

Simon Bowles: That is correct.

Q105 Mr Bacon: What is the purpose of making it clear that it is an accounting adjustment? We can see it is a change in the records. I am not with you.

Lin Homer: If you think about our approach to recovering overpayments, we don’t turn all overpayments instantly into debt. Some of them never go beyond being an overpayment. So we say to you, "Last year, because your income changed in year"-there is no fault; as we discussed before, we think much of this is error, not badness-"next year we will take 10% or 25% of that debt off your payment." Many people, although it takes them a long time, repay that money and their overpayment never turns into debt.

Q106 Chair: But they shouldn’t appear in this number.

Lin Homer: They don’t. This is a provision for the debts we expect to become irrecoverable. It is a proportion. Nick was talking earlier about overpayments in the system since it started. We are arguing about smallish amounts, so perhaps it is important to talk to you about what we are going to do about it, rather than the differences. We think it is £7 billion collected to £4 billion written off, plus another £4 billion or £5 billion still being pursued.

Mr Bacon: The C&AG is dying to come in here.

Amyas Morse: Only if it would be helpful, just to bring it together.

Lin Homer: He always says that.

Paul Keane: I think it does cover it. This amount covers the amounts that you are going to receive from ongoing awards as well. The £4.8 billion includes the sums you are recovering directly through your own efforts or through debt collection agencies, plus the amounts you are recovering through ongoing awards, which may be, in some cases, paying back small amounts over a long period. The provision applies to the entire amount of the sums that are recoverable.

Amyas Morse: The provision is justified if it is proposed on the basis that it is a realistic estimate of what is recoverable. That must be true; otherwise, you wouldn’t even be able to do it. It’s true that the debts don’t cease to be recoverable. If the point is that the debts don’t cease to be recoverable then there’s nothing preventing you from continuing to try to recover them. Just making a provision does not mean you are not trying to recover them.

Lin Homer: That’s right. That is the point Simon just made.

Amyas Morse: I understand that point; that’s true. None the less, what you provided must be justified on the basis that it was the most reasonable estimate of what was probably recoverable.

Lin Homer: But the link I’m trying to break is the link to overpayments. If you look at the sum that has been paid out as overpayments, it would not be true to say that we only recovered 30%. We have recovered more than half, and only a smaller proportion has solidified into write-offs, and that is the £4.8 billion.

Q107 Mr Bacon: I was talking about the £985 million.

Lin Homer: Overpayments has been closer to £16 billion.

Q108 Mr Bacon: Hang on. Irrecoverable debt only arises from payments that have been made, yes?

Lin Homer: Yes.

Q109 Mr Bacon: And, ultimately, payments have been made that should not have been made and therefore need to be recovered, but are deemed by you to be irrecoverable.

Lin Homer: Yes.

Q110 Mr Bacon: How can you break the link between that and overpayments? Aren’t they intimately connected?

Lin Homer: They are connected, but it is not true to say that 60% of that £16 billion has been written off; it is £4.8 billion-

Mr Bacon: I didn’t say it was.

Chair: Nobody is saying that.

Q111 Stephen Barclay: If I understand correctly, is there not a distinction between what you have written off and therefore will not claim-that is gone-and a further sum that you could reclaim, in theory, but in practice that is so unlikely that in the accounts you file that as a write-off? Legally, if one of those people wins the lottery, you could go after that money, but in practice you are not doing that.

Lin Homer: Yes.

Q112 Stephen Barclay: So, for the purposes of my constituents-to go back to Mr Bacon’s point-the amount of money overpaid that the Exchequer is unlikely ever to see again is the figure covered by the £985 million. Would that be a fairer way of summarising that, in layman’s language?

Lin Homer: Yes, I think so.

Chair: It is an even worse figure.

Lin Homer: No-

Stephen Barclay: So it is £985 million that the Government could have spent on other things, but which they will not be able to spend because they are very unlikely to see that again. As far as our constituency postbag is concerned, that would be termed as a Government cut, but in this instance it is a provision in the accounts for money we will not see again.

Chair: And that is only the recent addition to the £2.3 billion already written off.

Stephen Barclay: Indeed.

Q113 Mr Jackson: If you accept that the bulk of the losses is unrecovered overpayments that relate to changes in circumstances, let us put that aside and explore deliberate fraud. To use the Home Office expression "fraud prevention upstream"-I know you had success recently with child tax credits with the extended Czech family in Nottingham that had defrauded the public purse of many millions of pounds.

Let me tell a little anecdote. A constituent, who was a landlord, came to my surgery. He had tenants who were masquerading as Portuguese when they were in fact Brazilian. They were claiming working tax credits. He is, fortunately, receiving their bank statements and helpfully observing that they are taking out large cash payments, clearly having a good time in Rio, where they are now domiciled, despite his telling the DWP and HMRC four times that the working tax credit and child tax credit payments should be stopped. They are still taking money out, care of our constituents.

My question is: at the time of forms being filled in and the administrative work being done to set up these claims, particularly for child tax credits in relation to EEA citizens-mindful of the Romanian and Bulgarian free movement from January-what specific concrete steps are being taken to deal with fraud at its commencement?

Lin Homer: I will ask Nick to do that.

Nick Lodge: We do carry out a number of checks at that point. We have an automated screening process through which we screen all claims to try to weed out those that are fraudulent or contain errors before they get into the system, and we have some success with that. We carry out particular checks on documentation for certain types of claimant, and we screen our systems to try to find the kinds of awards that you are describing by understanding the characteristics that apply to them. I will have to sound very disappointed if we have been told about this and have not yet taken any action.

Mr Jackson: I have written to Ms Homer, so she will have my letter.

Nick Lodge: And we carry out a large number of interventions each and every year-about 1.5 million-to try to address different risks in the tax credit system to reduce the amount of error and fraud in the system, which we have done in the latest figures, which have been reduced to 7.3% from 8.1%. That is still too high, but it is a significant move in the right direction, at least.

Q114 Mr Jackson: And the existence of children who are at the centre of the child tax credit claim-are you able to verify that and how do you do it? Is it through birth certificates in the respective countries?

Nick Lodge: Yes. When we have a claim along the lines of the one you describe, our process normally would be to look at the original documentation-a birth certificate and probably one other piece of original documentation as well-and check it to make sure that the child exists, for example. Clearly, sometimes we see forged documents and we have to look out for those, but we do check the documentation.

Q115 Mr Jackson: How were you able to apprehend or successfully prosecute in this case? It has now been expedited and I assume you can talk about it; it was quite a wide-scale extended family fraud involving working tax credits, in the Nottingham area. Was it forensic documentation that you looked at, or was it a tip-off? How did you do it?

Nick Lodge: We liaised closely with other agencies, both here and abroad. I don’t know the absolute details of that case, but I do know that we worked with other agencies on it. These cases take quite a long time to come to fruition because of the evidence gathering and so on; but yes, we would have scanned our systems and drawn together all the information and, as you say, we managed to bring these people to court.

Q116 Mr Jackson: Finally, do you think you have the resource capacity to deal with deliberate fraud-the forensic examination and investigation that is necessary in these cases-going forward over the next two or three years?

Nick Lodge: The number of cases that we refer for prosecution is growing quite fast from a relatively low base. We are doing more and more of that. So far this year, we expect to refer 600 or 700 cases, whereas a few years ago it would have been only a few tens of cases, so it is rapidly growing. We work closely with DWP and are looking at capacity generally in the tax credits system for addressing error and fraud. Only this week we had opened discussions with the market, through issuing a prior information notice, so that we can talk to private sector organisations about how they can help us to reduce error and fraud in the system. This follows a small trial that we ran in the summer. We expect to do quite a lot more in that direction in the coming few months.

Lin Homer: We are also exploring how our Connect system can be used-our data analytics, which you have heard us talk about before. So if you think about that case, which has received quite a lot of publicity, we should increasingly be able to have our data analytics tell us if one set of bank accounts, one address or one agent pops up in too many places. So there will be more use of Connect. The final bit-which is for the future rather than now-is our investment in digital. We received £200 million in the autumn statement, and we will be rolling it out over the next couple of years. Clearly, this will be key particularly to customer service, but we think it will be really useful to us in compliance and fraud as well. Despite the short nature of the future of tax credits with us, we still think it will be worth checking what online work we can do.

So if we have had someone who has made a lot of errors in change of circumstance, let us say, we might require more regular confirmation of circumstance from them than our current system requires. We want to keep playing with this, because we think that observing patterns and behaviours will be one of the ways to get in earlier. It was a big catch to get that whole criminal gang-that is definitely serious organised crime-but at the moment that takes an awful lot of resource to stop. Jenny Granger will tell you that she believes she can take compliance more upstream and do more of that within our data analytics, which gives us more chance.

Q117 Ian Swales: I would like to follow on from what Stewart talked about. The report says that the estimate of error and fraud equates to about £2 billion. Lin, you have heard me ask this question in another context many times. Have you presented a business case to the Treasury to say, "If you give me this amount of money, I will reduce that by x?" It seems to be another example of, if there is £2 billion of fruit on the tree, it is worth paying some money to get at it.

Lin Homer: Yes. Our position is that we will present business cases to our Ministers, broadly, on a six-monthly basis: Budgets and autumn statements. I think that HMRC has probably had some investment at most of those occasions over the past four or five years.

Q118 Ian Swales: Have you done one specifically on this topic?

Lin Homer: We have already done some investment in this area. The notice that Nick alluded to is one of our current projects, so that would be the use of the private sector to increase our capacity in fraud and error.

Q119 Ian Swales: May I build on a question that has just been answered around the question of prosecutions? I can well understand the large resource required for a serious organised crime situation, but you often walk into a shop and see a sign, "We always prosecute shoplifters." Do you always prosecute people who knowingly submit fraudulent documents to you?

Lin Homer: No, I don’t think I could tell you that we do that. We seek to take action whenever we see a situation, but we will use civil approaches and behavioural nudges as well, just because of the numbers. We have 4.8 million families. For instance, in the compliance field, not in tax credits but in tax, we have recently used a "nudge" technique, where we believe there has been over-claiming of some reliefs. Where we would not have had the investigative power to look at all of those, we have asserted a belief that there is an over-claim in tens of thousands of cases, and had something like a 94% response rate to that of people conceding. So, that is not a prosecution but that is action.

Q120 Ian Swales: I understand that, but there is another form of nudge. If you are in a community where people are doing this sort of thing and getting away with it, they tell their friends. If they are prosecuted, they also tell their friends.

Lin Homer: And we tell their friends when we prosecute.

Q121 Ian Swales: Exactly. As I said, I can understand the high resource required where you have a complicated case, but where you have a simple case, where it is clear that somebody has attempted to defraud HMRC, I would have thought the resource required to prosecute was quite low and therefore you should be doing a lot of it. How do you respond to that?

Lin Homer: I think we committed to prosecuting more. That was one of my early commitments to the Chair. When you have the next team here I am sure they will happily give you more information about our prosecutions overall. We agree with you; we think we should prosecute more. You asked me a direct question whether I prosecute everyone; I told you truthfully that the answer is no. There would just be too many, and even a simple prosecution is quite resource-hungry.

Our experience is that of a mixture of civil penalties and refusal of award. You are absolutely right that if a scam does not work, people stop doing it. If we see a scam working, we will do a mixture of things in that space to signal that we are watching. Some of that includes simple advertising. I hope you noticed over the past year quite a lot of our eyes looking out at you. There is some good statistical evidence that adverts like that deter people from this kind of behaviour. We have to use the whole scheme.

Q122 Ian Swales: I will move on. I will leave the thought on record that once again I think the resource required to prosecute should be there, because the payback would be there.

Lin Homer: I might disagree with you on that one, but I hear your comment.

Q123 Ian Swales: Okay. May I move on to figure 11? We have talked about the reduction in fraud and error in certain categories. It is quite stark in figure 11 that the two categories where it appears to have reduced dramatically are both to do with children. I wonder whether there are any messages about what you have done there. Is it a coincidence?

Lin Homer: No, it is hard work.

Q124 Ian Swales: Have you done something specific? Is there any learning for the other categories here?

Nick Lodge: Yes, we have done some specific things. We break down fraud and error into these different risk categories and we plan around each one of them. When Lin and I were before the Committee in March talking about tax credits, we did not have this latest set of figures with us and were looking at the previous year, when the risk, for example, around children had grown quite substantially. We were saying at that time that we were going to do more data matching between tax credits and child benefit, which are two separate systems. Child benefit is a very old system so did not lend itself to much manipulation, but we have done more data matching. We have therefore found more cases where, typically, a young person aged 16 to 19 is no longer in further education and so qualifying for tax credits, and we have been told that-we have picked up those kind of cases. That has enabled us to reduce the amount attaching to children. With child care we have made a specific series of interventions talking directly to child care providers to check what is actually going on.

Q125 Ian Swales: That is helpful. The same chart also shows that the estimate for undeclared partners remains stubbornly high, having gone up in the past two years. Is there any learning you can apply from what you have been doing on children?

Nick Lodge: Yes, I think that there is. The undeclared partner risk and the amount associated with it has gone up roughly in line with entitlement. As you say, it has remained stubbornly high. We devised and began to implement a new series of interventions towards the end of the year to which the figures relate-in about December 2011, and the figures relate to the year 2011-12. We used financial information and information from credit reference agencies to tell us when a couple might be tied together financially but telling us that they are lone claimants. We began to do that and saw some effect, but not a marked effect in that particular year because we started late in the year. The following year, we carried out well over 100,000 interventions and prevented well over £200 million of losses. We would expect that to come through in next year’s figures.

Q126 Mr Bacon: I would like to follow up Mr Lodge’s point about children who are being claimed for because it is claimed that they are in education from the age of 16 onwards. The most recent communication I have from HMRC about this issue relates to a constituency case. I was told by the HMRC person who wrote to me that, "Unfortunately it is not a quick or easy process, and in the circumstances I ask for your patience in this matter." That was dated 23 September 2013. This is a matter that I have been drawing to the attention of HMRC and the CSA since June 2012, so there have been 15 months during which something hasn’t happened. Am I not right that, where a child benefit claim is being made on the basis that the children are still in education, the CSA is obliged to make a maintenance assessment for the ex-spouse to pay maintenance?

Lin Homer: I think that was a letter from me, Mr Bacon.

Q127 Mr Bacon: My constituent seems to be under that impression, because in this particular case the children are not in education, but the claimant, the other parent, has asserted that they are in order to claim child benefit-the allegation is that the claim is fraudulent. We are now at the point where the CSA is telling HMRC, "We have checked-these individuals are not in education," but there is nobody at the other end. It is like knocking on a vacuum.

The point that the CSA made to us is that this is rife. I am only raising it here because it is a systemic problem. There is a particular problem relating to my constituent-I am happy to get my office to draw it to your attention afterwards, Mr Lodge-but the CSA tells us that it is often facing this problem. HMRC tell me that unfortunately it is not a quick or easy process, but with the powers of Government, finding out whether or not someone is in full-time education is not actually a particularly difficult thing to do. My constituent has managed to find it out and the CSA has managed to find it out and verify it, but answers come there none from HMRC, over a period of well over a year.

Lin Homer: I think I have just signed a reply to you on this case, Mr Bacon. It might be a different one, but it sticks in my mind. First, I must say that if there is a systemic issue from CSA’s perspective, we should talk to them about it, and I will take that away.

Q128 Mr Bacon: That is what they told us-that they often have this experience of drawing things to your attention and nothing happens.

Lin Homer: That is helpful. That is not a comment they have made to me, and we of course should look at that. On the individual cases we see there are a couple of things. I am not an absolute expert and I don’t have my letter in front of me but one of the issues for us is that the definition of "in education" includes work where ongoing training is being provided in work. So one of our challenges is that one partner may not know as much about a case as he or she thinks. The second-many of you know this from the letters we write to you-is that we are not always at liberty to tell you if your constituent is one half of a partnership.

Q129 Mr Bacon: I understand that, but when it is asserted that someone is registered at a particular institution and there turns out to be documentary evidence that this is not the case and they are not registered and never have been, it is fair slam dunk. It is not that difficult.

Lin Homer: Well, I may still be in a position where because you are writing on behalf of one constituent and the tax issues relate to another, I am not able to tell you as much as I know. I do believe that in the letter I have just written to you I give you some general information about this. So my apologies if that letter has not got to you. That was much more recent than September. I will certainly take away the CSA issue and I will make sure that we talk to them.

Mr Bacon: I will certainly make sure that Mr Lodge sees the relevant correspondence.

Q130 Ian Swales: On the remaining point, having listened to this discussion, I feel that what we need is confidence that the processes you are operating and spending money on will be leveraged in the new universal credit world. If you are spending £200 million on this, then from a taxpayer, value-for-money point of view, it is self-evident that you should be closely tied to DWP. Either it is spending the same money-

Lin Homer: No. Absolutely.

Q131 Ian Swales: Or it is a joint project.

Lin Homer: Well, first we are spending £200 million over our whole system. This is not £200 million being invested in tax credits. We anticipate we will get a very good rate of return on that in a sense regardless of any benefit in the tax credits area. The universal credit system is innately better than the tax credit system at preventing fraud and error. It is a real-time system. It does not allow claimants to tell us something by error or intention, and require us to wait a long time before we can challenge that. I think the design will design out some fraud and error. RTI will allow that to be much more up to date. But we are determined to try to share fully our experiences of this system with DWP so that the learning is not done twice and so that the investment is not done twice.

Q132 Ian Swales: So no redundancy of cost and work now?

Lin Homer: Absolutely.

Chair: Let’s try and get Steve in on this issue and then we must come back to PAYE.

Nick Lodge: Chair, may I quickly clarify one point? The £200 million was not investment. It was £200 million that we had identified and addressed in error and in fraud. So that is not an investment in a system or anything like that. That is the amount from those interventions that we identified.

Q133 Stephen Barclay: Mr Lodge, may I go back to Mr Jackson’s point, which as I understood it was about benefits being paid into a UK account but withdrawn overseas outside the EU? Would your controls pick up such payments?

Nick Lodge: Increasingly they would, because we are now beginning to liaise much more closely with the banks so that we get a flow of information about such cases that we can then pick up. Historically that would not necessarily have been the case, but we are now doing that.

Q134 Stephen Barclay: But the banks would be aware of that, wouldn’t they? Payments had come in the form of a benefit and been paid out overseas.

Nick Lodge: They are able to identify that, yes.

Q135 Stephen Barclay: Might banks report that as a suspicious activity report to SOCA?

Nick Lodge: I am not sure whether they would, because my understanding of the particular arrangements for that is not good enough. They might do under some money laundering-

Q136 Stephen Barclay: My understanding is several years out of date, but is it the case that there were not industry requirements to report but that some banks would report such transactions as suspicious activity reports?

Lin Homer: I think the money laundering requirements have substantially changed over recent years. We put a lot of obligation on people who handle money to report.

Q137 Stephen Barclay: What I am driving at is, in a case like that identified by Mr Jackson-he has told me which bank it was-might it have filed a SAR with SOCA?

Lin Homer: As we said earlier, we will have involvement with the law enforcement agencies in those big cases.

Q138 Stephen Barclay: Sure. Perhaps you could let us have a note. In the case identified by Mr Jackson, I am interested in whether the information concerning that, as a suspicious activity report, was already sitting with SOCA. Could we have a note on that? And perhaps we could have a note on why that which is clearly preventable-in agreement with the banks, because the banks have that information-is not being addressed.

Lin Homer: We will seek to give you a note. I am not sure, and I might need to talk to Mr Jackson. The case we were talking about predominantly has come to fruition and led to a prosecution.

Q139 Stephen Barclay: Of course it is resource-intensive, and usually it is only triggered by other information such as in the Nottingham case and that leads to an investigation, but the point is that banks are perfectly able, at very low cost, to identify coded payments that come into an account from Government agencies in the form of benefits, and, as a matter of routine fraud controls, know the minute people are withdrawing from accounts overseas. Many constituents will know that if they go overseas and do not notify their bank, sometimes they cannot use their card. These are routine controls. Some-not all-banks, I suspect, report some of those things to SOCA, and yet another part of Government is not speaking to SOCA.

Lin Homer: That is not true. On the National Crime Agency, as it is now-Jennie Granger is heavily involved in the development of SOCA-the NCA. She sits on the programme boards with Keith Bristow. We are looking at ways to develop greater use by us of their data, but increasingly they are using ours.

Q140 Stephen Barclay: You have just confirmed to me that you cannot say there is a process whereby the banks are reporting information that they know about, which would be a standard fraud risk control. If the banks are looking at it with their own data and these payments are taken overseas, straight away that is a fraud alert. What Mr Jackson was alerted to-

Lin Homer: Not necessarily.

Q141 Stephen Barclay: Well, it is not if a customer is notified in advance, but what I am driving at is, if benefits are going into an account and the withdrawals are constantly overseas, I would expect that to be a finite population, which you would want to investigate.

Lin Homer: Yes. And some of the experiments that we are going into involve a greater use of what we would call third party data.

Q142 Stephen Barclay: But these are experiments you are going into. I am saying that this has been known for years. The sums at issue are very large. I cannot understand why we do not have today a process for the information that the banks have-we have confirmed that the banks have this information-on these accounts. It may already be sitting with SOCA in some instances; it may not for others. It would be an easy requirement to ask the British Bankers Association to make that an industry standard. That information is then simply an issue of data matching between who you are paying accounts to and who is withdrawing from the accounts overseas. That creates a population for you to investigate. It baffles me why that control is still not in place.

Q143 Mr Bacon: Particularly since I drew exactly this problem to the attention of the NAO and Customs and Excise many years ago.

Lin Homer: I understand your frustration. I think we are in this space. I do not think it is quite as easy as it sounds, because the threshold for referrals needs to be right so that you are not overwhelmed. There will be many instances of people who are entirely entitled to withdraw money, who may choose to do that overseas on occasion. I mentioned earlier that we think we are now reaching a space where we can do much more data analytics than we could before.

Q144 Stephen Barclay: With respect, you say "on occasion". I am talking about repeated withdrawals overseas. You could set a risk profile of countries that are a higher risk-an obvious one would be Brazil into Portugal-but you are saying that we do not have that.

Lin Homer: No. I am saying that we are now in that space. Traditionally, we were not. We did not have the data analytic capability before to watch those trends and do that matching. We do now and we are beginning to get benefits from doing it. We totally agree with you, but I do not think that Customs and Excise would have had the data analytics capacity to do it.

Chair: We will stop now because we have a vote, but we are coming back.

Sitting suspended.

On resuming-

Chair: We are moving off tax credits. Is everybody happy? Okay.

Q145 Austin Mitchell: Why was there a fall in the amount of overdue tax collected in 2012-13? I think it fell by £3.4 billion.

Lin Homer: In overdue tax collected? Generally or specifically, Mr Mitchell?

Q146 Austin Mitchell: The amount of overdue tax collected in 2012-13 was down £3.4 billion. See paragraph 1.5 of the Report.

Lin Homer: So this is back to debt?

Paul Keane: Yes. Tax debt rather than tax.

Lin Homer: The position we have been in on debt generally is that we have been separating out more clearly stock debt from flow debt and we have been trying to make progress on both at once; in a sense, to clear up the past while improving the future. I think it would be fair to say that it has taken a bit of adjustment to work out how much resource to put on each side of that equation. From my perspective, we have made some adjustments during the year to ensure we were pursuing enough of the flow debt as well as the old debt. It may be that that has led to the change that you are talking about. The difficulty is that I can’t see the figure you refer to in paragraph 1.5.

Paul Keane: It is in the paragraph with the £37.9 billion.

Lin Homer: I follow. Sorry, I just misheard. I was looking for a different figure.

Chair: The amount you have decided "not to pursue for reasons such as hardship or value for money…has increased from £5.17 billion last year to £5.31 billion this year".

Q147 Austin Mitchell: That could be due to more insolvencies.

Lin Homer: That is a different point.

Austin Mitchell: Why was there a drop in the amount of overdue tax collection?

Q148 Chair: Is that not what you meant?

Lin Homer: You are talking about different points, but that’s fine.

Chair: Oh sugar, sorry.

Lin Homer: Don’t worry. Overall debt fell during that period, but that is the amount we collected. That is because we were placing a greater emphasis on the old debt during that period. This links back a little bit to our conversation around tax credit. We have been trying to move into a position whereby we are active on all our debt, as opposed to just chasing the newest debt. In doing that, for a while, you might not be absolutely maximising your return. I think Mr Swales made the point about prosecution. We are trying to send a signal that we will pursue debt, whether it is small, big, old or new. We have altered where our resources go, but overall our proportion of debt collected has gone up and remains very high-Simon might have the figure to hand. We think that the approach of being active in relation to all our debt will allow us to send the message through the system that there is not a sort of line beyond which we will not go-if your debt is this old or this small, don’t worry, they won’t chase you. We are trying to change people’s perception.

Q149 Austin Mitchell: That’s a hope for the future, isn’t it? I mean, £3.4 billion is a big drop.

Lin Homer: Not in relation to-

Austin Mitchell: In relation to a falling take.

Simon Bowles: If we see it in the context that overall tax debt fell from £16 billion to £15.3 billion, there has actually been a reduction overall in the level of debt.

Q150 Chair: We are playing around with different figures. Can we all have the same figures? I am looking at paragraph 1.5.

Lin Homer: You are looking at the end of it.

Chair: In paragraph 1.5-overdue tax, that was the first figure; I went on to the second-in 2011-12, you collected £37.9 billion.

Lin Homer: It was £34.5 compared to £37.9.

Q151 Chair: So that is where Austin gets his £3.4 billion. You have collected less overdue tax.

Q152 Ian Swales: That is partly because you were chasing more old debt in the previous year. Is that what you are saying? That is what we would expect.

Lin Homer: Yes, and because we collect 99% of the taxes due, we are getting a smaller amount overdue. That is a sign of us getting more active in relation to debt.

Q153 Chair: How much was the overdue figure, of which you then collected £37.9 billion in 2011-12, compared with the overdue figure in 2012-13? What was it in the two years: the year where you collected more and the year where-

Lin Homer: It will be in one of the tables. I am sorry, I do not know if I can find it quickly.

Q154 Chair: Can somebody help us, otherwise it is pointless?

Lin Homer: Do you want me to give you a note on that?

Chair: Okay, it moves us on a bit.

Q155 Austin Mitchell: You are saying that it has nothing to do with staff shortages?

Lin Homer: No. We are putting a lot more into debt than we were. Our collection rates are good. As I say, we are moving out our sphere of influence, if you like, and chasing more debt. The question about how far back historically we can go, which is where we were on tax credit, is one issue, but our overall collection rate for debt in relation to tax is extremely good.

Q156 Austin Mitchell: Do you do any calculations about the increase in tax claimed or generated that is produced by employing an extra tax inspector?

Lin Homer: This is not done by tax inspectors. It is done by debt collectors.

Q157 Austin Mitchell: They produce claims about the amount of abuse discovered by DWP inspectors, and I am sure that the return from tax inspectors must be much better than hounding beneficiaries.

Lin Homer: Our return on each member of staff is very good. It depends how you want to calculate it. We collected £476 billion.

Q158 Austin Mitchell: It would be a useful weapon to urge against the Treasury when asking for more money.

Lin Homer: I think the Treasury thinks we are a good investment bet. I said earlier that I do not think we have had an autumn statement or a Budget in recent years without some investment. The Chair herself has quoted the return on investment in a number of our schemes, so you are absolutely right. But to be very clear on debt, we are deploying more resources than we used to and pursuing more of our debt. There is some that historically has not been value for money to collect, and one of the reasons why we have been testing the use of the private sector is to get a return on money that is poor value for money to collect, but you still get some return. One has to be practical. You may get 5p in the pound on some of that debt, but it would still be better than nothing. We would allocate our resources to the best return, so we are looking for models that would allow us to allocate more resourcing, even for the low return. That has been some of our practical work, both in fraud and error and in debt, in using the private sector.

Austin Mitchell: Okay. I want to ask about the Swiss-British agreement-

Q159 Chair: We will come back to that next week. On your private people-this goes a little back to tax credits fraud and error-you signed this contract with Transactis. What did it cost you? What did you get back?

Nick Lodge: We ran a very short proof of concept trial for about five or six weeks in the summer, just to see-

Q160 Chair: It says three months.

Nick Lodge: I think it started right at the end of May and concluded in July, so it spanned three months but it did not actually run for three months. I cannot remember the precise cost, but it was around £50,000-I can correct that if I have got the estimate wrong.4 The purpose of doing that was to check the proposition that a private sector company could carry out the kind of checks that we do to counter tax credits error and fraud: the checks that I described earlier on child care or children. It did show that they could do that successfully, so we are now talking to private sector suppliers-

Q161 Chair: They did not get extra money in.

Nick Lodge: They worked, to a conclusion, on over 5,000 cases and identified about £20 million-worth of fraud and error.

Q162 Chair: So they identified £20 million of fraud and error.

Nick Lodge: They did.

Q163 Chair: For the £50,000 you paid them.

Nick Lodge: Yes. As I said, I will correct that figure if I have remembered it incorrectly. That has shown us enough to want to talk to the potential private sector suppliers in a much more open way about whether we can extend those arrangements and bring them in at scale to improve tax credits capacity.

Chair: We will hear from Amyas first.

Amyas Morse: Following that, have you got a benchmark of what it is costing you to do that in-house so that you know-

Lin Homer: Yes.

Amyas Morse: Have you got the right costing information to do that?

Nick Lodge: We understand how many staff do this work and we understand the return they get from it in terms of losses prevented, so we have a calculation that we can do for that rate of return.

Lin Homer: For both debt and fraud and error, we can match the internal and external performance.

Amyas Morse: As I said, to make good decisions-

Q164 Chair: What is the advantage of the external?

Lin Homer: Potentially two. One is wider reach, and the second is that they may be better at it than us. Either is worth having, so even if they are not better than us, we could give them lower-value work and they could reach to a space that we cannot. We are looking at the sustainability of these models and ways to fund them sustainably, which may include payment by results.

Q165 Ian Swales: That is pretty spectacular: a 400-fold return on the money you spent, if I heard the figures correctly. Did you say £50,000 for £20 million?

Nick Lodge: I will need to check the figures and, clearly, I will correct them if I have-

Q166 Ian Swales: It almost beggars belief. Without necessarily going into all the details, what kind of work were they doing? Was this more sophisticated data matching than you do, or was it boots on the streets going round? Because you do not get that much for £50,000.

Nick Lodge: Transactis are a specialist data company, so they did some data analytics to select cases, and we saw some benefit from their expertise in that in selecting and homing in on the right cases to pick up. Then they were phoning up and writing to claimants to check the circumstances for those cases, to correct them if they were incorrect and pass that information back to us. So it was a combination of data analytics and some-but not very many-boots on the ground. It was quite a small trial; it was not very large numbers.

Q167 Ian Swales: This comes back to something that I keep talking about: investment in order to get out. These amounts of money-these billions-that we see in these sorts of reports buy you an awful lot in terms of people or third-party effort or systems and so on. Is that not a perfect example of how perhaps we should not be as quick either to write things off or to say, "There is this percentage of fraud and error" when that kind of expertise is out there? If they can do that, that suggests to me that, even though we think our systems are good in Government and in the Department, what sounds like a relatively small company has far better systems. What is that telling us?

Lin Homer: We talked earlier today about what we have achieved ourselves. To put that in context, the 1% improvement over the last year that Nick referred to earlier has been achieved internally with relatively small resources, and that also adds up to a pretty big sum. My maths is not very quick, but Nick will probably calculate what proportion of the £1.9 billion less error and fraud that means there is in the system. I would therefore slightly challenge the view that we do not have the capacity to do these things ourselves.

The question for us is whether we can create mechanisms whereby we don’t think we have to do it all ourselves-we can use a mixed economy. One thing you can do is deploy your resources more flexibly. We are trying that in contact centre work. Our view is that, in a number of these areas, having a mixture of resources to deploy is useful. Very often, ours will be the most skilled, but we can share and deploy and use other people. We think this is a rich field and we are very interested. We are looking at it within customer service debt and fraud and error, and we think it is a big opportunity for the future.

Q168 Ian Swales: Can I pick up another couple of points? Figure 5 of the NAO Report summarises the work with which the Committee is familiar-the catch-up work you have been doing. It says that the tax forgone over the eight or so years concerned is £953 million. The first item shows that you temporarily changed the threshold for cases that you looked at-have you now changed that back?

Lin Homer: Yes. This linked to the introduction of NPS. I know that you have spent a lot of time looking at this. For the debts that had occurred in the period from 2008-09 through to 2011-12, when we retrospectively tried to pursue them, we changed the threshold from 50 to 300.

Q169 Chair: How much did that mean in terms of estimated loss of revenue?

Lin Homer: That is the £266 million. We have changed that back. If I am honest-you alluded to this earlier-I think that that was a practical response to the impact on the taxpayer and us, but we now believe that we can and should maintain that lower level of tolerance. As we were discussing earlier, with people having greater understanding about their position, it is hopefully less easy for taxpayers to get into debt without realising and we can therefore pursue more confidently because there has not been confusion on their part-it has been more deliberate.

Q170 Chair: May I come in on that issue? It is a lot of money, in the end. Congratulations on having sorted it out-in the end-but £1 billion of revenue was lost because of the bad way in which it was handled. Has anybody ever been held responsible and accountable for that?

Lin Homer: I am afraid I think it is a systemic issue. I do not think that anybody sat there and said, "Write off £953 million." I suspect that the question about the open cases-how many there would be and how long they would run-was not well anticipated. We have to learn the lesson of anticipating that better, but a fairly big chunk of that ended up being time served-we just ran out of time to pursue them. We have tightened up our systems and you would rightly be able to hold me to account if in four years’ time I was back telling you-

Chair: You’ll be off then.

Lin Homer: You hope.

Q171 Ian Swales: It is a hidden cost of the systems and reorganisation overload.

Lin Homer: To some extent, our old PAYE system was not as accurate as we or others thought for many years. In a way, the good news is that it is more accurate now, but in the years we were making it more accurate, we had an opportunity to pursue those greater accuracies in a sense that was greater than we managed to grasp. I would accept that.

Q172 Ian Swales: A final point from me. We have been looking at paragraph 1.5. The last bit, which we have not yet mentioned, says that the amount you "decided not to pursue for reasons such as hardship or value for money…has increased from £5.17 billion last year to £5.31 billion this year." Again, that is an enormous amount of money. That could cover a wide range of circumstances from too small to chase, to bankrupt companies or whatever. Can you tell us a bit more about what is behind those figures because, as I say, they are enormous?

Simon Bowles: Yes, I can tell the Committee that over 90% of that related to insolvency.

Q173 Chair: Over 90%?

Simon Bowles: That’s right.5

Q174 Ian Swales: So that would be tax assessed but the company had gone away or gone bust. I guess this is partly a policy point, but do you do any work to look at where a company closes down and the directors-I forget what the term is-

Chair: Phoenix.

Ian Swales: That’s it. Often HMRC is one of the big losers in situations like that. Do you do any work to pursue directors who may have engaged in that kind of activity?

Simon Bowles: We are doing quite a lot of work on that. I would not want the Committee to think that we just accept that over 90% being insolvent is the end of it. We are trying to improve risk assessment and have better trained and earlier interventions. We are also doing some more work to look at how we might better secure HMRC as a creditor.

Q175 Chair: What about the phoenix company?

Lin Homer: We try through our criminal investigation and our tax investigations base to tag and watch people with a history.

Q176 Chair: Any success on that?

Lin Homer: Yes. We do have some success but our systems allow companies to go down and come up. One of the spaces you will sometimes perceive us in, and you may indeed pressurise us to be kinder, is in time to pay. We will not always exercise our discretion to offer time to pay if we believe the company is at risk of insolvency, whether deliberately, or just that they are no longer sustainable. If giving them longer to pay leaves them falling off, we are then standing in line with creditors. In areas where we think we can envisage known characters appearing again, we will exercise more observation and closer scrutiny of those companies.

Q177 Ian Swales: Being more specific, what powers do you have with phoenix companies? I know enough about corporate law to know that a company is like a person. If they die you cannot resurrect them. Do you have any powers to deal with people who do that kind of thing?

Lin Homer: I haven’t come with this as my specialist subject today. I think it would be a fair question to put to the team on 28 October. I know that we exercise scrutiny of those companies. I am afraid I am not good enough on the company law side of this to be able to say whether we can exercise discretion. What we can do is use our criminal powers to investigate where we perceive a pattern of behaviour. So, for instance, in some of the MTIC fraud, where we have been very successful, we saw people appearing more than once. At the end of the day, if the legal entity is properly stood up and put down, there are limits to what we can do. We have discussed some of this around charities, haven’t we, as well? So we exercise what powers we can but I am afraid that I could not off the top of my head tell you our position in relation to phoenix.

Q178 Ian Swales: Last question: of the £5 billion-ish written off, in what categories of tax does that mainly fall?

Paul Keane: That is in note 7.2 to the accounts. Income tax in total was £1.3 billion; VAT £2 billion and so on.

Ian Swales: You caught me out. I know the rest of it. I had just forgotten that bit.

Simon Bowles: It is a gripping read.

Q179 Mr Bacon: I should like to ask about the construction industry scheme, which as far as I can see rates one mention in the Report on page 169 about a bulk remission. It concerns £206 million of penalties relating to 1.2 million cases that were considered irrecoverable or where there was a low likelihood of recovery. Do you think that the construction industry scheme is now an expensive anachronism?

Lin Homer: The construction industry scheme is quite complex. We have gained some experience of penalties in a number of elements. Two things are going on here. The first is the scheme itself, which the construction industry would probably say is very necessary and important. The other is the way we tie penalties to schemes to encourage them to be well managed.

Q180 Mr Bacon: I was alluding to the penalties because they were the only reference I could find in the report. My question was not about penalties, but about whether you think it is now an expensive anachronism.

Lin Homer: That is a policy question at its heart, so I am not going to answer that. From the letters I get, it remains strongly supported by the industry itself.

Q181 Mr Bacon: When you say that, it’s interesting to me. Which bits of the industry write to you supporting it?

Lin Homer: We get a lot of pressure when payments due under it are either challenged or slow.

Q182 Mr Bacon: Do you mean payments by you back-

Lin Homer: Back.

Q183 Mr Bacon: That is not quite the same as supporting it. "Dear HMRC, I had some deductions taken from me by my main contractor as I, a sub-contractor, have to be a member of this scheme, and I am not given gross payment status. The contractor has duly paid over this money, but you haven’t paid it back to me yet. I have therefore given you a free loan for 15 months. Please pay." Frankly, that does not sound to me like support for the scheme; it sounds like me wanting my money back. Are you seriously praying in aid that type of letter as support for the scheme?

Lin Homer: No, but I am saying I get-

Q184 Mr Bacon: I think you will agree that it is not evident that that is support for the scheme.

Lin Homer: My point is that the people using the scheme do get in contact with us about it.

Q185 Mr Bacon: I didn’t expect to have to belabour this point, but that is not support for the scheme. They are entrapped in it, and they want to make it work, because they want their money back.

Lin Homer: Your view.

Mr Bacon: You mean they don’t want their money back?

Lin Homer: If they are due it, they will get it back. On the broader issue, you know that we put into effect the policies we are given.

Q186 Mr Bacon: I understand that. I am interested in the effectiveness of the scheme-whether it achieves its aim. If someone who is necessarily enmeshed in the scheme wants it to work for them, that is not the same as saying that they support the scheme.

Lin Homer: No. I am happy to take back the word "support".

Q187 Mr Bacon: Who writes to you saying that they support the scheme? You said, "From the letters I get from industry, the scheme is well supported."

Lin Homer: I am happy to withdraw the word "support". What I am saying is that I get a lot of letters indicating that people are dependent on the scheme. I am not going to enter into a debate with you about the policy. I have not had anything like the amount of conversation about the effectiveness of the scheme as I have about some other areas, so I have not been lobbied that it is ineffective or administratively burdensome.

Q188 Mr Bacon: What does the scheme cost to administer?

Lin Homer: I don’t know that off the top of my head.

Q189 Mr Bacon: The three numbers that I have from Government-these are all Government figures-are, a first estimate of £52 million, which went down to £30 million, and which later was admitted to be much higher than that-possibly £100 million.

Lin Homer: I am happy to give you a note. It is not something I know off the top of my head.

Q190 Mr Bacon: It sounds to me like there isn’t a clear handle on how much it costs. I do not know if work is now being done on establishing what it costs-

Lin Homer: That is not true, Mr Bacon. I just don’t come with every fact about all our systems in my head. I am very happy to give you a note, and I am confident that we have a handle on it. It was not a question that I anticipated. As you say, it was not a big feature of the annual report.

Q191 Mr Bacon: No, indeed.

When the scheme was set up-I think, 40 years ago, in the early 1970s-there was a structure in the construction industry whereby a great deal of cash was paid. People would turn up with security vans with literally lorry-loads of wage packets; they were called wage packets. They were counted out and put into envelopes, and then handed to people in cash. It was to combat potential tax evasion in those circumstances that the scheme was originally introduced in 1972, wasn’t it?

Lin Homer: I wasn’t there.

Mr Bacon: Neither, I hasten to add, was I.

Lin Homer: I think we were probably both at school in East Anglia at the time.

Q192 Mr Bacon: Well, yes.

Given that that is not quite the picture now, do you happen to know what proportion of workers in this sector are paid in cash?

Lin Homer: No. I think we have a briefing coming up with you on VAT. I am sure we could come along with a note on the construction industry scheme.

Q193 Mr Bacon: You spent taxpayers’ money on asking Ipsos MORI to ask this question. If you ask subcontractors how contractors pay them for subcontracting work, cash is 4%. It is a tiny amount.

Lin Homer: I would hope so.

Q194 Mr Bacon: Indeed. Not only because it is more secure, cheaper, quicker and more traceable to do it by direct payments; a lot of it happens now by direct payments, and the industry is not anything like what it was. Even for those that are not deemed large enough businesses to qualify for gross payment status, a significant proportion of the others will be small to medium-sized businesses that derive their own money from larger audited businesses, where it can be easily traced, and even the really small self-employed ones are getting their money mostly as direct payments, not through cash, are they not?

Lin Homer: I am not an expert on this scheme. I am happy to enter into a conversation with you about it, but I am not going to give you good value this afternoon. You clearly do not think it is a good scheme.

Q195 Mr Bacon: It has been put to me by people, including the Institute of Chartered Accountants in England and Wales, that now-

Lin Homer: They have not put that to me, interestingly enough, and I see them quite regularly.

Q196 Mr Bacon: They may be knocking on your door.

Lin Homer: I would be very happy to talk to them.

Q197 Mr Bacon: Given the structural changes that we have seen in the industry, there are now questions that might not have been there at the beginning about the scheme’s effectiveness and efficiency, and indeed its economy, because of the cost of running it.

Lin Homer: Well, I would be very happy to talk to them.

Q198 Austin Mitchell: The agreement between Britain and Switzerland over unpaid tax-

Chair: Austin, we are going to deal with Switzerland and all the tax avoidance stuff next week-the whole lot together. We all have lots of questions to ask on that, so save it until next week.

Q199 Austin Mitchell: You will not raise an estimated £4.4 billion over the next three years, particularly since part of that will be raised directly from individuals who will suddenly break down and confess, "I’ve got a Swiss bank account" and hand over the money. That was my thought. I was going to give you that thought.

Q200 Chair: Okay. Give it to them next week. I have two final questions. One concerns customer performance. We talked earlier about your failure to answer calls on the child tax credit. If you look at appendix B of the Report that we got from the NAO, you will see that performance appears to have deteriorated in the first quarter of 2013-14.

Paul Keane: It is appendix B to the brief we have given the Committee, but it is from HMRC’s Business Plan Indicators-Quarterly Performance: Quarter 1-April to June 2013.

Q201 Chair: Okay. If I just take that. Post received and dealt with within 15 days was 85% in the fourth quarter of 2012-13, and in the first quarter of 2013-14 that was down to 70.3%. The percentage of calls "attempt handled"-I don’t know what means; I assume it means answered-

Ruth Owen: Answered, yes.

Q202 Chair: "Percentage of calls" answered "by our Contact Centres within 5 minutes"-which is dire-is now at 57.7%, and those seem pretty grim statistics. Apologies if you have not got them, but Ruth will be familiar with them.

Lin Homer: Before Ruth gives you the update, we do not think customer service is good enough yet, so do not for a moment think we are yet content. We are making some underlying improvements. Again, I would credit Ruth and her team with working very hard in this space. The best comparators are quarter to quarter, because our business is very different in its seasons.

Q203 Chair: Let me take you from quarter to quarter. Quarter 1 of last year, on the answering of the letters, was 77.4%; it is down to 70.3% this year. On the other one-calls handled-you obviously were not collating the stats last year.

Lin Homer: I think we were, but Ruth can give you the detail.

Austin Mitchell: How do we tie in the policy, which is mentioned in paragraph 7, of getting more tax by co-operation and by encouraging businesses and individuals to comply with their obligations-the job of an accountant or whatever-with the fall in responses that we have just had?

Lin Homer: Quarter 1 of this year.

Ruth Owen: So on calls answered, in quarter 1 of this year, we answered 78% of calls compared with the same period the previous year, which was 67%. The same quarter the previous year was 56%. None of those is where we want to be, but there is an underlying trend, even in what is a busy quarter for us, of us actually understanding how to manage supply and demand.

Q204 Chair: The first quarter, of course, is when most people ring you up, isn’t it?

Ruth Owen: It is a busy quarter. I gave you some figures the last time I was here to talk about 0845 numbers. At the year to date, we are at 73% over the whole year, so just over halfway.

Q205 Chair: You are at what?

Ruth Owen: 73% of calls answered.

Q206 Chair: Calls answered within five minutes?

Ruth Owen: No, calls answered.

Q207 Chair: Calls answered full stop? This is where Fiona hangs on for 20 minutes before she gets an answer.

Ruth Owen: Hopefully less now. Compare that with last year when it was 66%. Year on year, we are trying to get better. We know the first half of the year is always difficult. In the second half of last year, we were consistently at 90% and I expect to do the same this year, which is why when I last spoke to you-I wrote to you afterwards-solving how we manage the summer and the peak of that tax credit activity is the key to getting this customer service right.

Q208 Chair: You have also closed your-people cannot walk in any more.

Ruth Owen: We have not closed any offices, no.

Lin Homer: We have a pilot going in the north-east, which is providing telephony as an alternative, and we will be looking at that. I have sat in on some of those calls, and I must say that they are providing an extraordinarily high level of service to vulnerable customers.

Q209 Chair: When they get through.

Lin Homer: No, no, these are calls that are being answered quite quickly. They are also being consistently answered by the same person, who is giving as long as it takes. This is not our normal contact centre. We have not closed an inquiry centre and they are just joining the normal queue. We are directing the vulnerable customers to our most specialised telephonists. I think it will be worth talking it through with you because it is going to be an improvement in service. Ruth has faced up to the bits that she is still not happy with, but on the last occasion that I was here with her, she was committing to changing to 0300 numbers, and we completed that by the end of September. That has gone faster and was easier than we expected.

Q210 Chair: It is faster, is it?

Ruth Owen: It was faster to deliver it-that is what we mean, not faster to get through. It is cheaper to get through.

Lin Homer: That has moved some of those costs, which you rightly challenged us on, off the customer. We do continue to be very alive to our need to improve customer service.

Q211 Chair: Okay. We will keep coming back to it.

Lin Homer: Of course you will.

Chair: The other thing is to talk about the child benefit changes. You were quite robust last week. You said that 165,000 people should get off their backsides and register with HMRC. First, 165,000 is what percentage of the cohort where at least one earner earns more than £50,000?

Lin Homer: So 1.1 million families were affected by the change in high income.

Chair: So it is about 16%.

Lin Homer: Some 325,000 opted out by 7 January. Another 65,000 or so have opted out since, and 425,000 were in self-assessment anyway. We also needed to get in another 325,000. At the point where I used the "backside" word, which may not have been wise but did get some attention, our anxiety was that those people were facing penalties if they continued to take no action between now and the end of the year. About 165,000 have registered and about 165,000 have not. I actually gave out an overly optimistic figure about a week ago, because immediately after the weekend, we thought that that had dropped to 110,000, but we got over-excited and double counted some people in that weekend. We still have 165,000 people that we need to get to do something; otherwise they will still owe us the child benefit that they have had, which for 7 January to March for one child is about £263, but the worst-case scenario is that they could then owe us as much again. So from now until the end of the year, we will continue to try-by direct letter, by media and by any means we can-to get the rest of those people to take action before the due date.

Q212 Fiona Mactaggart: How much is it costing you to do that?

Lin Homer: Much less than we anticipated, because so many people have been compliant. Over 90% of people are now compliant in the system. The vast majority of those did it online. We had far fewer calls than we expected. I have reduced the amount we expected to spend on this from the early 20s to about £13 million; that is not bad for a 1.7 billion response.

But I am keen, in this first year, that we put all the effort in that we need to, because these people are losing a benefit. That, in a sense, is what the Government wanted to happen, but we do not want them to face a penalty as well. So you will see us continue to try and chase down that 165,000. It is actually about 145,000 now.

Q213 Chair: These are all households where one person earns-

Lin Homer: Or the other.

Q214 Chair: So they are not how we would define vulnerable citizens.

Lin Homer: No.

Q215 Chair: Yet getting them to change their behaviour is proving extremely difficult.

Lin Homer: Well, getting around 90% compliant within the first six months of introduction is pleasing. There is an element of human nature-we see it in tax credit renewal and in self-assessment-where people leave it really close to the end date before they do anything, so-

Q216 Chair: I was going to observe that if it is that difficult-I accept that 90% do, but 10% is quite a lot of people.

Lin Homer: It will be less than that by the end of the year. I am confident of that.

Q217 Chair: What does that tell you about how you implement these changes? I am thinking about universal credit, on the other side of the spectrum, where you will be dealing with vulnerable citizens and you are expecting behaviour change.

Lin Homer: Well, throughout the tax system in this country, we consistently see about 90% compliance. It tells us that we are a great deal better than many other countries. The natural level of voluntary compliance, I believe, is high, but people have a right to expect the system to help them keep their affairs in order by being simple and understandable.

Q218 Mr Bacon: May I clarify, for the avoidance of doubt, that those 425,000 who are in self-assessment anyway would not be subject to penalties, but would just be expected, when the self-assessment comes, to make a declaration?

Lin Homer: The self-assessment has a very clear box in it which says, "Have you or your partner received child benefit?" We can tell them, effectively, how much they have received. If they do their self-assessment by 30 December, we can code that sum out so they do not even have to pay it back in a lump sum. If they leave it to the last minute and do it on 31 January, they are more likely to have to pay us back in a lump sum, but it is built into the SA system. So yes, it is there for them.

Q219 Mr Bacon: Okay. That is the first part. The second part is: in terms of claiming child benefit, if you are a person who comes into this category-I should know this-are you then able to continue claiming child benefit ad infinitum but it is then completely taxed away?

Lin Homer: Yes.

Q220 Mr Bacon: There is not a sword of Damocles that comes down and prevents you from claiming it.

Lin Homer: No, because the Government’s view was that child benefit has been universal and they wanted it to remain like that. So if you choose to keep having it here, and have it coded out there, that is one choice. For people with between £50,000 and £60,000 of income, that is the right thing to do, because they will be partial beneficiaries. So it is very important that people make-

Q221 Mr Bacon: It is when you earn over £60,000 that you are not a beneficiary at all.

Lin Homer: Over £60,000, you would pay back what you get, but it is income, not earnings. That is why we put the calculator on the site, so people could work out if it applied to them.

Chair: Thank you very much indeed.