Public Accounts Committee

Oral evidence: Child Maintenance Scheme, HC 455-i

Wednesday 9 July 2014

Ordered by the House of Commons to be published on 9 July 2014

Watch the meeting: http://www.parliamentlive.tv/Main/Player.aspx?meetingId=15750

Margaret Hodge (Chair); Guto Bebb; Mr David Burrowes; Jackie Doyle-Price; Chris Heaton-Harris; Meg Hillier; Mr Stewart Jackson; Mrs Anne McGuire; Austin Mitchell; John Pugh; Nick Smith; Justin Tomlinson

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

Witnesses: Caroline Davey, Director of Policy, Advice and Communications, Gingerbread; James Pirrie, Resolution Child Support Committee, gave evidence. 

              Q1 Chair: Welcome to you both—and a particular welcome to Caroline, for stepping in at the last minute. Please send our best wishes to Janet Allbeson.

              You all know the purpose of this first short session. It is for you, from your experience on the front line, to feed into us what you think are the key issues that we need to address when we question the accounting officer.  James, would you like to start, given that you have had a little bit longer to prepare than Caroline has?

              James Pirrie: I suppose we should congratulate the teams on how much work has been done. A lot has been done successfully and there has been a lot of good communication. I suppose the purpose of today, though, is to think more about the bits that are not working so well, although perhaps without disappearing into too much detail and technical difficulties.

              It is still quite early for us. I have not yet been seeing the quantity of cases that would enable me to report with any deep level of profundity on where the straws in the wind are going, but I think we can spot some things that are not perhaps as happy as they might be. Can I just fire them off in any random order?

              It is unfortunate that there is not more of an effort being made to solve the historical debt cases. The way I am experiencing the cases at the moment is that there is still this endless, almost holy grail-like, search for the answer to how much debt is outstanding, as opposed to a focus on bringing the parents together, to assess how much father can pay—if one is going to stereotype for a second—and how much mother might accept.

              It is particularly galling for these parents and children, when what is happening—instead of payments and solutions—is that they are actually paying me to investigate, endlessly, how the formula might apply over decades. Brokering agreements would be lovely.

              No. 2, the system has rotated through 180 degrees. When I first saw this thing arriving over the horizon in 1991, it was going to suck in every single maintenance arrangement in the jurisdiction. We have now reached the point where it wants no maintenance arrangements at all, ideally; it would prefer—maintenance works perfectly adequately without any intervention from the CMS. However, we still have the old legal structure.

              All the amendments over the years have just been to change the 1991 Act, so we have a load of accretions that are doing exactly the opposite of what we want. For example, even though we are encouraging people to self-determine, they cannot self-determine by having a court order. They can step outside the CMS system for a year, but a year later either side can then make an application to the CMS and that will terminate the court order. These sophisticated financial—

              Chair: Sorry, I do not follow that. Explain that again.

              James Pirrie: My client will go to court and we will agree a court order, but a year later either parent can then make an application to the CMS and that automatically terminates the court order, without any by your leave. My point is that there is a tension between trying to encourage self-determination and a system which prevents it.

              Q2 Chair: What would you like changed there? The court order is almost a self-determination, in your definition.

              James Pirrie: For four years. It should not be disappearing after 12 months. I am seeing regularly—I know the service does not accept this—either mother or father making an application to court because they see financial advantage, and then the whole case trips into applications through the appeals process and variations, and so on, and the warfare moves into the service, at great expense. We will end up spending two days’ of tribunal time on those cases.

              It is similar with the family-based arrangement. We have been promoting an idea that people could create a careful, sophisticated agreement that would be legally binding. What we have courtesy of the service is really a Mexican stand-off arrangement, where we clarify what we agree we will pay, but either of us can walk away from this arrangement at any time. So what I think we will be experiencing more and more often is that receiving parents will simply see their payments dwindle down to nothing with promises that something will be made good at a later point, which does not happen. What we have seen from Gingerbread’s research is that, over time, more and more parents have to move into a regulated arrangement. I think we will see more tensions coming on to these families rather than less, and I worry that the hope for the family-based arrangement will be unfulfilled and that the formula will be the maximum and that people will negotiate downwards from that point, which is worrying because we will therefore see less money go to children.

              Q3 Austin Mitchell: What are family-based arrangements? Inherently, it sounds to me, as a mean Yorkshireman, that you will give more power to the men than to the women and that they are going to be more aggressive about their money.

              James Pirrie: Yes, it is a non-binding contract. We would like to see binding contracts.

              Q4 Austin Mitchell: Enforced how?

              James Pirrie: Not enforced. That is why I am saying that it is a Mexican arrangement. When somebody gets fed up, you can step back to the CMS. The arrangement is that this is what we will do until we decide to do something else, which is inherently dangerous.

              Caroline Davey: It is essentially the new terminology for private or voluntary arrangements. I think our position has always been that where parents can make their own arrangements between themselves—plenty of them do—they do that.

              Q5 Austin Mitchell: That implies an equal balance in the relationship, which does not exist in most cases.

              Caroline Davey: Where parents can make that arrangement amicably, that is great. You are right, however, to say that where parents have an unequal power relationship—inevitably, if one is paying and one is receiving payment, that is inherent—if they are being pushed towards a private arrangement, when it is not something that they can come to between themselves, that is where there is real concern about that power imbalance.

              Q6 Chair: Have you got one more?

              James Pirrie: Two more. I just want to highlight the dangers of the new system for what I accept is a small minority of the very rich. I am seeing multi-millionaires paying £7 a week and there will be more of that, because we have terminated an arrangement of lifestyle inconsistence. Under the old rules—CS2—as an applicant I can refer to the lifestyle of the paying parent and the tribunal will assess how much income you must have. That has gone now. So we are moving into a much more technical regime where it will just be looking at what last year’s tax return says and many people will be falling into the cracks.

              Q7 Chair: You’re seeing more of those?

              James Pirrie: Oh yes. If I can access sophisticated professional advice and if I can understand the system well enough, it is now much easier to move into the cracks. It is not the bulk of cases, but it will be difficult for a minority. Of course, you cannot use the courts to bolster it for the reasons that I have discussed.

              My last point would be on charging, but I can pass that ball sideways.

              Caroline Davey: Thank you for accepting me at short notice as substitute.

              It is always worth restating that our starting point at Gingerbread is one of the objectives of the child maintenance statutory scheme, which is very helpfully in the summary of the NAO Report: “One of the primary objectives of the…scheme is to ensure more parents pay the child maintenance they owe in full and on time.” Another objective is to “maximise the number of children benefiting from child maintenance arrangements”. I would hope that that is self-evident to the Committee, but it is worth stating.

              In terms of the NAO’s excellent Report on progress so far, we are pleased that phase 1 has been going pretty well and that the approach taken has been a phased implementation, and we recognise the positive comments made about the phasing. Some of the delaying in phase 1 has been entirely appropriate and has been the best way of managing risk. The IT has also been performing reliably so far. There is therefore some good news about how the new scheme has been introduced.

              The Report also highlights some warning lights. Of the two that are of particular concern to us, one is the significant question that has been already raised about family-based arrangements. A lower than expected proportion of parents are already saying that they intend to make private arrangements. We would come back to the point that we have made throughout this whole process that we think the Department has been quite unrealistic about how many parents will be able to make their own private arrangements that sustain. We think this concern raised quite early backs that up. We think there is much more to unpick around that.

              The Report highlights clearly that the real test for the new scheme comes with phase 2. It comes with charging and it comes with case closure. The NAO has raised concerns around the lack of contingency for parents not behaving as they might be expected to. There are delays in the data warehouse, which is a very significant IT piece of the jigsaw and was commissioned five years ago. There is still no clear date for when it will be implemented. We think there are some real risks in the implementation of phase 2. In some ways we would like to see much greater delay in the phasing of phase 2, as there has been of phase 1, in order to get it right for those parents as they move across. However, we recognise that that also carries some quite significant financial risk.

              Q8 Chair: How do you explain the cut of a third in the number of family-based arrangements?

              Caroline Davey: To be honest we are not massively surprised. Partly, we think, it is a slightly different population. If you are requiring everyone, as the Department now is, to go through the child maintenance options gateway, you are bringing in a much greater number of people to have that conversation. Previously, and it is instinctively logical, people who know full well that they cannot have that conversation with their child’s other parent or they have had a private arrangement and it has broken down, which is why they are applying to the statutory service in the first place, would have gone straight to the CSA. They would not have had to go through options. They would not have had to have the gateway conversation.

              Because those people are all now going through the gateway, there are an awful lot of people who may well be calling options and saying, “I’ve had a private arrangement for a few years.”—as James said, there is clear evidence that private arrangements break down over time—“It has now broken down. I now need some help.” Those people are being asked, “Would you consider a private, family-based arrangement?” That seems extraordinary. Similarly, those parents for whom a private arrangement was never likely, where there has been violence, hostility or unequal power arrangements, those people are saying, “We’ve split up. There is no way the ex will pay without support from the statutory service. I need to go through.” Again those people are being asked, “Have you considered a family-based arrangement?” That seems inappropriate, to be honest. That is our shorthand take on why that proportion has gone down. It is a point we have raised consistently throughout. We think the Department has been quite unrealistic.

              Q9 Chair: I saw some figures to show that 50% of the people who approach the statutory body are victims of domestic violence or abuse. Is that right?

              Caroline Davey: Yes. That was based on a DWP survey conducted of the CSA clients a couple of years ago. You are absolutely right; 50% said that they had experienced domestic violence at the hands of their ex-partner.

              Q10 Chair: So they won’t be charged, will they?

              Caroline Davey: They won’t be charged the application fee but they will be charged the collection fee of the 4% deduction if the new service has to step in to collect that money.

              Q11 Chair: Finally, the options gateway means you either have to ring up or go online?

              Caroline Davey: You’ve got to ring up.

              Q12 Chair: You can’t go online? Is that good, bad or indifferent?

              Caroline Davey: James and I talked earlier about what was being provided online for parents. It feels as if parents are being pushed very much towards the telephone.

              Q13 Chair: I hope it is not an 0845 number. We will ask the officials about that.

              Max Tse: It’s an 0800 number.

              Caroline Davey: It will switch to a free phone number within the next couple of months, as I understand it, following some fairly concerted lobbying. There is a lack of help and support online. Some of the leaflets that parents are being sent are only being sent out in hard copy. They are not available as PDFs online. We know that there is the promise of a self-service portal for parents but we don’t quite know what that will look like and what possibility there may be for interaction through that portal or when it will be delivered. Our understanding is that parents are being pushed very much to the telephone gateway as their first port of call.

              Q14 Meg Hillier: I represent a constituency where there are a lot of families like you described. The women—it is mostly women—have no contact with their partner for all sorts of reasons, including violence. Seeing the projection in figure 8 and figure 7 about the number of family arrangements going down, have either of you done any analysis, or can you point us to any analysis, of the impact on children of not getting money into the household? We are looking at the value for money of this. Part of that is looking at the money going into households in greatest need, and some of these women are central to that group.

              Caroline Davey: I can certainly point you to some research that we published last year in partnership with the National Centre for Social Research, which looked particularly at the population of single parents on benefits, so that lowest income group. I suppose our findings were the flipside of what you are saying.

              We found that in cases of single parents on benefits, where they did receive child maintenance, that lifted one in five of those families out of poverty. We are very clear that where child maintenance is paid across the board, but particularly to low income families, that can have a very significant impact on lifting those families out of poverty. The tragedy that the report also showed was the relative minority of cases where that was the case.

              Q15 Meg Hillier: Could both of you highlight some of the practical challenges? You have given some useful evidence about the practicalities of the system. In those situations—a big issue in my constituency—tracking down, usually the father, and trying to pin down his budget, is this system better or worse than the old system? None has ever been perfect. What are the barriers in place to getting an understanding of what dad, if it is dad, can pay and getting the money out of him, and what is the failure rate?

              Caroline Davey: You are right that it looks like there is much more efficiency in the system in terms of some of those administrative processes. I think we will have to wait until more cases go through to see if that is playing out in practice. One of the real positives is the use of HMRC data for employed non-resident parents, in terms of assessing income and doing that much quicker and more reliably. As James indicated, and might want to say more about, there is still an outstanding question where the paying parent is self-employed and establishing accurate assessments of income.

              Q16 Meg Hillier: Working off the books?

              Caroline Davey: We know that has been a long-standing issue in the current systems and I don’t think they have cracked it in the new system, about how you pin down self-employed parents in terms of getting accurate income data.

              James Pirrie: Perhaps the only other aspect is about the impact on behaviours of fees or fines, whatever you want to call them. The 20% levy is very high. I think one has different concerns about it in different scenarios. One that has worried me greatly has been about how one missed payment—and it may not have been missed, there may have been administrative mess-ups—has a 20% levy for six months before the person can escape the system again. There is not really an external system for assessing whether that was the right decision to have made.

              That is on one side. On the other side, I have been looking forward to the fees or fines system because I think it may transform compliance. I think we are going to move away this time from the “get away without paying for as long as you can” to a system where it will be as the service would like: pay on time, in full, all the time. We may see more of that because of the fear of the fine.

              Q17 Austin Mitchell: I am personally horrified by the prospect of charges. Was Gingerbread consulted on the principle and the level of charges? What do you think of the principle of charging? I think it is £20 for a consultation and then 4% and 20% for the collection and distribution. How does that compare with charges made by the courts under the old system? Gingerbread first.

              Caroline Davey: Gingerbread has been consistently opposed to charges from the beginning, as you might expect. We have been pleased to play some role in influencing the reduction of the charges from what was originally proposed. The original proposal was to have £100 application fee, which has been reduced to £20, and to have a 7% collection fee for parents with care, which has now been reduced to 4%. We still disagree with both of them on principle.

              The research I referred to that we published last year looking at single parents on benefit showed that for a significant proportion even a £20 application fee will be a barrier to entry. One must remember as well that we are currently in a very constrained economic environment—some may agree or disagree—where single parents have been disproportionately hit by many of the cuts and are struggling financially in a huge number of cases. For many of those parents, £20, which may sound not very much to a lot of people, is a huge amount of money to find and could put up a barrier to application.

              In terms of the collection fee, we have really struggled with the principle of it. In all cases, the paying parent will be offered the opportunity of direct pay. So once they are in—single parents pay the £20; they have had the assessment of the amount. The paying parent is then given the opportunity to pay voluntarily—they are all given that option. It is only after they have proven that they will not pay voluntarily—

              Austin Mitchell: The collection has never been particularly efficient, as far as I could see.

              Caroline Davey: Perhaps a separate issue. Only once someone has demonstrated that they will not pay voluntarily through direct pay will they be able to be moved into the collection service. At that point, we think a 4% deduction from essentially the child for their non-resident parent’s failure to pay is completely punitive and serves no purpose. One might argue, and we have joined with James in saying that the 20% collection fee for non-resident parents is quite a blunt tool and is quite a high amount and one must not count against there being the risk of a reaction against the single parent if they are faced with an additional 20% fee on top of the existing maintenance amount. However, at least there is logic to that: all of those parents would be offered the chance to pay voluntarily and if they do not pay voluntarily, they get the fine—that is in their gift, to a greater or lesser extent. For a single parent and a child, who are entirely dependent on the other parent to pay, having 4% taken from their payment feels very harsh.

              James Pirrie: Provided the administration gets it right every time, a 20% fine may be an effective way of ensuring compliance. I think my worry is about when that is got wrong and there is not a good system for sorting out whether it was a failure to pay. I think a fine is an effective way of ensuring compliance, but it is a pretty terrifying prospect: having got something wrong or having been misjudged, you then pay effectively 30% of your net income. For some families, that is quite a lot.

              Q18 Chair: Caroline, you say that Gingerbread has done research that shows that the introduction of a fee will dissuade people. Can you explain the basis of why you think 10,000 fewer people will apply, so we can understand that? If we understand and accept the basis, what is the impact on child poverty or benefit bills?  Have you looked at that?

              Caroline Davey: Yes. The starting point, and it is the caveat for the research that we have done and the research the Department has done about the impact of charging, has all been about intention, because charging was introduced last Monday. We don’t know until it has been introduced what, in reality, the impact will be, but the best estimate that we have been able to get and the Department has been able to get is to ask people, “If there was a fee of this amount, would you pay it? Would you apply? Would you think about it? Would you not pay it?” I can certainly provide the details of that research to the Committee. A significant minority of single parents on benefits said they would not be able to pay the fee and the Department’s research said something similar. The point in the Report I might draw your attention to, which I think is helpful, is figure 7, on page 21. That shows the Department’s estimates, based on their research, of how child maintenance arrangements would break down over time. If you look at the historical cases, 25% have no arrangement; phase 1 shows that 25% have no arrangement, but the longer term forecast for 2019 is that 31% will have no arrangement. I think the Department has made some estimates for the number of families that will drop out of the service—the number of current cases in the CSA that go through case closure who will not reapply to the new system, and as a result, the number of children who will lose out on maintenance. We think these are very worrying figures.

              To my knowledge, no one has done the next step of the analysis, which is the impact on child poverty, but I suspect it is not a good one.

              Q19 Chair: Or on benefits?

              Caroline Davey: Or on benefit received.

              Chair: We will ask the Department about that.

              Q20 Mrs McGuire: I wonder whether you have had any positive feedback from people who are using the new system? I appreciate Gingerbread’s position and it has been consistent all the way through, but have you as front-line practitioners seen any response? Do people think it is not bad, or slightly better, notwithstanding the criticisms that have been made?

Caroline Davey: To be honest, like James, we have not seen that many people who have been through the new service yet. I think it is a little bit too early. Slightly inevitably, if they have had a good experience with the CMS they are probably unlikely to call us. So the handful of people that we have spoken to, certainly on our helpline, who have been through the CMS have called us because they have experienced various glitches. It is not a number on which I would like to base any assessment of the new service.

              Q21 Mrs McGuire: Is the Department using your expertise in monitoring the new system? Again, you are the front-facers in engagement with those who might not find the new system acceptable. Is that expertise being used?

              Caroline Davey: Not on a systematic basis. We have regular discussions with the Department and they invite us in and ask us to share our views. We have those channels of communication open and we are using them.

              Q22 Austin Mitchell: Proceeding on my wife’s axiom that all men are bastards, do you anticipate that the number of parents who feel free to make family-based arrangements will increase, as the Department estimates, or whether it is likely to stay the same or even fall? It seems to place power in the hands of the men, but I won’t go on about that. Do you expect the number of family arrangements to increase?

              Caroline Davey: In order for that to be the case and in order for the number of effective family-based arrangements to increase, there would need to be more and better support for parents. That is an area where the Department talks a good game on additional support being provided to parents but in practice for those people calling the child maintenance options service there is very little, if anything, in addition that they are being offered. They are just being told, “You can apply to the new service; you can apply to the court; you can make your own arrangement. We think that is probably a good thing.”

              It is a script-based call centre and they do their best, I’m sure. For parents who are struggling to make their own arrangements but might get there with a bit of additional support—and James might be able to talk about some of the courses he has run that help parents get to that point—they need quite a serious intervention. That is what is missing from this process. There are some ongoing pilot projects, which will not be evaluated until the end of next year.

              For parents, certainly those facing case closure now who have been separated for some time, and even for new parents who are struggling, having a conversation with a call centre operator, however well meaning, is not going to make the difference in their being able to make effective family-based arrangements. The one thing that is important to remember is that if people do not apply to the new service, either on the basis of charges or their conversation, that does not necessarily mean that they are making a family-based arrangement. We think there is a high risk of drop-out altogether.

              James Pirrie: It has always been the difficulty of focusing only on maintenance. Maintenance is one seventh of the financial question and the financial question is a seventh of separation. It is very much the tail wagging the dog. It would be terrific if we could have a much broader discussion that is trying to encourage parenting. Gingerbread came up with a lovely idea where instead of moving into fines we would go to parenting courses. I have always found when I work as a mediator that if you can get people to come at the financial questions through the parenting ones it is much easier. Otherwise, it turns into a slog over the finances, and that is what I worry we will have too much of.

              Chair: Thank you so much both of you for enabling us to have that conversation. That was really clear and helpful.

Examination of Witnesses

Witnesses: Robert Devereux, Permanent Secretary, Department for Work and Pensions, Ian Wright, Change Director for Child Maintenance Group, DWP, and Susan Park, Senior Responsible Officer, Child Maintenance Scheme, DWP, gave evidence. 

              Q23 Chair: We have an empty chair at the moment, but I will start.

              I said before you came in, when we were having our pre-meeting, that it is a delight to have a report coming out of the DWP in which it looks as if things are moving in the right direction. So we will start by saying congratulations to you. No doubt we will focus on some of the issues as we go through, but it is really heartening for us to see again something with a shared objective across the political spectrum, and that here is a programme that is working well. I think probably part of that is that you have both been at it for a few years, and that is something that we often say in this Committee—that if we could only have consistent leadership and management of projects, we might be better at implementing the policy.

              Can you describe to us why you think it is going well? What are the ingredients? What can we learn across Government? And what might you learn within DWP as to why you’re doing better here than elsewhere in Government and certainly elsewhere in the DWP?

              Susan Park: I will start and then Ian, who is the programme director, will come in and talk about specifics. First and foremost, we read the history books.

              Chair: That’s a good idea.

              Susan Park: If Richard Bacon were here, we’d say that we took particular note of case study 1, which was what went wrong with the last two schemes. That’s the first thing we did. Then we talked to people who were around and they told us most emphatically what had gone wrong.

              The second thing we did was to take a step back and look at the ingredients that make for a successful programme, one of which you touched on: having a consistent set of senior people who actually are talking to each other. This is about policy people and operations people, who are very key to all of this in terms of turning policy into delivery.

              The third element is that you actually make sure you have a system that is deliverable, and you do that in a safe and managed way. Again, that is what we did. We switched it on very slowly; we took it very carefully; and we de-risked as much as we possibly could.

              So those are the things we learned from reading our history books, and Ian will tell you precisely what we have got as a result of that. Then, I am sure that the permanent secretary will want to add to that.

              Robert Devereux: I am sorry I am late. I was supposed to be here for 3 pm, so my apologies.

              Chair: You missed the praise. [Laughter.]

              Mrs McGuire: You can hear it tonight on “Today in Parliament”.

              Robert Devereux: I shall look forward to it. It was good of you to be quorate for it.

              Ian Wright: I agree with Susan. From my perspective, I think there are a couple of things that were key in terms of how we got to where we are. One of them was right at the start in the design. When the high-level design of this new scheme and system was done, it was done in conjunction with the policy people. One of the things that we learned from the past was that if you design a policy to a large extent in isolation, and then hand it over and try to implement it, that way lies problems. So, learning that got us off to a good start.

              I absolutely agree with Susan about the way we’ve implemented it. With big complex systems, my personal view is to imagine that you can test them in the lab, go live, and then take on all of the applications, all of the business from day one. That is a recipe for significant problems. All the way through this programme we have challenged ourselves and said, “Are we ready to go live with this particular element? If not, we are not going live. If yes, then let’s do it but let’s do it sensibly and let’s grow it.” Those two elements are certainly two big elements that have led to where we are now.

              Q24 Chair: The only thing that you did go live on earlier without testing, as I read in the Report, was the IT systems, and that led to some glitches.

              Ian Wright: When we increased the volumes in November, which moved us to taking all new applications into the system, we had some issues in terms of stability for two or three weeks. Having said that, we were still able to process applications. We never got big backlogs and so on. It is one of those judgment calls. One of the things you’ve got to assess at any time is how much testing you can do in the lab versus the need to go live and see what works. With some of the problems that we hit, one of the questions I put to our IT people was, “If we had tested for another two or three months would I have found this problem?” They said, “I can’t guarantee that.” So it is a judgment at any point in time. To flip it the other way, when we had our first major projects review, they said we were being too risk averse with testing and that we should compress it and overlap it more. So it is a judgment call. We think we have made more or less the right judgment calls to date.

              Robert Devereux: May I just add to that? This is quite important. I do think there is a sense abroad that you can test something, and that is obviously right and it is how we work. What we have done here, and are doing with other things, is recognise that there is a limit to what can be proved in a lab. In real life, real humans are different. There is no amount of testing you can do. So the right way to control that is to start with very small numbers. So when we started with 200 cases a month, with the entire management team clustered around the board when we were taking the first phone call, everybody could then get a grip of what was going on. If something goes wrong at a level of 200 cases, you can work it out yourself. So it is quite an important lesson. We should do all this testing as much as possible.

              Chair: Hear, hear.

              Robert Devereux: But testing only gets you so far and real life is real life.

              Chair: Can I just ask a question relating to my case load in my constituency? In a lot of cases where the system breaks down it is where the absent parent claims to be self-employed. So you have no record or the HMRC record is out of date or comes too late to assess it and obviously they are not on benefits. This new system does not capture that any better than the old system. I do not know whether my experience of that is disproportionate and whether colleagues around the table have a different experience.

              Q25 Chris Heaton-Harris: Can I phrase the question another way? This is one of the first times you have used real-time information from HMRC, isn’t it?

              Robert Devereux: We are not. Just to be clear, real-time information—the monthly flow of your earnings, my earnings into the Revenue, is being used for real in universal credit. The child maintenance system is working on the previous tax year’s total—

              Q26 Chair: So you will get the self-employed previous tax year?

              Robert Devereux: Correct. So let Susan answer the question.

              Susan Park: That is exactly the answer. We get the self-employed, the self-assessment from HMRC. So it is more efficient.

              Q27 Chair: Is that better than the previous system?

              Susan Park: Yes.

              Robert Devereux: Much better.

              Q28 Chair: One of our earlier witnesses—I think both Susan Park and Ian Wright heard this—said that that he accepted it was a small number but that under the system the very rich are getting away with paying less. Did you hear that?

              Susan Park: I did.

              Q29 Chair: And? You’re pulling a face: what does that mean?

              Susan Park: I think it is tiny numbers. Even the very rich have to make tax returns and so if somebody tells us that they believe there is some unearned income, then we can go and get those tax returns as well. So we can go and get the unearned income from—

              Q30 Chair: So do you not recognise the issue that was raised?

              Robert Devereux: We do. The last time we were here, we had a long conversation about someone with Woolworth shares, if you remember, with Richard. The position is that the system is automated to collect earned income from HMRC—employed and self-employed earned income—because, for the vast majority of people, that is all I need to automate.

              In the event that there is any suggestion that there is unearned income to be had, the Department will go to Revenue and find that. We have not built a system that scours everybody’s very complicated tax records for bits of unearned income, but we will go and find it. The rich person whose wealth lies in shares and the like would be captured by the system.

              Q31 Mrs McGuire: Could I deal, for a few moments, with figures 6, 7 and 8? I will take them in reverse order.

              To a certain extent, some of my questions on figures 7 and 8 are predicated on some of the analysis in figure 6. Unless I am misreading it, the family-based arrangements should rise by 6%. Is that right?

              Robert Devereux: Which figure are you on?

              Mrs McGuire: Figure 7.

              Robert Devereux: Yes. You have 25% family-based arrangements in the historic scheme, and 31% in the long term. Is that what you were referring to?

              Mrs McGuire: Yes. So you are anticipating a 6% rise in family-based arrangements?

              Robert Devereux: Correct.

              Q32 Mrs McGuire: How are you going to monitor that number? How are you going to assess whether you actually have a 6% increase?

              Robert Devereux: Our knowledge of what is going on in the wider world outside our own transactions is based on the surveys that we do of family lives. These are not monthly surveys, but we have them every other year, I think.

              Susan Park: Yes.

              Robert Devereux: So every other year, we are going out and doing a proper evaluation, and we are asking people, “What arrangements do you have?” The reason we know how many families there are who are not—

              Q33 Mrs McGuire: But it will be survey-based, and it won’t be a definitive figure—although you have indicated 228,000 in figure 6?

              Robert Devereux: It is survey-based, but an awful lot of things are survey-based. We cannot possibly go out and count everybody in this position.

              Q34 Mrs McGuire: Okey-doke. Could I take you to the other end of that bar chart? It shows an increase in no arrangement. Again, there is a 6% increase in no arrangement. Could you explain why there would be that increase under a new system?

              Robert Devereux: The only trouble with this picture—we had a conversation with the NAO about it—is that you ought to read what, in my eyes, is the red bar—the no arrangement—alongside the nil assessed and the nil compliant piece.

              What is going on in the move from the old world to the new world is, you can see a shrinking number—going from 17 down to eight—in the nil assessed and nil compliant. A lot of the people who are booked as having no arrangement in the top line are people who have no arrangement because they are nil assessed and nil compliant. In the new world, it is much easier for us to establish exactly who’s who.

              If you take those two numbers together, you will find that, across the course of this time series you have here, we are going from 42% who are either with no arrangement or nil compliant down to 39%. In that 3% improvement overall, you have more people benefiting.

              Q35 Mrs McGuire: So you think it is a representational issue in terms of the bar chart?

              Robert Devereux: Yes it is. You need to read the middle one—

              Q36 Chair: Does the NAO agree with that analysis?

              Max Tse: To the extent that nil arrangements could become no arrangements because they should not be in that category anyway, yes, but these are forecasts, and they are uncertain.

              Robert Devereux: The forecast is for 2019. The other two bars are actuals.

              Max Tse: Yes, sorry.

              Robert Devereux: I am telling you that the classification of the actuals is the one I have just described.

              Q37 Mrs McGuire: So are you agreeing with the interpretation by Mr Devereux?

              Max Tse: Yes. That is a possibility.

              Q38 Chair: So the interpretation of the two sections, you think they are both the same?

              Max Tse: Yes. It is possible that people will move through that from one category to the other, and that’s fine. I think the concern we would have generally is just the uncertainty about where people will go—whether they will go to family-based arrangements or no arrangement—and how that will split. It is uncertain. To your point, it is about how we monitor that and make sure that we understand what is happening.

              Q39 Mrs McGuire: Figure 8 indicates that the number of parents choosing family-based arrangements is not yet increasing. Susan, do you have any explanation as to why that is happening?

              Susan Park: This is about intention. When people phone they state their intention. At the moment, I would say that it is way too soon to take a definitive view on whether that is a long-term or short-term trend. In the period that we are looking at, where the actual number does reduce, that was during the period when we first switched on the mandatory gateway.

              We said that instead of everybody who wanted to make an application to the state just going to the state, we now say you have to go to the gateway. So that increased our calls and for a period of time we actually allowed them some easements in terms of the conversation that they were having, so that we did not block and cause a spike. For the period we are looking at we allowed some easements in that conversation. That is what is possibly distorting it, as well as the fact that we put a lot of people through that process who were clear that they do not want a family-based arrangement.

              Q40 Mrs McGuire: They don’t want a family-based arrangement.

              Susan Park: Because they are clear. These are the people who would have automatically come to the state and now we are asking them to reconsider and whether there is any help and support we can provide to make a different decision. But a lot of those people will already be clear that they want a statutory scheme. It is a very different mix.

              Robert Devereux: If I could just add one thing. Although it is true that we gave the options to provide some easements, all things being equal, that meant that more people were coming through to the scheme than would otherwise have been the case. As soon as they come through to the scheme the first thing that our staff do is ask, “Have you thought about family-based arrangements? Do you understand charging?” As you can see from the earlier charts at figure 5, the best part of 11,000 applicants do not proceed with their application, once they have started talking to us. It is not as if the fact that they don’t turn up on this chart that you started with is indicative of what is going on in the real world.

              Q41 Mrs McGuire: Given that a significant amount of the success of the scheme from your point of view is going to be based on encouraging people to make the family-based arrangements—the nudge in terms of behaviour—you are not yet concerned about that reduction.

              Robert Devereux: You’re absolutely right. If those green lines stay where they are over the next 12 to 15 months then we will be concerned. Our view is that that is a temporary phenomenon based on the fact that to start with in those months we virtually doubled the number of calls going in, and to start with they were not managing them altogether. When they are managing them well, you would expect that green line to be going up.

              Chair: Amyas wants to come in.

              Sir Amyas Morse: We do not dispute that. We think the proof of the pudding will be if this continues—as you rightly say, if we are sitting there in a year’s time and looking at the same proportions, there would be deeper questions to be answered.

              Q42 Chris Heaton-Harris: As it is a day for business cases, I thought I would raise one about the business case, looking at paragraphs 1.14 and 1.15. Essentially, the business case estimated the net present value of the programme to be £835 million in July 2013 and costs have gone up. Total costs have increased by about £70 million since 2012. Can you reassure us that they will not increase any further in this programme?

              Robert Devereux: The figures you have quoted have come from the business case at the time, and include that £70 million. We have spent most of the money we are going to spend on this programme. It is up and running. A lot of the future costs in here are the costs of operating, so it is numbers of staff multiplied by how long it takes to do it, with some assumptions about efficiency in it. So it is not as if I have got 80% or 90% of capital spending still to go over which I am uncertain, but it is a business case that goes to 2022-23. I will not say that I absolutely know that it is right, but a business case that produces an £835 million NPV is quite a good one.

              Q43 Chris Heaton-Harris: Yes. As we said at the beginning, it is a good Report. We are just picking holes in it now—unfortunately, the holes are not that big to pick. How much of that £70 million will be IT? Since I have been on this Committee, IT issues have always been a big thing for your Department and the Government in general.

              Robert Devereux: If we can turn to figure 3, because there is one thing that I wanted to say about that. In arriving at the calculation of what is going on with the costs, one of the things we identified was that £30 million of that £70 million is actually reallocation of operating costs into the cost of the programme, so, strictly speaking, it should be a £40 million net change.

              That £40 million, though, is of the order of £60 million or £70 million because of things that we have done through extra testing: doing a pathfinder, doing it in two phases and the costs of phase 2. To bring that back to £40 million, that is offset by some of the savings you see lower down in the table.

              Q44 Chris Heaton-Harris: From where you started with the programme to where you are now, were you happy with your cost estimates? Was your testing of them rigorous enough?

              Robert Devereux: Over the period covered by figure 3, my answer is yes. It has changed because we consciously made changes to the programme for all the reasons that you are congratulating us on. That came with costs which we knew about and consciously engaged with. Paragraph 1.17 talks about what was going on in the previous years: clearly the control was not so good and the NAO records that. But over this period, broadly speaking, since Ian arrived and Noel Shanahan became the SRO, this has been in a pretty tight world.

              Q45 Chris Heaton-Harris: What is your assessment of the risks in closing old cases on the legacy systems—will many costs be attributed to that?

              Robert Devereux: There are a couple, really. First of all, as the National Audit Office recognises, some of this is uncertain. We have a set of policies, charges and intentions in place and we have made our best possible assessment of how many people will want to transition from the old world to a family-based arrangement as opposed to from the old world to the new scheme. Those numbers could be wrong: the more people that come our way, the more our costs will be—it is as simple as that. So there is a behavioural response that we do not know about.

              In terms of the costs, I come back to the answer I gave earlier. We are now through the large uncertainties around the capital cost of getting a big system running. We are into the “business as usual” category of how many people we have got and what is their efficiency, which, broadly speaking, I will assess myself as being more comfortable with.

              Q46 Chris Heaton-Harris: Can I ask a couple of questions about this exciting data warehouse? It was first contracted in 2009 and we had policy changes announced in 2011; will it be fit for the purposes of those policy changes?

              Ian Wright: Yes. Part of the reason for the delay is that there have been changes in terms of policy and we have had to raise some change requests through the database. But yes, I am confident that, when the database goes live, it will encompass those elements.

              Chris Heaton-Harris: Cool.

              Q47 Mr Burrowes: Can I follow up on the data warehouse? Obviously there are some behavioural uncertainties, but the planning for your cost savings expected the data warehouse to be up and running. I understand that it is not up and running and it will not be available to select cases for the first segment in July 2014 and that, therefore, you needed to amend legacy systems to adapt to that. Given that you can only apply arrangements to the first three segments, what about the highly complex cases that need to be dealt with as a legacy? There is a financial implication in not being able to deal with those at the earlier segmentation stage. Surely your figures need to be revised.

              Robert Devereux: The story on the data warehouse programme is that we have had to put in a lot of effort to get it to where it is. You are absolutely right that we are going to select the first segment that is going to transition using a different method of values in the data warehouse. We have no reason to believe, as we sit here now, that the data warehouse will not be fit for purpose for those complex cases by the time we come round to them because, as you see from the time frame, it is quite an extensive period over which we will be doing transitions. The early ones are relatively straightforward and our contingency arrangement will work perfectly well.

              Ian Wright: I agree. We have tried all the way through this, particularly on case closure, to have a twin-track approach to make sure we had contingencies in place if the warehouse was not in place. We know we have those complex cases at the end. One reason for putting them to the end was that, by the time we get there, we will have gone through our processes and have much more information about what is working and whether our assumptions were correct. We will then be able to use that information to work out the best way of closing those cases, but I am confident we will have the data warehouse there and available to do that.

              Q48 Mrs McGuire: What is your oldest legacy case that is sitting somewhere in this data warehouse?

              Susan Park: The first scheme started in 1993. I suppose you could have entered that scheme and just been born at the end of 2002—so, 2002 plus 20 years. That would be your oldest, if I have worked that through correctly.

              Q49 Chair: Are you sure the parents are all still alive, whom you want the money from?

              Susan Park: The parents? Yes.

              Q50 Mr Burrowes: If these complex cases remain open, what is the additional annual running cost if your assumption is not realised?

              Robert Devereux: The big cost saving that is baked into the arithmetic is that we will eventually turn off the old legacy systems. If we are unable to transition all the cases off the old legacy systems, we will have the huge cost of continuing to operate them. I don’t have to hand the running cost of the old system as it stands.

              Q51 Chair: Because you are taking this slowly—which we think is a good thing and we are with you on that—there is likely to be a cost implication because you will have to run the old system in parallel. The IT has gone up massively and the actual cost of that is around £390 million. But if you run it for longer, there will be a cost.

              Robert Devereux: We can have a conversation about whether it has gone up massively in a moment. In terms of us taking this slowly, that is baked into the business case. The thing you are crediting us for is not some additional cost on top of the business case; the business case reflects that.

              Q52 Mr Burrowes: If there is any drag or lateness in the data warehouse being available, have you factored in the financial implication of that?

              Robert Devereux: I don’t think we have.

              Ian Wright: In terms of the actual closure of legacy cases, we are confident we can use the contingency solution we are developing to do segments 1, 2 and 3, so it would only be when, in a number of years’ time, we got to the more difficult cases that we would have any issue. If the data warehouse is not ready by then, we’re in real trouble. In terms of managing case closure, I don’t believe there will be an impact. The data warehouse is currently scheduled to go live at the end of October, but if that extended, we would obviously have costs associated with the suppliers as part of that extension. They would not be huge costs, but that depends on the length of the delay.

              Max Tse: Paragraph 3.14 on page 41 of the Report states that “extending case closure by two years would increase net costs by £140 million.” It is a rough estimate but that would be increasing it by a couple of years.

              Ian Wright: I think that is on the assumption that we don’t have the contingency solution that allows you to close cases, because most of the costs of the data you are talking about will be on the basis of not moving those legacy cases. In terms of the contingency solution that we built, and are continuing to build, I am very confident that that will cope with segments 1 to 3, so I do not think that that is a risk.

              Robert Devereux: We would of course be happy to come back in two years to tell you that it is all going well. Last time we came, you told me it was not going to go well, and it has, so another invitation would be fantastic.

              Q53 Austin Mitchell: Having said that the family arrangements give too much power to the biggest earner—normally the man—a woman with a child who is not working is very vulnerable in that situation. Both sides have an equal power to force an assessment, but it is possible that the highest earner can make direct payments, and only if he misses a direct payment, which is outside your charging system, and she raises the need for sanctions is there any discipline to it. Does this system of family payments give too much power to the man?

              Robert Devereux: I think the answer is no. We are positively trying, for policy reasons, to get people—

              Austin Mitchell: But they are in a better position to bully a vulnerable woman.

              Robert Devereux: They are only in a better position to bully if the threat and execution of not paying direct payment led to getting away with it, but because of the way in which the system is constructed we are conscious of who is in direct payment. They come to us. We have established what the sum is. We are telling them every year how to revalidate that payment, based on the latest information. In the event that the receiving parent is not getting the cash, we expect that they would call us and we would immediately go straight back into enforcement arrangements. Direct pay is not a soft option. It is one where we know what is going on, and we are ready to go straight into enforcement if we need to.

              Susan Park: It is, but I think you were referring to family-based arrangements, which of course are outside of the scheme altogether.

              Robert Devereux: I thought you were talking about direct pay within the scheme. Which were you talking about?

              Austin Mitchell: I was talking about direct payments.

              Robert Devereux: Yes, within the scheme. So my answer is right.

              Q54 Austin Mitchell: But you are predicating some of the improvement on more people coming to family arrangements. He quoted the gateway opinions that the number might well go down. Aren’t you really saying that family payments will increase because of the charging system? That will be the stick that increases the number of family arrangements, rather than turning to child support.

              Robert Devereux: Just to rewind, we are trying to—

              Austin Mitchell: You are going to bully them into it, in other words.

              Robert Devereux: We are seeking to go down this course, because all the evidence says that that has a better outcome than coming via a referee. We are trying to encourage people and support them. That is why we offer them the helpline, or the other things going on. But we are reinforcing that support with some incentives. That is what the fees are.

              Q55 Austin Mitchell: Okay, but on the level of charges, 20% for collection is a very heavy charge. I mean, bailiffs don’t charge that much.

              Robert Devereux: So get on and pay directly.

              Austin Mitchell: You’re taking bread from the mouths of children.

              Robert Devereux: No, I’m not.

              Austin Mitchell: Yes, you are—by having such a high charge for collecting what is due.

              Robert Devereux: With respect, the paying parent who owes £100 will pay £120. We will keep £20 and the receiving parent will get £96.

              Austin Mitchell: That’s a big deterrent to paying.

              Robert Devereux: Therefore the paying parent might be wise to go away and—

              Austin Mitchell: Twenty per cent on collection and 4% on delivery is a big charge.

              Robert Devereux: It’s a charge that at the moment is a small contribution towards a heavily subsidised taxpayer-funded scheme, and we are trying to get the individuals in question to get on and sort themselves out without being subsidised by the taxpayer.

              Q56 Austin Mitchell: Well, how did you set that level of charge? Was it to raise money or was it just that you thought it would be a bearable charge?

              Robert Devereux: This business case would be even staggeringly better if I put in an economic charge for this operation. I am not seeking to recover the full costs of the child maintenance arrangements, so you the taxpayer are still paying for this, even with these charges. We sought to put them in at a level that we think is reasonable and politically defensible, and at a level that people would recognise as being a reasonable deterrent. At every stage there is the option to do this without charges. You have only to ring up and organise it directly and we will underpin it with our compliance, or sort it out without coming near us and there will be no charges.

              Q57 Austin Mitchell: Is it higher or lower than commercial bailiff charges?

              Robert Devereux: I have no idea.

              Austin Mitchell: You should have found out.

              Robert Devereux: We have extensively consulted on this. I cannot tell you how many people have given us advice.

              Q58 Chair: Can I ask about the cost-benefit? Chris drew our attention to something on page 15. He talked about the cost. Part of the business case is the benefit. You are assuming a fee income, which is from the charges, of nearly £900 million—paragraph 1.14 on page 15—right? If your assumptions are wrong about how many people will pay the fees, what does that do to the cost-benefit picture?

              Robert Devereux: It works both ways, really. In the event that the fees cause fewer people to come, we will get less fee income and we will have less work to do. If more people come, we will get more fee income and have more work to do. One of the charts that the NAO produced—

              Q59 Chair: But the only thing is: do you remember that one of the charts in the NAO Report shows that the marginal extra cost to you, of more or less, is tiny? This is quite a big sum in the whole calculation in the business case. The marginal cost is very small.

              Max Tse: Figure 17 on page 37 shows two lines. One line is what happens to the gross cost—the cost of administering everything—if the number of people increases. The yellow line underneath shows the net cost after fees are taken into account so from the Department’s point of view, volume changes are more or less cost-neutral.

              Q60 Chair: So you can’t play that argument. Of course we would like more people not to come to you and to pay for their children in a voluntary arrangement but I am just interested in the business case aspect here. In the cost-benefit you assume £900 million. That is a lot of money. How robust is that and what is the impact of that being wrong on your whole business case?

              Robert Devereux: Just to repeat what I said: I said that because if volumes go up, I will have more work to do and more fees. I believe I am right in looking at that draft and saying that the effect of fees dampens the effect on me because I will have more work to do and more fees. The blue line is a lot more horizontal than the red line.

              Chair: Which one are you looking at?

              Robert Devereux: I’m on figure 17, which Max has just been talking about. Coming back to your question—we have made the best projections that we can for a world that we have never previously been in and we could be wrong. You are absolutely right. In the event that fees are higher or lower, the effect of that is higher or lower. The chart tries to give some indication of the changes. To be very clear, in arriving at this, I believe that the NAO has cut in all the child maintenance shares of the entire Department for Work and Pensions so this is not just the marginal cost of the actual activity you would see in a child maintenance centre; it has a tiny share of my salary in. Some of these things are fixed, in some sense.

              Q61 Chair: I accept that. Focusing on the business case, I understand that we have not been there before so it is difficult to predict but let’s say it was half that number. Let’s say that it was £450 million, not £890 million.

              Robert Devereux: Let’s take 2018-19. In 2018-19 the business case has got fee income of £144 million. In the event that we had half as many people coming in to pay those fees, we would have fee income of about £70 million. There is a sort of £70 million delta. This is in a business overall that, by 2018-19, is going to be £4 billion or £5 billion. That sort of variation is relatively small for our organisation. The change in JSA—

              Q62 Chair: I understand but, in the business case for this particular issue, what does it do?

              Robert Devereux: I am trying to do the sum in my head. I cannot help but feel that the reason the graph is the way it is in Max’s chart is that it probably makes very little difference to the business case.

              Max Tse: It makes very little difference to the business case from the Department’s cost point of view.

              Chair: Cost.

              Max Tse: I guess if there are more people paying fees, the overall amount of fees is higher and that is borne by the parents. I guess it depends on whether we are counting the cost perspective, in which case it is relatively—

              Q63 Chris Heaton-Harris: I thought if the costs go down by that amount, something in society is working and we should all be a lot happier. Is that not right?

              Robert Devereux: The fact is that the net position, after fees, in the chart that Max is referring to is basically flat. I think we would therefore assume that the taxpayers’ business case is pretty stable in all circumstances.

              Q64 Chair: What I am getting at is the business case has costs in it and it has savings. If any of the savings are wrong, all I am trying to get is an assessment as to what impact that would have on the viability of the whole thing.

              Amyas Morse: I think they are persuaded by the fact that if the fees go down and there is less work to do, the costs go down. Therefore, they move sympathetically.

              Chair: You are sympathetic to that?

              Robert Devereux: The business case is essentially going to give you the same sort of return, just at a slightly lower cost and with slightly lower savings. Same return on the investment.

              Q65 John Pugh: In the more predictable world of IT procurement, you use Tata Consultancy, Capgemini, British Telecom and Hewlett Packard, but initially under a contract that was time and materials. Then you changed, according to the NAO, to a traditional programme management. Who advised you to use a time and materials approach first? Is it a reasonable assumption that most of the escalation in costs took place during that phase of the procurement?

              Ian Wright: I was not in the role when decisions were made to use time and materials. The way the contract was structured, in terms of the time and materials, was one that is used fairly regularly, not just in the public sector, but in the private sector. The attempt was time and materials, but there was an amount of money kept aside as well and the intent was that if the supplier did well and came in early and under cost, they would get a risk-reward type share. That was the intention, as I understand it, of the original contract. I could not say what advice the people who put that in place did or did not get, I’m afraid.

              Q66 John Pugh: But did most of the escalation of costs for IT take place during that phase?

              Ian Wright: I think the impact from that phase led to the subsequent movement of the programme and the costs that we then incurred. That phase was the detailed design phase. That detailed design phase took much longer than was originally anticipated, which meant that all of the other costs moved to the right. So, yes, I think—

              Q67 John Pugh: When you change, do any of the contractors you previously had drop out?

              Ian Wright: No, we used the same contractors, but for two of them, TCS and Capgemini, we moved from time and materials, because we had seen that it was not giving us the outcomes we needed in terms of hitting dates and time scales, to fixed price, which is, “This is what we will pay you as long as you deliver x, y and z.”

              Q68 John Pugh: And am I right in thinking the new kit you have got, or will end up with, will enable you to get real-time information about PAYE, tax returns and so on?

              Ian Wright: No. We are not using the real-time information; we are using the PAYE system, as we said earlier.

              Robert Devereux: This is simply—

              Q69 John Pugh: Can I just explain where I am coming from on this? It is obviously very good to have as much interoperability and interaction between HMRC and your good selves, but there is a sort of urban myth that people often switch from PAYE status to self-employed status to working for an agency to setting up limited companies and paying themselves in a variety of different ways that elude your comprehension. Does your system pick that up, and if so, how quickly?

              Robert Devereux: Yes. Because we are deliberately looking at the last four years’ tax-earned income, and we are not trying to follow people’s employment status month by month, we are basically saying that if you make an application now, July 2014, we will look back at the 2013-14 tax-earned income, and whatever you said—self-employed, employed or however it goes—that is the number we use. The advantage of that is that we can very quickly make the calculations. We are now assessing people in 12 days or so, where previously we were taking weeks or months.

              Q70 John Pugh: So it is an urban myth that people can evade you by calling themselves self-employed—not just the very rich, but people working on building sites and the like?

              Robert Devereux: It may be an urban myth that applied to the older schemes, but it is not an urban myth in respect of the current scheme.

              Q71 John Pugh: Okay; one last question. On figure 5 on page 18, I was interested in what seems to be very good performance with regard to arrears—paying parents who pay less than 90% of maintenance due. I was intrigued as to where you got your expected or target figures from—you clearly did better than the targets, in terms of getting the arrears down, if I am reading that correctly, but what was the 17,360 expected based on?

              Robert Devereux: You may be misreading that—

              Q72 John Pugh: Am I? It may be healthier?

              Robert Devereux: That is not a calculation of arrears. This is saying that we expected 17,000 parents to be paying less than 90% of the maintenance. In practice, it was 14,000.

              Q73 John Pugh: Yes; I understand that. What is that figure based on?

              Robert Devereux: The expectation would have been based on what we knew about the existing system.

              Q74 John Pugh: So it is based on historical data?

              Robert Devereux: Yes, with some assumptions about what was going to improve in the new world. We wanted direct payments to be a higher proportion. Because direct payment is a higher proportion and because it is underpinned by this compliance, we are assuming that is going through at 100%, so we were allowing for that in our expectations, but even with heightened expectations, we are actually beating them.

              John Pugh: So it is not a figure plucked out of thin air—there is empirical data behind it.

              Q75 Chair: I thought the interesting thing was that you were quite out in your expected figure for people using the phone, as against the actual figure. There is quite a gap. Have you got an explanation for that?

              Robert Devereux: This is back in the options service part. I am not sure—I was paying more attention to the lower numbers—I am not sure why we were quite so far out on that, but this is our estimate of how many people might go via this mandatory gateway, post-November ‘13.

              Q76 Chair: What you are left wondering is, are there are more people just thinking, “I’m not touching it,” and therefore not getting the money they should? You just don’t know.

              Robert Devereux: It that was the case, we would not be almost spot on two lines below, where it says that the number of people applying for statutory cases, at 55,600, is precisely 4% fewer than we thought. I suspect it is to do with our inability to know about things beyond our shores, rather than our inability to forecast things coming across our shores.

              Q77 Chair: Have either of you got an explanation for that?

              Susan Park: I think that is the best explanation.

              Chair: Okay.

              Q78 Meg Hillier: First of all, as we have touched on legacy, I want to speak on one legacy issue first and then we can cover other questions.

              I have been contacted by a colleague who has a constituent whose son is now 21. I just wanted some general advice; this is typical of what you have in the legacy backlog and it does potentially pertain to the cost. The constituents are still being pursued by the CSA for an outstanding debt of £2,000 despite not having had statements, which is, I think, symptomatic of past problems. Although the parent has said the debt should stop, they have been told that they need to pay it to the state. This has been going on for over a decade. Is that typical? Are you really sure that the data warehouse system will solve those problems? It sounds like that is one of the really intractable ones.

              Robert Devereux: The thing about the new system, which is absolutely top class, is that every night the system balances to the nearest penny.

              Q79 Meg Hillier: But this is obviously an old one?

              Robert Devereux: Exactly, but in answer to your question about how this will go forward, when you are on the new system everything is balancing. The case you have read out sounds to me like one in which the parent with care was on benefits. At the time, the policy was that some of the money coming towards the parent was diverted to the Secretary of State, so that is probably why it is owed to the state and not to the parent, and it is taking so long probably partly because the last two systems have not worked very well.

              Q80 Meg Hillier: Okay, so what is the scale of that kind of problem, going back to what the Chair and Chris Heaton-Harris were asking about the cost of dealing with the backlog and the impact it has on implementing the new system? You cannot migrate that—the child is now 21—to the new system. That is not the plan, is it? If it has taken this long to resolve, despite the problems of the old one, you would think there must be some crack plan.

              Susan Park: That sounds as if it is a case that is closed because the child is over 20, but it still has arrears. Those types of cases will be kept right until the end. When we get right to the end—that is closed cases with outstanding arrears—then we will start to work those. As we work each tranche of cases that are closing, we go in and we have a look at the arrears. No arrears are automatically written off. There is a clear category in which you can write off. We have a conversation with the parent with care if we are still in the old system.

              Q81 Meg Hillier: So people who are caught up in the problems of the previous two systems could be waiting how long?

              Susan Park: We are going to close cases over the next three years.

              Q82 Meg Hillier: So it could be another three years.

              Susan Park: It just depends where they fall in that.

              Meg Hillier: It underlines why it is important to get systems right in the first place.

              Q83 Chris Heaton-Harris: In Meg’s case—you said that some of these cases will last three years—I am assuming that you would have some contact with these people in the interim. If you haven’t had contact in the interim, and all of a sudden you are getting a bill for something you thought had been tidied up, you are going to get a lot of cross people, even though they might well be owing the money. Could you give us an assurance on what you are doing to maintain contact, and trying to work on them in that interim period?

              Susan Park: Again, one of the distinct differences between the old and the new is that the new will have an automatic annual review. Every year, we will ensure that we update and bring those arrears and liabilities up to date. That is not a feature of the existing CSA schemes. I am afraid that, unless there is activity, we rely on clients to contact us if there are changes.

              However, given that we are going to close every single case between now and in three years, we will be contacting them and having those conversations. We are already doing lots of work in the meantime to look at what sort of client reaction we will have to these and how we will deal with that. We are doing all sorts of scenarios.

              Q84 Chair: Just before we move on, when do you, Mr Devereux, think that the NAO will be in a position not to qualify your accounts on this issue?

              Robert Devereux: It would be quite nice to imagine that the Auditor General will start to partition the world that is the old schemes, where he has been for a long time. I’d quite like him to engage with the question, “Could he give me an unqualified set of accounts for all the things that have happened since December 2012?” My guess is that he will not do that.

              Q85 Chair: I think he won’t. So when do you expect to be in a position where your accounts won’t be qualified on this particular, knotty issue?

              Robert Devereux: That was a good try.

              Sir Amyas Morse: No, I don’t think it was.

              Robert Devereux: If we have closed the other cases over the course of three years, we will have had to get to the bottom of this. As Susan said—

              Chair: Three years.

              Robert Devereux: All those debt-only cases are at the back of the queue. When we have tackled all those, I am sure that Amyas will give us an unqualified account. But on the way, I am sure that he will say nice things about how accurate the current system is.

              Sir Amyas Morse: If we get to a point that we don’t have to qualify in three years, I will be happy to buy this team a glass of champagne.

              Robert Devereux: Bring it on.

              Q86 Meg Hillier: I have two quick questions. Earlier, you described the situation in which you ensure that you look at each tax year in turn. Of course, people can file their tax return as late as 31 January the year following the end of the tax year. By the time it has gone through the tax office, it could conceivably be in April of that year. How do you deal with that time lag? It can be tough on the payer, who can be paying on historical data, or tough on the paid, either way, depending which way it goes.

              Susan Park: We use the latest available, and we go back six years. If the latest available is six years ago, that is what we will use. But we tell the individual what we have used, including the figure, so if they want to dispute that and tell us that there is a more up-to-date figure, they can do that. We might not use it, because it has to be a difference of 25%, but if it is more than that, we will use that figure.

              Q87 Meg Hillier: The other big question for me is about parents who are not in this system. I have many constituents who choose not to engage, for all sorts of reasons, with any system; and sometimes just with systems like this, which deal with difficult partners and money. They will choose to eke out their existence on benefits rather than pursue the errant father. Many of them sit in front of me at surgeries and, when I ask where the child’s father is, often the response is “I don’t know,” or “I don’t want to know and I have no desire to pursue it.” Those people will not show up in any of charts 6, 7 or 8. How do you know about them, and are you doing any work to assess the people who drop out completely? Although we are looking at the value for money of a system, in the end it is the children who should be getting support and who could be lifted out of difficult situations who matter, which is why we set up these systems in the first place.

              Susan Park: Actually, we are interested in the whole population of separated families, not just those families that touch the state. That is 2.5 million people. I think the Permanent Secretary talked about a combination of survey data. We look at the longitudinal Understanding Society survey, which tracks households—

              Q88 Meg Hillier: Sorry to interrupt, but it is quite important to know how you select those households.

              Susan Park: It is not our survey.

              Q89 Meg Hillier: Is it possible to get a sentence or two about that in a note? A lot of the people I deal with are what you might call “below the radar” for all sorts of reasons. They choose not to engage and they certainly wouldn’t answer a survey if it was put in front of them.

              Susan Park: Yes, well this is a longitudinal survey—a household survey that tells us as a society many things about households, not just the people living in them. We pool data for it, but it is not specifically for our purposes. We track it.

              Q90 Meg Hillier: It is going to be an imperfect system.

              Robert Devereux: Let us write you a note about how this assesses the inbuilt bias. It is a pretty classic survey problem. Some people are less prone to respond, which typically means that they oversample that category to adjust for it.

              Q91 Meg Hillier: A corollary would be the census where in areas like mine we always have a problem sorting it out afterwards because it is not quite real.

              Robert Devereux: As Susan said, it is a broader survey than just, “Tell us about your missing parent,” so it is possible that people are engaging with that because it is picking up other things. We can drop you a note.

              Q92 Meg Hillier: Just one follow-up, although it is perhaps more about process than policy. Do you have systems through government whereby if someone does interact with the system, be it through a health visitor, a GP or a school, they get advice about how to interact with this system? They might still not choose to do it but do they get that put in front of them?

              Susan Park: Actually, one of the programmes of work that we have been really focused on since 2008 is doing a lot more work with the sector out there that already deals with families. That includes people like Bounty, who interact with every mother; education; the Ministry of Justice; Gingerbread and Relate. We do a lot more work with those organisations about including training packages, information and education and with all of the existing bodies that deal with families. We do what we can to interact with all of these organisations. That was part and parcel of the ministerial steering group that was set up to try to better co-ordinate all of those services. It is certainly one of our objectives to do that.

              Q93 Chris Heaton-Harris: I have a couple of small questions. On page 20, paragraph 1.23 says that “Staff compliance varies geographically” on administrative processes. I could not work out why that would be, apart from perhaps the management of those geographic offices. I assume that you have done something to sort that out.

              Susan Park: Yes, this is a feature of bringing in the scheme gradually. As we build up the scheme, so we build up our people. I know exactly where this geographical area is. It is the last area where we have introduced our colleagues to working the scheme. So, actually, it is a feature of people gradually getting to grips with the new processes. I had a look at the underlying reasons and one of them is that they have not addressed all of the issues with the client that they should have. For example, one of our standard questions is, “Do you require information in another format?” If they haven’t covered that then it is something that gets dropped down in terms of the quality analysis that we do. As people build their experience, so that quality builds as well. That is exactly what we have tracked through the process.

              Q94 Chris Heaton-Harris: If you flip over the page to paragraph 1.28, reference is made to the operational changes. It says that the change in contractor in September 2013 reduced the number of staff with experience in helping callers overcome barriers to family-based arrangements. What do you see as an outcome of that? If people have struggled to talk to someone who knows what they are doing, could that be behind the number of people choosing either not to engage with the service or to do something on their own?

              Susan Park: It was actually the response that we provided earlier: we allowed some easements in the process to allow this new provider to get up to speed and take way more calls than we had originally forecast. It was during that period of easements where—

              Chris Heaton-Harris: So that is done and dusted now.

              Susan Park: Yes.

              Chris Heaton-Harris: I have one final thing. I want to say thank you, because when I started as a Member of Parliament, Child Support Agency cases were a huge chunk of my casework. We checked before I came here today and I haven’t had one for six months. I view lots of things through my mailbag and something is working.

              Chair: Good.

              Q95 Mrs McGuire: After that I feel that I am a bit of an anticlimax. I think that you have processed about 56,000 cases and you have all of these cases in a warehouse—I cannot imagine what this warehouse looks like, but never mind. What proportion of your staff are working on current cases and what proportion on legacy cases, given that there are 10 times as many legacy cases?

              Susan Park: Again, we are balancing it so that as our legacy cases reduce—you will see that they are reducing substantially—we move our caseworkers from that, train them and put them on to the other cases. At the moment we have about 1,600 to 1,700 people working on the 2012 scheme. The remainder are working on—

              Mrs McGuire: What is the remainder?

              Susan Park: We have about 8,500 colleagues.

              Q96 Mrs McGuire: Right; thanks. I suppose that this question is for the permanent secretary: what lessons have you learnt? This has been a really nice session for you and it is not often that you get one of these at the Public Accounts Committee. It is not perfect, but have a gold star.

              Chris Heaton-Harris: You go for it—

              Mrs McGuire: If anyone gets too cheeky I will take half a gold star back. What lessons have you learnt?

              Robert Devereux: The lessons are pretty much the ones that Ian has been through. I was going to bring Richard Bacon’s book with me, but I am glad that I did not because he is not here. Chapter 1 of his book is—

              Mrs McGuire: We haven’t read it—we are too busy reading NAO Reports. It is only people like you who read it.

              Robert Devereux: It is a good read. Obviously I have read it from cover to cover—

              Chair: He is speaking in a debate, to be fair to him.

              Robert Devereux: Every single one of the chapters at the front—the disasters—are characterised by one particular thing: things get turned on and rolled out nationally instantly. Everything—whether tax credits, individual learning accounts, you name it—has been done in the same way. What is different about this programme and all the other things we have done—it is the same with benefit caps and universal credit—is that we are staring with things that you can actually—

              Mrs McGuire: I think we were doing quite well up until that last sentence. Let’s stick to this programme.

              Robert Devereux: But it is sort of true—

              Chair: I wasn’t going to torture you with any questions on universal credit this afternoon at all.

              Robert Devereux: That is very kind, but it is still the case that what I have said is true. Given that we have established that you cannot test in a laboratory how an IT system works to absolute perfection, you have to start with small numbers. And when you know that it works with small numbers, you can give it to a few more people and you can unwrap it. So doing it progressively with decent testing is lesson 1.

              Lesson 2 is just being really confident that you do it at the point at which it is ready to go, not by some pre-announced date. In this programme, twice, we have sat down with the Secretary of State—he has been engaged in this programme weekly ever since we arrived—and gone through it and tried to make risk judgments. We have asked, “What do you think—should we go now, or is it better to do this?” As Ian said, there comes a point at which, actually, you could do another six months of testing and still not know anything, so eventually you have to jump. But the two big things are doing it in a progressive fashion and making the individual steps at the point at which you are confident, not because the date is pre-announced.

              Q97 Mrs McGuire: Although the processes are there—you have tested that, done all the IT and the things you can objectively assess—the thing you cannot assess is behaviour. One of the reasons there had to be a statutory system in 1991 was that, with the greatest expectations that people would come to arrangements, we know that family breakdown is not quite so straightforward. In 14 months’ time, we will know whether—even if your IT systems are perfect—the whole process will be the success you think it will be.

              Robert Devereux: That comes back to this business about doing proper longitudinal surveys of the whole population. Just looking at our data will tell us if it is working and we are efficient. The $64,000 question is whether we have created difficulties out in the wider world. We are consciously planning on doing, and we said we would come back with, an evaluation of the—

              Q98 Chair: Can you give us a time frame for that?

              Robert Devereux: We have committed in Parliament with the passing of the Welfare Reform Act to be back in 30 months after charging, which we decided was a long enough period to start understanding how behaviour works.

              Q99 Mrs McGuire: Some of that survey work also looks at the impact of having some of your other responsibilities, including the alleviation of child poverty. Will that be factored into your analysis if you don’t get the behavioural change?

              Robert Devereux: Yes—

              Q100 Mrs McGuire: So you can get value for money in this section of DWP, but if it impacts on another section of your work, you are not getting strategic value for money?

              Robert Devereux: In a lot of the things we do—pensioner poverty, child poverty, family breakdown—we necessarily have to survey the population at large to see what is going on. Ministers have committed to doing things in respect of all the three things I said, and they therefore need to know what is happening. In the event that we roll this process forward and have a perfect system but are not actually getting the right answers in society, Ministers will have to reconsider. That’s what they are there for.

              Q101 Chair: I am just going to put a slight dampener on this. If you look at figure 9, “Parents’ satisfaction”—the purpose is to try to get more parents to pay for the children. It is disappointing that 23% still don’t get a call back when promised and nearly 30% don’t feel their issues are dealt with in a reasonable period. The number of people kept informed is still four out of 10. It hasn’t shifted.

              Robert Devereux: It is also a table headed “October to December 2013”, so it was at a relatively early stage. I agree; we want to improve on these metrics. Remember that some of the people sampled are people we are trying to take money off who don’t willingly want to be part of the process, so their attitude to us is not wholly unusual.

              Q102 Chair: Of course, but things like ringing them back when they’re expecting to be rung back is an easier way of making sure you get the money.

              Robert Devereux: Correct. We have put in a system. One thing we had a conversation about many years ago was the general way in which the Department operates. We had a session with the Public Administration Committee about complaints. Lots of good things are happening in terms of improving how the Department performs. We have to improve in this area as well.

              Q103 Chair: When are you moving to a freephone number?

              Robert Devereux: I think I told you last time I was here that I have already done it. This is about 0800 versus 0300. When we had that hearing 12 months ago, I went away and we made some changes. Virtually all our benefit inquiry lines now offer both an 08 version and an 03 version.

              Chair: Are you telling us you are on freephone?

              Susan Park: Child Maintenance Options is a freephone 0800 number. All the others are 0845 or 0345.

              Robert Devereux: When we went through this previously, we established that 0845 is actually still cheaper for landlines, but 0345 is cheaper for mobiles. We now offer pairs of numbers, so the customer can choose.

              Q104 Austin Mitchell: We have the second highest number of legacy cases in my constituency after posh Southport and the highest arrears owed of any of our constituencies here. Why is that?

              Robert Devereux: I’m afraid I don’t know. It’s not deliberate.

              Q105 Austin Mitchell: Can you send me a note on it?

              Robert Devereux: I will try. Can I make one observation? The one person who is not here and is positively referred to here is Noel Shanahan, my chief operating officer. He has been the SRO for this and he was the chief executive of the Child Maintenance Commission before it moved back into the Department. He, Susan and Ian have done extraordinary amounts of work and I would like to record that he has been a driving force behind this. It would be nice to recognise that.

              Chair: We recognise that stability of leadership and personnel has really helped—a good story. Thank you.      

              Oral evidence: Child Maintenance Scheme, HC 455-i                            1