Public Accounts Committee
Oral evidence: HMRC’s contract with Concentrix, HC 998
Wednesday 25 January 2017
Ordered by the House of Commons to be published on 25 January 2017.
Watch the meeting
Members present: Meg Hillier (Chair); Mr Richard Bacon; Philip Boswell; Charlie Elphicke; Kevin Foster; Kwasi Kwarteng; Nigel Mills; Anne Marie Morris; Bridget Phillipson; John Pugh; Mrs Anne-Marie Trevelyan.
Sir Amyas Morse, Comptroller and Auditor General, John Thorpe, Executive Leader, NAO, Claire Rollo, Director, NAO, Rob Prideaux, Director, NAO, and Richard Brown, Treasury Officer of Accounts, HM Treasury, were in attendance.
Questions 1-86
Witnesses
I: Philip Cassidy, Senior Vice President, Concentrix, Nick Lodge, Director General Transformation, HMRC, Mark Oatridge, Commercial Manager, Concentrix, and Jon Thompson, Chief Executive and Permanent Secretary, HMRC.
Written evidence from witnesses:
– [Add names of witnesses and hyperlink to submissions]
Report by the Comptroller and Auditor General
Investigation into HMRC’s contract with Concentrix (HC 915)
Philip Cassidy, Nick Lodge, Mark Oatridge and Jon Thompson.
Chair: Good afternoon, and welcome to the Public Accounts Committee on Wednesday 25 January 2017. We have two sessions, both related to HMRC. One is to look at the investigation that the National Audit Office have completed on the outsourcing by HMRC of some of the tax credit work to Concentrix, so this is a subject we have returned to a lot. It has been a big feature on the Floor of the House of Commons and in the country, but we are now hoping to get to the bottom of what happened, through this inquiry.
Our second session, which we hope will start in about 40 minutes’ time—that is a hint to witnesses about the length of the answers we are expecting—is on the HMRC estate programme. It has been a feature of HMRC’s future planning for some time but is now in year one of a 10-year programme, so we are looking at it at a conveniently early stage.
For the first session, our hashtag, for anyone following, is #concentrix. I would like to introduce our witnesses. From my left to right we have Nick Lodge, who is the director general for transformation at HMRC—welcome, Mr Lodge. We also have Jon Thompson, who is a frequent flier here as chief executive and permanent secretary of HMRC. Welcome back, Mr Thompson. Then we have Philip Cassidy, who is the senior vice-president, Concentrix, and Mark Oatridge, the commercial manager, Concentrix. Perhaps it would be helpful if I could ask our Concentrix witnesses this. Titles like senior vice-president are confusing. Can you just explain, both of you, what your involvement was with this contract? Just briefly, Mr Cassidy, where do you fit in the pecking order?
Philip Cassidy: My name is Philip Cassidy, senior vice-president. I suppose my responsibility is looking after a number of different areas within Concentrix globally—pricing, solutions and project management. As part of that, I am on the senior executive team for Concentrix, and as part of that, because it was such a high-profile case or client, I was asked to be sponsor for that account, so I was sponsor for the overall account.
Q1 Chair: You were sponsor throughout the terms of the contract?
Philip Cassidy: Yes.
Q2 Chair: Mr Oatridge, were you closer to the ground?
Mark Oatridge: Yes. My name is Mark Oatridge, commercial manager at Concentrix. I was around on the inception of the deal, just after when it was priced and when it was brought on. Then I was back on to the deal from around December 2015 and throughout last year in a commercial manager role.
Chair: The acoustics in this building, designed though it is for speaking—clearly the Victorians had louder voices—are terrible, so if you could all speak clearly, and we do the same, we will get through the business more quickly. I am going to hand straight over to Charlie Elphicke, who is going to kick off our questions.
Q3 Charlie Elphicke: Good afternoon, gentlemen. Mr Cassidy, can I first address this question to you: looking at the contract throughout, if one looks at figures 12 and 13 of the Report, on pages 34 and 36, would you accept, first, that it is clear that over half of the performance targets were missed, and that overall claimants got a poor deal in the management of this contract?
Philip Cassidy: A number of performance targets were added throughout the period. I think these refer to black and white, in terms of it was delivered or not. We received volumes of over 400% of what was expected, and as a result service levels were impacted within certain periods within that. Ultimately I do accept that there were service levels that were missed, but we prioritised on the service levels that were going to be impacting claimants specifically.
Q4 Charlie Elphicke: Mr Oatridge, you were in charge of operations. During this two-year period that you were in charge of the operations, all these performance targets were being missed. What was your analysis of that, and what was your response?
Mark Oatridge: To clarify, I was in the role from December 2015. From last year our performance targets were slowly increasing. A change was adopted in October 2015 that crystallised change and helped to improve on the performance. Throughout 2016, our KPIs were slowly increasing, although as Mr Cassidy states, we accept that some were below target.
Q5 Charlie Elphicke: Mr Thompson, you came before this Committee in June 2016. It was clear at that time that Concentrix was not meeting targets around accuracy, opening post and answering calls. The KPI situation was deteriorating. Why did you give us the impression in June that you thought that things would get better?
Jon Thompson: My letter to you of 26 July 2016, paragraph 3, does indeed say that it has been the case that levels of customer services have not been acceptable, but we had been working hard with Concentrix to improve, particularly the speed of answering the telephone.
Q6 Chair: That was in July. What you are citing was in June.
Jon Thompson: Sorry, I was going to go on and explain that, in June and July 2016, as the Report says, the speed of answering the telephone was within the KPIs. That is why I said that I appreciate this had some anxiety for tax credit claimants. I apologise for that—let me repeat that apology again—but we had been working hard to bring the service up to an acceptable level in that period. The Report tells you what then happened in August. That is why I gave you the answer I did on 26 July.
Q7 Charlie Elphicke: Even in that time, about 35% of calls were answered in five minutes. The call answering, even at that time, was shocking. Why was action not being taken more seriously at that time and why was that not being monitored?
Jon Thompson: The data shows that, in June, the average speed of answering was 2.3 minutes. That seems to me to be a perfectly reasonable performance. To answer your opening question to Mr Cassidy, I would like to apologise to all customers. Customers should be at the heart of this system. That clearly was not what happened. The payment of a timely and accurate benefit should be the goal. That was not what happened.
Q8 Charlie Elphicke: Mr Cassidy, do you accept that you are at fault here and, as Mr Thompson says, that this is your failure?
Philip Cassidy: Absolutely not. We received more than 400% of volume than we planned for and agreed with HMRC. We planned what we were expecting to receive using HMRC models and our own experience. We estimated what sort of volume we were going to get in. We planned our headcount around those volumes. We delivered our headcount around those volumes. What actually happened in early August was that 45,000 annual declaration terminations went out from HMRC, with no intervention from Concentrix whatsoever. They went out as a stop payment. That was more than double what our expectation was and what our forecasts were. That is what drove significant volume into the system, and even though we received more than 400%, we actually answered more than 300% of the volume that we had planned. Mr Thompson refers to the fact that it was a failure of customer service. We absolutely do not accept that that was the sole responsibility of Concentrix.
Q9 Charlie Elphicke: You are saying that you suddenly ended up with a huge number of cases, you did not expect that huge number of cases and you could not manage that huge number of cases. Is that what you are saying?
Philip Cassidy: We had planned for a certain amount of cases. We had staffed to significantly more cases—about 15% more cases that we could handle. Then we put in contingency plans that included getting overtime, moving staff off casework and on to telephones. There were a number of different areas and we did all those things. We moved so that we could answer telephone calls and work to get the claimants’ issues sorted. We did all that and we got emails from HMRC staff—operational people—saying, “We really appreciate all the work you are doing.”
Q10 Chair: We’re replaying this from a previous hearing, so because of time—
Philip Cassidy: We did everything we said we could do. The only additional thing we could have done was get additional trained staff on to the system. Part of that was the HMRC staff, which we requested, but that did not come through.
Q11 Charlie Elphicke: Mr Cassidy, 31 July was the review deadline, so you knew this was coming. Why were you suddenly caught in a position where you were surprised?
Philip Cassidy: There were 45,000 cases against what we expected, which was 21,000. That went out and stopped payments—
Q12 Charlie Elphicke: Why do you think there were so many more cases than you were expecting?
Philip Cassidy: I can’t answer that. Claimants did not reply to the annual declaration request that came from HMRC. We had no intervention in that whatsoever.
Q13 Charlie Elphicke: Mr Lodge, you have ownership of this in HMRC. Can I just take you paragraph 3.30 of the Report? It says that, in 2015, there was effectively a six-week grace period—terminations were mandated over a six-week period—whereas in 2016, they were much more immediate. Can you explain how that happened and why that was?
Nick Lodge: Yes. Our normal practice is to give claimants four months. As you have mentioned, 31 July is the deadline for people to renew their claim and, in effect, make a new claim for the current year. Obviously, we send reminders and exhort people to contact us, but if they do not make a claim by that date, we stop payment. If we don’t stop payments quickly, claimants find themselves in a position of building up an overpayment—a debt, if you like—that they then need to repay.
It is very important, and is our normal practice every year, to stop payments very soon after 31 July for those reasons. That’s what we did in 2016, as you said. There was no reason for anybody to think that we were going to do anything other than that or to be surprised by that.
Q14 Charlie Elphicke: But you didn’t in 2015. In 2015, it was over a six-week period, but it wasn’t in 2016. Why?
Nick Lodge: Because in 2015, Concentrix came to us and expressly requested that we do something different with terminations, as I think the NAO Report sets out. We agreed to do that as a one-off exception.
Q15 Charlie Elphicke: Can I take you to figure 17 on page 41? If you look at the call answering line, you will see that call answering basically goes down from June and July onwards, and the same pattern is repeated. It didn’t end in calamity in 2015 and, from the evidence, it seems to me that that was because there was a six-week period so that it could be smoothed over that time.
This time, you just did a much faster system. Mr Cassidy is saying that ended up with him unable to cope with the level or renewals that were happening, which then ended in calamity. Do you think you made a mistake in not having a six-week review period, as was done the previous year?
Nick Lodge: Mr Cassidy and I take a different view of what happened over that period. As the NAO Report sets out, a wide range of factors came together in August. Concentrix didn’t stick to its plan for clearing cases; it built up big backlogs. They also had an IT outage on, I think, 11 August. There were a whole raft of things that generated additional calls. We had information from claimants saying that they had to phone repeatedly to get through and so on and so forth.
The 45,000 cases would, in fact, have been fewer had Concentrix stuck to its plan, as is set out in the NAO Report. It wasn’t was because they did not or were not able to stick to that plan. When those payments were stopped in the normal way, Concentrix’s plan should have taken account of that.
Q16 Chair: Mr Cassidy, it is only fair to give you an opportunity to respond to what Mr Lodge has said.
Philip Cassidy: Yes. Mr Lodge refers to the fact that this was caused by an IT issue that happened over a one-day period. We had an issue on 11 August, as Mr Lodge said. It was something that impacted a number of Government Departments, and it was one of our third parties. We got it resolved within 26 hours, and we worked very hard to get it resolved within that time period.
Whatever way I look at this, a one-day outage does not have an impact on the sorts of volumes that we got. We got 400% more volume than we jointly expected.
Q17 Chair: What about what Mr Lodge said about the backlog of cases that he says you built up?
Philip Cassidy: We manage the cases in order to ensure that the service levels were achieved throughout that period. We would have reduced casework in line with phone calls coming in, and that was the driver behind that. It was to manage the phone volume coming in, and that was something that was agreed with HMRC. At all times, we had daily calls, weekly calls and monthly calls. On none of those was the fact that Mr Lodge refers to called out and raised as a major issue to us. In actual fact, it was the reverse of that. We received many congratulatory emails, many congratulatory—
Q18 Charlie Elphicke: Hold on. The key question here is: if there had been staggered terminations, as there were in 2015, would we be in this position today?
Philip Cassidy: I believe that if we had staggered it, we would not be in this position today.
Q19 Charlie Elphicke: Mr Lodge, do you accept responsibility that you ought to have staggered the terminations like you did in 2015?
Nick Lodge: I do not think that staggering the terminations would have made the difference that Mr Cassidy says. I think that one of the lessons for us in all of this, as Jon has said, is testing more thoroughly the contingency arrangements that should have applied during this period. We could and should have tested those more thoroughly to ensure that when they kicked in, they worked better. If those terminations had been staggered, we would clearly have looked at that, had we been asked to do so, but we were not. Had those terminations been staggered, it might have made a marginal difference, but I think the list of issues in the NAO Report clearly says that that was only one of a number of issues that contributed.
Q20 Chair: While we are on the issue of staggering—a simple answer to this question, please—when you took on the contract, you had done it the previous year with a six-week period. Did you expect it to be six weeks in 2016?
Philip Cassidy: We expected it to be over a longer period, not necessarily six weeks.
Q21 Chair: One week was a surprise to you.
Philip Cassidy: It was.
Q22 Chair: Did you at any point ask Mr Lodge or anyone at HMRC for longer? You said that you did not request a longer period. Can we just be clear? You are saying that you did not expect it to be one week.
Philip Cassidy: No, normally there would be a discussion about a change like the one made to the six-week period. There was not a discussion about that in the daily, weekly or monthly—
Q23 Chair: Who didn’t initiate that discussion? Mr Lodge seems to think that you should have asked. You think that they should have talked to you before making it one week. Am I paraphrasing correctly?
Philip Cassidy: My view is that if it was going to be over a much shorter period and there was going to be significantly more volume than planned—I think there are two issues here—we should have at least had a discussion about it, and that would normally be instigated by HMRC.
Q24 Chair: Mr Lodge, on that point, who do you think should have initiated the conversation? You said that they did not ask you, quite pointedly, I thought.
Nick Lodge: We offered help in a general sense throughout the period, but we were not asked to do anything different to our usual process. Our usual process, as I have explained and as is set out in the Report, is to stop payments where people have not—
Q25 Chair: You heard Mr Cassidy say that he did not know that it was going to be a week. Do you agree with that, or do you think that you made clear what your position was in normal circumstances?
Nick Lodge: The way that renewals work was made clear to Concentrix. There were workshops to run through in detail precisely how that renewals process worked, and there were reminders that terminations of those who had not made a claim for the current year were coming up. So far as I am aware, Concentrix would have known about that.
Chair: Mr Cassidy, I think you need to come back on that point. I am going to bring in Sir Amyas Morse after that.
Philip Cassidy: If I could read the email that was sent, it says: “I can confirm that the start date for all the terminations will be Sunday 7 August 2016. Concentrix terminations will be included with the rest of HMRC. The estimate is that approximately 550,000 terminations will be processed.” We did not get the number or the fact that there would be more than double what the expectation was in the plan.
Q26 Chair: What email are you quoting, to be clear? When was it sent?
Philip Cassidy: That email was sent on 20 July by Carol Thompson.
Q27 Chair: That email was sent by HMRC to you?
Philip Cassidy: Yes, to two people in our business.
Q28 Chair: That was giving you notice, then, of the—
Philip Cassidy: It said that there were all the terminations going out. It did not refer to the number of the terminations that were going out, or that it would be double what our joint plan was.
Sir Amyas Morse: I just want to check what the deal actually said. There is quite a lot of “shoulda, coulda” stuff going on. What was your obligation on whom? Are you saying that it was specific in the contract—that you agreed a range of volume that you would be ready to deal with, and that this was outside the range and you could produce evidence of that?
Philip Cassidy: There is a plan that was agreed with HMRC. We staffed to that plan. We received significantly more volume as a result, and we handled significantly more volume. On the point that I referred to earlier, in August we received more than 400%—
Sir Amyas Morse: But what you are saying is that, in contractual terms, the protocol for agreeing how much you will be resourced to handle—what does it exactly state, without going into great length?
Philip Cassidy: We agree the number of staff, and we staffed more than that.
Q29 Charlie Elphicke: Listening to what both of you have said, it seems to me that Concentrix were expecting a six-week staggered termination, as happened in 2015, whereas you, Mr Lodge, expected to revert to a week and it would all sort itself out. This seems to be a problem that goes to the heart of the issue. We have to remember how many people suffered from this situation. It is all very well for you both to pass it off as a lack of communication, but it was very important for this contract to be effectively managed and to make sure that a situation like this did not arise. What lessons do you draw from this, Mr Lodge, about contract management at HMRC?
Nick Lodge: First, you are absolutely right. Uppermost in our mind as we put this right was that claimants—people—were at the end of this. I think that clearly there were issues of communication. I do not think it is quite right that Concentrix were expecting six weeks, because that would have been clearly abnormal—
Q30 Chair: You say “clearly abnormal”. We have a real difference of opinion here. Mr Cassidy, what did you expect?
Philip Cassidy: We expected 21,000 cases over a period longer than a week.
Q31 Chair: Longer than a week—so, two weeks or three weeks?
Philip Cassidy: We just had it over a period longer than a week.
Q32 Chair: It’s quite vague, given that you are trying to deliver a service to a certain number of people.
Philip Cassidy: The main issue was 45,000 over the space of a week.
Nick Lodge: There are lessons about communication. As I have already said, there are lessons about testing contingency arrangements in the event that they need to be triggered. I think we have taken that as a key lesson, making sure that arrangements are robust to circumstances that might arise.
Q33 Charlie Elphicke: Mr Thompson, you’re running HMRC. What is your view of the future of private sector involvement in this area and outsourcings generally? Would you do another one, or do you think you would not do any more?
Jon Thompson: First, it is worth clarifying that this was not an outsourcing. It was additional capacity and capability, as the Report sets out. I have already given evidence to the Treasury Committee and the Work and Pensions Committee saying that I would not intend, under my watch as the chief executive, to involve the private sector in this kind of activity, particularly in a full payment-by-results method, which this contract was set up to deliver.
Q34 Charlie Elphicke: How could a payment-by-results model have been made to work for this contract?
Jon Thompson: On the face of it—I hate to do this, because you have been having a really interesting exchange between the people on my left and right, and there may well be a dispute between us about the volumes—it seems to me that the NAO Report is pretty definitive about why this didn’t work, as set out in paragraphs 3.16 and 3.17. For example, it fairly clearly sets out the definitive view about what went wrong, which is about resourcing customer contact centres, compliance planning and IT systems. Mr Cassidy may or may not dispute that.
For once, perhaps, I would point you directly to what the NAO Report actually says. It is backed up by various graphs. Figure 20 shows the expected number of people in the contact centre versus the number that actually were in the contact centre.
Referring to figure 20 on page 46, it does not appear that Concentrix put the number of people on the telephone that was in the plan that we agreed. To be fair to Mr Cassidy, we have not actually discussed that. If the NAO is to be believed—and one assumes that the Public Accounts Committee would indeed believe the NAO—then you have got some other evidence in this Report that you may wish to debate.
Could this be made to work? To answer your question: conceptually, I have my doubts. That is why I took the position that I did at both previous Select Committee hearings—in fact this is the fourth one we have given evidence to—that I would say in this particular area one of the lessons to be learned is whether a full payment-by-results model would work for this kind of work. That is set out in paragraph 3.50 right at the end.
Q35 Charlie Elphicke: Will HMRC use a private sector supplier to undertake tax credit error and fraud checks again?
Jon Thompson: I think I was fairly clear in answering the previous question that the answer to that is no.
Q36 Chair: You are in accord with the Prime Minister, you may be pleased to hear.
Jon Thompson: I have indeed agreed with the Prime Minister, which is always good for the career, I suppose.
Philip Cassidy: Madam Chair, could I make a point?
Q37 Chair: Yes, Mr Cassidy. I think you have some issues about that charge.
Philip Cassidy: On the paragraph that Mr Thompson referred to, the NAO also referred to the fact that there is a difference of opinion on the number of staff. The number of staff is very clear. In the plan, we referred to the actual number of staff that we would deliver and in all cases, on all days, we delivered more staff.
What is referred to—and it was a late change to the NAO Report at the very end and could not be clarified from a time perspective—was that the numbers that are referred to in the chart that Mr Thompson referred to only include talk time that is included within the headcount. It is taking total talk time from HMRC systems and dividing by the number of heads. It doesn’t take into account lunches, breaks and shrinkage, as we would call it.
Therefore, it is not comparing apples with apples. If you compare apples with apples, in August we said we were going to have 509 staff within the plan and we delivered 579 on average throughout that period.
Q38 Chair: I want to bring in Claire from the NAO, who conducted the investigation, to clarify, please.
Claire Rollo: We do understand Mr Cassidy’s concerns about this. The data that we have used are the data that were available and were used in contract monitoring. It does represent, as Mr Cassidy explains, the talk time: the actual time that staff were engaged in talking to claimants. It is on a slightly different basis from the way that the plan was prepared, which would have accounted for lunch breaks, logging on the office system, but they were the data that were available for us to use.
Philip Cassidy: In August we had 70 more people than we said we were going to in the plan.
Q39 Charlie Elphicke: Mr Thompson, with so many HMRC staff working on this contract, why do you think no one spotted a major problem sooner?
Jon Thompson: One of the lessons to be learned—and I said this when I was here in October and it remains a lesson to be learned that is sound—is about what happens in the organisation and how it accelerates through the organisation to decision makers such as Mr Lodge and myself. That is the third of the five lessons in paragraph 3.50, and I gave them to you in October.
We have been doing quite a lot of work since then. We have taken all the top 600 people in HMRC through a seminar led by Mr Lodge on what we learned from this process and the importance of escalating issues in contracts and in other areas of our own administration, so that we actually resolve them. It seems to me that, if you have got a problem, the key thing to do is to try to resolve it as fast as possible in the interests of taxpayers and other customers.
Q40 Kwasi Kwarteng: Clearly, something did go wrong, and I understand that effectively you are blaming each other for what went wrong; that is how it may appear to a layman. Specifically, however, I want to ask the HMRC representatives this: why do you think you got the balance wrong—or do you think you got the balance wrong—between focusing on savings and also thinking about the actual claimants, for whom this whole system has been set up in the first place?
Jon Thompson: That is exactly why I have my doubts about whether a full payment-by-results methodology could work, but also about whether the subtleties, if you like, of delivering a public service really work in a full payment-by-results model. So I doubt whether this could work in this particular area of our administration.
That’s exactly why I have that doubt—how do you balance those issues off? Because essentially what this contract does is absolutely incentivise my colleagues to do things which they otherwise might not do. That’s not to doubt in any way the integrity of Concentrix, but it provides an incentive framework that might not be a very appropriately balanced incentive framework. I think the combination there of two of those lessons is exactly the question you ask.
Q41 Kwasi Kwarteng: When did you come to that realisation? At what stage in this process did you suddenly—
Jon Thompson: To be completely frank with you, our first reaction was, “Okay. We’ve got a problem. Let’s sort that problem out”, so as to be really, really focused on resolving a situation for customers as fast as we could for those who were clearly in the system with Concentrix and the 181,000 people who had part-worked cases that we took back.
However, once we got through the bulk of that, then we sat down and thought, “Well, what do we think are the lessons to be learned from that?” And I think paragraph 3.50 is a direct repeat of the evidence that I gave on 26 October 2016 to this Committee. So it’s between it happening in early September and us getting in and sorting it, and our reflection in the middle of October.
Q42 Kwasi Kwarteng: So, just as a follow-up—this is to everyone—your understanding is that you reached this conclusion at the end of 2016. Is that right?
Jon Thompson: Yes, at some point in September or October.
Q43 Kwasi Kwarteng: And talking about 2015 now, as I understand it, the commission rate was increased from 3.9% to 11%, was it not? Yet at the same time, in October 2015, the company—Concentrix—was not meeting its targets. So what was the rationale for increasing the commission?
Jon Thompson: Let’s just go back a bit. We need additional capacity and capability to try to reduce the fraud and error rate in the tax credits system. So we’ve made a lot of headway from a fraud and error rate of more than 9% down to under 5%. We agree with Ministers that we need additional capacity and capability; we agree that the best way to do that is to source it from the market, and we go around a competitive process.
The year 2015 is the first year of that operation. There are clearly some lessons that were learned through the 2015 experience, which I personally was not part of but other colleagues were. It was apparent, I think—and you’ll have to ask Mr Cassidy about what the financial state of return was for Concentrix. That’s clearly one of the issues, and the Report says that Concentrix made a loss of £20 million in the end on that contract.
So clearly, there was some financial pressure, but we also wanted to renegotiate some of the performance measures. The trade that we made was that, in exchange for a higher commission rate, we wanted clearer planning and to change the performance regime, and the Report also sets that out, because it was the first year of operation and what we were striving for at that time between the two of us was to try to make it work. That’s a perfectly natural reaction.
Q44 Kwasi Kwarteng: But the intuitive response to that, and what seems bizarre looking at this, is that, at the very moment in 2015 when you knew they were failing to hit the targets, and at the point at which you renegotiated the commission, you increase the commission. That seems a kind of bizarre way of going about things. You could see that the targets were being missed.
Jon Thompson: It seems completely logical to me that—
Q45 Kwasi Kwarteng: You would pay more for failure?
Jon Thompson: No. It seems completely logical to me that, if there is a contract where the contractor is clearly struggling to make any kind of financial return—I wasn’t a party to it but I assume that’s the case, and you can definitely infer it from the Report—one of the things you have to do is to renegotiate that contract, to see whether you can make it work for an appropriate rate of return. Otherwise, and perfectly naturally, Mr Cassidy and his company may say, “Look, we can’t make this work financially for us.” That is a perfectly natural reaction. We are trying to make this work. We understand it is experimental. Just to respond to another point, we are not blaming each other. I am, in the end, accountable for this. We are all responsible for it. That responsibility lies between us in terms of different parts of the Report.
Q46 Kwasi Kwarteng: My understanding is that the savings realised were £193 million.
Jon Thompson: Yes.
Q47 Kwasi Kwarteng: And you anticipated £1 billion in savings, according to the Report.
Nick Lodge: Originally, yes.
Q48 Kwasi Kwarteng: So there’s a pretty big difference between the actual amount that was saved and the amount that you anticipated. Do you think that this whole process was worth it? Do you think it is something that, in hindsight, you would do again?
Jon Thompson: I’ve said on the record, so I’m not going to back away from it, that if you were the Chancellor of the Exchequer, the investment of £32 million of public money for a return of £193 million for the Exchequer would still look like value for money. I understand, though, that for customers this experience was completely the reverse. But if you are the Chancellor of the Exchequer—
Q49 Chair: Exactly. It rather highlights what you are measuring. As HMRC, I hope that you are measuring what you are delivering for your customers—in this case the tax credit claimants. Do you want to rethink what you have just said? You were very careful to say “from the point of view of the Chancellor of the Exchequer.”
Jon Thompson: Indeed I did say that. From the point of view of the Chancellor of the Exchequer, that looks like a reasonable rate of return. From the point of view of a customer who has been through this experience, it is terrible. It is a terrible experience.
Q50 Chair: Where do you fit, then? You are not the Chancellor of the Exchequer or a customer, Mr Thompson. Mr Kwarteng asked you what you thought.
Jon Thompson: Indeed he did. I thought I was being rather clever in giving you the point of view of the Chancellor of the Exchequer.
Kwasi Kwarteng: I wasn’t asking you what you thought the Chancellor of the Exchequer thought. I was asking you what you thought.
Q51 Chair: Between the money saving and the service to customers, where do you settle?
Jon Thompson: I think I was very clear in response to Mr Elphicke’s questions that I wouldn’t do it again.
Kwasi Kwarteng: Okay. That’s all I wanted to highlight.
Q52 Philip Boswell: May I clarify one thing, Mr Thompson? In something you mentioned about what you feel about contracts and what would be fair, you intimated that you felt that Mr Cassidy’s company wasn’t making profits, so you gave them some more work to enable them to make some more profits. But were you sure what those profits were? Does a company just have to hint that they are not making profits to get more work, even when they are failing to deliver?
Jon Thompson: The Report gives you the definitive position, does it not? I’m sorry to refer to it. Actually, I don’t see why I should apologise for it. I am going to refer to the NAO Report. The NAO Report says that they lost £20 million on this contract.
Q53 Philip Boswell: But did you know they’d lost £20 million at the time you made the decision?
Philip Cassidy: We did have open-book accounting, and it was very clear. My position would be slightly different. We were very clear on the financial losses. We were also clear in terms of some of the assumptions that were outlined to us from the pilot and from HMRC data. Those data were not actually proved to be correct—possibly because it was the pilot—but that was the reason that the financials were changed. It was not just because we said we were losing money, as you can imagine.
The other thing I would say is on the point about Concentrix being incentivised to do more. Mr Thompson talks about the fact that that would not happen internally. We received an email when we went through the 1.5 million records and reduced them to 324,000. From undeclared partners, we reduced claims from 1.1 million down to about 139,000 through deselection. The email that we got when we went through that was: “Given improvements in data analytics you think you have made, combined with your learnings from last year, I’d still expect a higher value. Certainly, if we”—that is HMRC—“had that caseload in our business, we’d expect a higher return than you’re forecasting.” To me, that is very clear that it is not to do with a private company versus a public organisation.
Kwasi Kwarteng: May I rewind? This touches on the line of investigation I want to pursue. Let’s wind the clock back to before the contract was given and you identified £1 billion as what you thought you could save by going down this route. If you had actually stuck to doing it in-house, as it were, and pursuing that line, how much of the £1 billion do you think you could have saved yourselves? Is that a clear question?
Jon Thompson: We cannot answer that question. It is not possible to answer that question.
Q54 Kwasi Kwarteng: You have no idea?
Jon Thompson: Just to rewind to the question that I gave you earlier, this is about trying to reduce the overall fraud and error rate in the tax credit system, in which you need additional capacity and capability. The option of us doing that through employing our own staff was not available to us at that time. Therefore, we went to the market.
Q55 Kwasi Kwarteng: But I am asking you a hypothetical question.
Jon Thompson: Yes, but I think they are quite dangerous.
Q56 Kwasi Kwarteng: I am asking you a hypothetical question: if that option had been open to you, how much do you think you would have saved?
Jon Thompson: It is possible for us to model, if we employ a certain number of people, what the likely fraud and error that we could detect is. I could not do that off the top of my head—
Chair: We are not expecting you to do that.
Jon Thompson: But we could attempt a written answer to your question.
Q57 Kwasi Kwarteng: I think it is quite nice to have a ballpark figure in terms of the savings—you have identified the savings and, in order to pursue the savings, you went through a third party, which at the time seemed a logical thing to do. But having identified the savings, what I am asking you is, to what extent could you have made those savings—how successful would you have been?—if you had relied on your own resources and your own staff? That is an entirely legitimate question as far as I am concerned.
Jon Thompson: It is a good question and we will try to give it an answer.
It is also worth saying that this was not just about providing people, but about bringing in some new skills. One of the things that we were looking for was to bring in different kinds of analytical skills and access to other third-party data, which actually, in my reading of it, was brought by Concentrix.
Q58 Kwasi Kwarteng: Do you think, going forward, that your institution will be able to make the kind of savings that you anticipated before you awarded the contract? Do you think that you can save more money working on your own, as it were, without the third party—Concentrix—being involved? Just looking at next year, for example.
Jon Thompson: The situation we are in now is that almost 250 colleagues have transferred from Concentrix to us, which gives us additional capacity to try and do fraud and error ourselves. The Office for Budget Responsibility estimate is that this event will have reduced the amount of fraud and error found in the system by £30 million.[1]
Chair: £13 million?
Jon Thompson: £30 million. This happening—
Q59 Chair: “This happening” sounds very like a sci-fi movie or something. We wonder a lot about the realms of fantasy in this place, but which happening—the fact that Concentrix ran the contract for a bit or the fact that you took it over?
Jon Thompson: The fact that Concentrix ran it for a bit did deliver £193 million, but at current estimates that is about £30 million less than what the Office for Budget Responsibility considered. That has been factored into the last fiscal event.
Q60 Kwasi Kwarteng: Another bit that is puzzling is that you paid the company something like £6.9 million over and above the commission payments due the company, according to the Report.
Jon Thompson: Yes.
Q61 Kwasi Kwarteng: But at that point you knew that it was falling quite far short of the targets you had set.
Jon Thompson: Yes.
Q62 Kwasi Kwarteng: So—I am going back to my initial point—it seems odd that, at the moment when you realised that the company was not doing well and not meeting its targets, you gave it more money.
Jon Thompson: Are you asking me about the renegotiation in 2015 or the termination in 2016?
Q63 Kwasi Kwarteng: I am asking about the renegotiation and the increase in the commission in 2015.
Jon Thompson: I thought I had answered that question, which was that we are trying to make it work, and that one of the ways in which we were trying to make it work was to work together on planning, performance and reward.
Q64 Kwasi Kwarteng: But do you understand my difficulty? It seems very odd that, there you were, and you could see. This is just repeating what my colleague and I have said, and I want to get a bit more clarity, because I do not really understand the motivation.
Jon Thompson: Okay, so you are asking whether—
Chair: You are asking him to clarify his answer.
Q65 Kwasi Kwarteng: Yes. I am just wanting to get a bit more. We ran back the plot to September 2015 or whenever it was. You can see that the company is not doing well and that the targets are being missed, and yet you decide to incentivise them more because you think that somehow that will turn the performance around. That seems like an odd call to me.
Jon Thompson: Your alternative is, after one year, to terminate the contract. You are trying to explore what the options—
Q66 Kwasi Kwarteng: No, I am asking you what the alternatives were. I am asking you whether the decision you took was entirely logical, with hindsight.
Jon Thompson: Fine. I cannot answer that question, because I was not party to that decision. I don’t know if you can answer it, Nick.
Nick Lodge: Yes. As we have said already, it was not working financially for Concentrix. They were very clear with us about that. We had open-book discussions about what we should do. There would have been an option to terminate, I guess.
Q67 Kwasi Kwarteng: But you decided to give them more money.
Nick Lodge: Both organisations decided that we could make this work if we made some changes to the contract. The commission rate was increased, but it wasn’t the only change we made. There were a number of changes to the terms, because we clearly had similar performance on the customer service levels, and we wanted to make a raft of changes to better balance off those things about performance and customer service with a reasonable term for the supplier of a service. If your contractor is not making a return, that’s not much good for—
Q68 Kwasi Kwarteng: Just on this point, as a layman, in terms of common sense, it seems quite an indulgent thing to do. Let’s say I had a house and wanted someone to do some work on it, and halfway through the period, they had completely failed on the targets that I had set them, and that then they turned around to me and said, “If you give me more money, I’ll get it done,” and I said yes. That would seem slightly eccentric, to put it mildly. You can see the difficulty that we have with this issue. Does that make sense to you?
Nick Lodge: I understand the point you are making. However, what we wanted to do was both secure the savings for the Exchequer and improve levels of customer service. If you look at the levels of customer service—Mark Oatridge mentioned it earlier—immediately after the variation in October 2015, they did improve for a number of months. We were trying to balance those two things off on an open-book basis. We were not throwing lots of money in Concentrix’s direction so they made a huge profit. That is already clear from the NAO Report and from figures we have quoted.
Q69 Kwasi Kwarteng: To wrap up—this is really to Mr Cassidy and Mr Oatridge—the loss on the contract was £20.5 million?
Philip Cassidy: Yes.
Q70 Kwasi Kwarteng: At what point during the contract when you were making those losses—they were not just crystallised in one moment—did you think that this was not something that made much commercial sense to you?
Philip Cassidy: From the start, we had concerns, and we raised those concerns. Part of that is why the renegotiation was done in 2015. You talked about a house. My analogy to that is that you asked for a three-bedroom house and then asked us to deliver a 20-bedroom mansion. Ultimately, a lot more staff were required to deliver what was being delivered. That is the reason why the contract was renegotiated.
Q71 Kwasi Kwarteng: Just to wrap this up—obviously, our time is limited—Mr Thompson, you suggested that you had learned five lessons as a result of the experience. What are you actually doing as head of an organisation at the moment to ensure that those lessons are actually implemented? Intellectually, you have understood the lessons, but practically, what steps are you taking to embed them in future contracts, if such there be, or your organisation?
Jon Thompson: First of all, it is worth repeating that, as I said, Mr Lodge had a seminar for our top 600 leaders about escalation of issues. That is one of the lessons to be learned, and it is set out in paragraph 3.50. Secondly, we are going through a re-evaluation and a revamping of our overall commercial and contract management function in order to make it more appropriate to what we need. In relation to the delivery of the tax credits system, although you have not debated it here, we will be making some changes for the 2017 round, for example the famous letter, which is in the back of this. We will be improving and clarifying letters in order to—we have been experimenting with this—improve the response from claimants when they receive this kind of letter, and we will put in an online change of circumstances. So we are trying to change the tax credit system. We have been making some improvements there, but we also need to think about these lessons systemically for HMRC.
Q72 Kwasi Kwarteng: Just to cap it all more broadly—this is a question for Concentrix and for Mr Cassidy in particular—what do you think the lessons learnt are in relation to this kind of private sector delivery of what HMRC and what we are trying to do? Given your experience, what are your broad thoughts about the private sector’s capability to deliver this sort of service?
Philip Cassidy: I understand what happened with claimants and certainly we apologise for that, but we performed as well as the contract allowed us to perform. We do not believe the contract was designed in the right way to deliver what was required for Government.
On the escalation that Mr Thompson refers to, I am surprised that the escalation was not within HMRC. Given that I had board members in Asia and on the west coast of the States coming and asking me questions about it because they had seen it in the news, I find it hard to believe that senior people within HMRC did not have visibility of these issues at that time. From our perspective, I would say that the process review is significant, and one of the biggest issues is the burden of proof that is currently within the HRR cycle. That burden of proof is on the claimants. The process most definitely needs review because that causes a lot of pain for claimants and for whoever is servicing those.
The last thing I would say, which again surprises me and has not come up as a point, is that the only ability to add headcount at those periods—at that period in August—was really with trained people. It would have taken us six to eight weeks to get people up and running when that volume hit in August, and the only people that could have done that were in the HMRC environment. That didn’t come until after the crisis had expired, although that had been requested a number of times and the operational team on the ground had been requesting that. To me, they are the main lessons.
Q73 Kwasi Kwarteng: But you think that the private sector can be engaged—
Philip Cassidy: I believe so. We have done these sorts of contracts before very successfully. We are one of the top five business services companies. We are all around the world with 100,000 people. Ultimately we do this. This is our bread and butter. No matter what organisation would have been delivering this contract, they would have ended up in the same or a worse position.
Q74 Chair: Mr Cassidy, you are saying there were problems with the contract. I will go back to Mr Lodge about some of those points. Why did you agree to go so quietly at the end of it?
Philip Cassidy: In the end? We very much looked at whether, from our perspective, it was worth following up in some sort of legal wrangle or in focusing on how we just get on, put this behind us and focus on generating more value to our business and our shareholders. We took the view that it was not worth getting into a legal wrangle and to let’s try and get this resolved as amicably as possible.
Q75 Chair: Did any money change hands when you walked away? Did you pay HMRC or did they pay you?
Philip Cassidy: It’s all within the report. There were cases that we had partially worked and we got paid.
Q76 Chair: No, I am asking whether there was any particular payment one way or the other where you bought yourself out or you were bought out.
Philip Cassidy: It cost us money, but it is all within the report. It is not anything different from what is in the report.
Q77 Chair: I want to go back to Mr Lodge. You have been with HMRC all the way through this contract. There were three senior responsible owners. Am I right? We have heard issues about the contract. We haven’t gone into the issues around the contract’s start date, which was delayed. There were disagreements along that path. From the beginning—I think Mr Cassidy alluded to it—there were problems with the contract. Would you agree with that?
Nick Lodge: Yes, I would.
Q78 Chair: So there were problems then. The problem is a year in. Who is ultimately responsible for drawing up that contract, and why was it so badly drawn up that it went so wrong?
Nick Lodge: It was drawn up in accordance with our normal contract terms and conditions. We really got going—Concentrix really got going—in I guess November, December, but really ramping up in January.
Chair: January 2015, just to be clear.
Nick Lodge: That is right; that is when the sort of volume started to go through. Then, as has been discussed, both parties could see it was not working as envisaged. It was not generating the kind of revenues, or return, that Concentrix had expected or wanted; and it wasn’t producing the kind of performance levels that we had expected, either. That is why we went through the renegotiation in October ’15, to try to put it on a better footing, given the experience that both parties had had, and the increased commission rate, which we have discussed, and the different arrangements for performance management. Thereafter we did see an improvement.
Q79 Chair: Mr Thompson, you have talked about increasing the capacity within HMRC. You arrived—you are still relatively new; we are giving you credit for being a new boy, so far. How bad was it at HMRC that these mistakes in contract letting were made, and what would your vision be about how it is going to be improving?
Jon Thompson: If you are asking me—
Chair: Was it fit for purpose when you arrived, is really my simple question.
Jon Thompson: Clearly not.
Chair: Thank you for your honest answer.
Jon Thompson: If you are asking me about the commercial function, HMRC has struggled over a number of years to get the kind of commercial function it needs, and in the end we actually went to the market to get a commercial partner to run the commercial organisation.
Q80 Chair: But if you couldn’t do the contract with the commercial partner, you would be in a bit of a vicious circle of bad management.
Jon Thompson: That may be ironic, but that is the situation—that the commercial function of HMRC has not been at the level that we needed; it does not have the capacity and capability that we needed, and we have hired Proxima, a private sector specialist commercial organisation, to run that function for us.
Q81 Chair: So you acknowledge that there were problems at the HMRC end as well as the Concentrix end, in the management of the contract.
Jon Thompson: Yes, and as I think I said earlier in answer to Mr Kwarteng’s questions, in the end the buck stops with me, does it not? We were all party to this. There are failings on both parts, and in the end I would like to apologise to customers.
Q82 Chair: On the customer point, we had some discussions here about compensation received as a lump sum or as weekly payments. While the lump sum was important for some people, there is a consequence of receiving a large amount of money suddenly in your bank account when you are on a low income or on benefits, and that is that it has an impact on your other benefit entitlements, which is still causing havoc for some customers. Can you tell us what, if anything, you are planning to do to support those clients of yours who are now having problems with other parts of the benefits system? I know that is not your Department, but it is part of Government.
Jon Thompson: We gave an undertaking to the Work and Pensions Committee to look at this, because they said that it bumped into the housing benefit system. Mr Lodge has been looking at that.
Q83 Chair: Mr Lodge, any updates, please?
Nick Lodge: Yes, we have talked to DWP about how that process worked. What we have clarified—I have to confess I am not an expert on housing benefit—is that the payment of a lump sum should not, in fact, impact on housing benefit. We exchange information regularly with DWP and through DWP to local authorities. That flow of data on changes to tax credits goes through in an automated process, and if there are any lump sum payments made, they are flagged as lump sum payments, which should not then impact on somebody’s housing benefit. We have had those discussions and I think there are further discussions to have, to make absolutely certain that claimants who have been on the end of this do not suffer—that they get the amounts of money that they are due as a result of this. I am confident from what I have seen so far that that is the case.
Q84 Chair: Can I just say that we hope this doesn’t ever happen again? The failure was catastrophic for claimants. In all this discussion about who is responsible for what, and contracts, you mustn’t forget it turned their lives upside down, and as I have just alluded to, many people’s lives are still being complicated by the outfall of what went wrong last summer. I have to say it has been quite unedifying at times during this hearing that we are seeing arguments between HMRC and Concentrix about what went wrong. You are all highly professional people—or you should be; you are highly paid and experienced. Yet we saw a level of contract management—contract implementation—that fell far short of what the public or the taxpayer should expect. It has been hard to fathom the level of miscommunication between HMRC and Concentrix. My final question to you all is whether you have anything further to say to the customers who have really lost out and had their lives turned over as a result. Mr Lodge?
Nick Lodge: I would like to repeat the apologies that we have already made. I am—and the Department is—acutely conscious of the impact of this on claimants. We put things right as quickly as we could, and we must not let it happen again.
Q85 Chair: Mr Thompson? You have already apologised a bit already, but do you have anything final to say?
Jon Thompson: Let me repeat my apologies to customers for the third time. They should be at the heart of a benefits system, and the payment of a timely and accurate benefit should be the goal. That is not what happened here, so I give my apologies.
Chair: Mr Cassidy?
Philip Cassidy: We apologise unreservedly for what happened. Certainly, in many cases, we focused in on how we could get things resolved as quickly as possible. Part of what you—
Q86 Chair: I am asking if you’ve comments for customers; it didn’t get resolved at the time for them. Are you saying you are sorry?
Philip Cassidy: Most definitely.
Chair: Thank you very much, gentleman. We will now move on to our second panel. The uncorrected transcript will be up on our website in the next couple of days, and our report will be published in due course. We have a vote at 4.05 pm, so we will get cracking on our second session, but we will probably break just before the vote so that we can try to get to the front of the queue.