CORRECTED TRANSCRIPT OF ORAL EVIDENCE To be published as HC 1627-i

HOUSE OF COMMONS

ORAL EVIDENCE

TAKEN BEFORE THE

PUBLIC ACCOUNTS COMMITTEE

MEANS TESTING BENEFITS

WEDNESDAY 2 NOVEMBER 2011

ALISON GARNHAM, PROFESSOR JOHN HILLS and SALLY WEST

ROBERT DEVEREUX, STEVE LAMEY and EDWARD TROUP

Evidence heard in Public

Questions 1 - 114

USE OF THE TRANSCRIPT

1.    

This is a corrected transcript of evidence taken in public and reported to the House. The transcript has been placed on the internet on the authority of the Committee, and copies have been made available by the Vote Office for the use of Members and others.

2.

The transcript is an approved formal record of these proceedings. It will be printed in due course.

Oral Evidence

Taken before the Public Accounts Committee

on Wednesday 2 November 2011

Members present:

Margaret Hodge (Chair)

Mr Richard Bacon

Stephen Barclay

Jackie Doyle-Price

Matthew Hancock

Chris Heaton-Harris

Meg Hillier

Joseph Johnson

Fiona Mactaggart

Austin Mitchell

Nick Smith

Amyas Morse, Comptroller and Auditor General, National Audit Office, Michael Kell, Director, NAO, Gabrielle Cohen, Assistant Auditor General, NAO, Paula Diggle, Treasury Officer of Accounts, HM Treasury, and Marius Gallaher, Alternate Treasury Officer of Accounts, HM Treasury, were in attendance.

REPORT BY THE COMPTROLLER AND AUDITOR GENERAL

Means testing (HC 1464)

Examination of Witnesses

Witnesses: Alison Garnham, Chief Executive, Child Poverty Action Group, Professor John Hills, London School of Economics, and Sally West, Income and Poverty Strategy Adviser, Age UK, gave evidence.

Q1 Chair: Welcome to all three of you. We are grateful that you have agreed to give evidence this afternoon. For those of you who have not given evidence before, let me explain that we are different from other Select Committees. We normally talk only to accounting officers. We have introduced a new mechanism where we talk beforehand to experts where relevant, and we have decided you are the experts. We would like to inform our questioning of the three accounting officers, so we have a tight time frame.

I am going to start by addressing the three of you. This is a huge amount of money with an impact on a heck of a lot of people. If you were Chancellor or, dare I say, Prime Minister, what would you change? What is great about it? What is damaging? Looking to the future with all the changes coming-Universal Credit, changes to higher education and pensions, and so on-where are the biggest dangers for you? I will start with Alison and then go across the panel. It will be helpful if you keep your answers tight.

Alison Garnham: Overall, one of the biggest things we would change is probably the balance between means-tested benefits and non-means-tested benefits. Over the years, there has been a big swing towards more and more means-testing. Claimants are dealing increasingly with very complicated benefit systems, such as Tax Credits and, in future, Universal Credit, away from the simpler benefits, such as Employment and Support Allowance and Jobseeker’s Allowance, that are based on contributions.

Benefits such as Child Benefit are going to be income-tested for the first time. Child Benefit is in many ways an example par excellence of what works. It is a targeted benefit-targeted at people who have children-and it hits nearly 100% of its target, whereas means-tested benefits do not tend to. They are described as targeted and they are income-tested, but they are targeted in a different way. They are targeted on grounds of income, and for that reason they become very complicated and sow the seeds of their own problems, because they tend to miss their target more often than simple, straightforward benefits such as Child Benefit.

Q2 Chair: We are in a fiscally very tight situation. I come from an era when we always argued for universal benefits. That is not possible in the current climate, so which one would you prioritise? Child Benefit?

Alison Garnham: Yes, absolutely. That would be a top priority.

Q3 Chair: One thing I wondered when I read about it, was would we be better off having just a living wage, or raising the basic tax threshold? Would that be a more efficient and effective way to run it, rather than try to do all these means-tested benefits?

Alison Garnham: That is a very interesting question. More of the burden is being placed on means-tested benefits than on employers, for example. You could ask for more of the heavy lifting to be done by employers, through a higher national minimum wage, for example, and less through means-testing. That would be one option. We have also learned quite a lot through Tax Credits about a lighter-touch approach to means-testing. That has its problems, of course: we know about the problems of overpayment. We also know that it leaves people alone for longer periods of time, for example, and that you are not required to report changes so often.

Under Universal Credit, there will be monthly means-testing and reporting, depending on what kind of income you have to report, plus the introduction of capital rules. That is going to introduce much tighter means-testing. Arguably, although it helps people on the lower level to get the first step into work, it may be less good at helping people to progress in work.

Professor Hills: Your first question is a very big one, because you are covering such an enormous canvas. I would echo what Alison just said about thinking carefully about the overall balance. I spent several years as a member of the Pensions Commission. Our main concern, when thinking about how the pension system was working, was that at that point we were drifting towards a much greater extent of means-testing in old age. We put forward a series of arguments that suggested that that was undesirable in terms of the incentives to save that it set up. All parties accepted that that was the wrong direction to go.

I am concerned at the moment, with some of the directions of policy, that we are moving in the other direction, as an almost inevitable response to the combination of the desire to reduce the fiscal deficit without putting too much burden on taxation, but simultaneously trying to protect the poorest. The almost inevitable result is that all over Whitehall people are inventing new means tests. We have the Department for Work and Pensions producing its new Universal Credit proposals, designed to smooth out some of those means tests and to get rid of some of the highest peaks of means tests, while at the same time elsewhere new crenellations are growing.

Q4 Chair: Where?

Professor Hills: You asked what the biggest challenges were, and maybe we will come to that later, but I am worried about some of the moves that will create greater complexity in the system. The report that this hearing is based on had a useful way of setting out the different dimensions in designing means tests, which lead to either more or less complexity.

I am concerned that there is a new dimension that was important before 1972. Over the past few years, the idea of localised means tests has not been so important, but that is now coming back in. Maybe if we have a chance later on in the session, we can talk about how, under the localised reforms, each local authority is being asked to devise its own Council Tax Benefit system in future and how universities are inventing their own dramatic means tests under the new increased fees with bursaries system. That degree of complexity, which is happening at a decentralised level, will produce a system that will move in the opposite direction from some of the intentions of Government policy.

Q5 Chair: What is your observation from a value-for-money perspective, because we are a value-for-money Committee rather than a policy Committee?

Professor Hills: Okay. From a value-for-money perspective, setting up systems that undermine savings incentives and work incentives may put additional costs on the Department for Work and Pensions and reduced tax revenues in the hands of the Treasury. So, the value-for-money concern is that by trying to target spending very tightly, you are setting up systems that may lead to feelings of injustice that pop out elsewhere in the system.

On the complexity front, there is concern about the time periods and how tightly you tailor the system to people’s direct changes in circumstances from minute to minute, which might interact bizarrely with different parts of the system that change at different rates and times. Again, that is something that Universal Credit is trying to address by putting it all together so that everything changes at once within one system. I do not know whether that will happen and whether the ambitions of the Universal Credit system, through this real-time adjustment, will be achieved so that the system is right from month to month, precisely. That ambition seems to involve some rather heroic assumptions about how the administrative systems will work.

Q6 Chair: And if it does not, presumably fraud and error will go through the roof.

Professor Hills: That is one of the dangers.

Q7 Mr Bacon: You talked about a feeling of injustice popping out somewhere else. I am not clear how such feelings pop out anywhere, but can you be more explicit about what you mean?

Professor Hills: As economists, we tend to think about incentive and disincentive effects of heavily means-tested systems, when people are effectively gaining only 10p in the pound for every extra pound they earn. There is an equally important effect, which one might call the "prodigal son’s brother" effect. Why is someone who has behaved badly in some way treated very well in the system? You will find a lot of resentment from people who are just above the eligibility threshold in the means test, who wonder why they have lost out by comparison.

Q8 Mr Bacon: You are talking about incentives to fraud, basically.

Professor Hills: No, I am talking about feelings that someone did the right thing, but possibly regretted it afterwards. The more the means tests pile on top of each other and get closer to 100% rates-or even above that in future, looking at some of the systems that are coming in-the more people will say, for instance, "Oh, for goodness’ sake. If I’d realised that, I wouldn’t have taken those extra hours in the year that my child was going to university."

Sally West: We would like to see a world with a lot less means-testing of older people. Before I talk about some of the problems with means-testing, I should say that pensioner poverty figures have gone down considerably over the past decade, largely as a result of Pension Credit and more generous means-tested provision for older people. We would like to see a system where there is less means-testing because people have better state and private pension provision. We are positive about the proposals that the Government are putting forward for a higher, single-tier State Pension. We are certainly positive about the aims, which are reducing means-testing, encouraging saving and having a fairer system. We need to see more detail about exactly how that will work.

The problems that you have at the moment are about take-up. The Government have decided the levels of incomes that are acceptable for older people, and yet we know, for example, that about a third of pensioners are missing out on Pension Credit, which is meant to be a minimum income. Means-tested benefits are complicated and expensive to administer, and they are very difficult for people to understand. There are many barriers to people getting benefits and a lot of individual costs in the process of having to claim means-tested benefits as an older person, which was picked up in the NAO Report.

Q9 Chair: Before you came in, Jackie was saying that page 17 of the NAO Report sets out-it is an average, so it reflects the fact that take-up is not as good as it should be-that, on Pension Credit payments, pensioner households receive £476 on average, whereas the richest fifth of retired households receive £488. Reading that, I assumed that that was because of fuel allowances, television, etc. and all those benefits that are non-means-tested. Given that what you are about is presumably alleviating pensioner poverty, what does that say to you? Does it say that we ought to start means-testing more? Given the fiscal environment or in terms of value for money, what does it say to you? I am sorry to constrain your thinking a little.

Sally West: One thing that is worth thinking about is where these people would have been in the income distribution had they not received benefits such as Pension Credit. People do get higher incomes for good reasons, including, for example, as a result of being severely disabled. Sometimes, if you are just looking at who gets Pension Credit, it may well be that getting the Pension Credit has bumped them up the income distribution, so you have to be slightly careful about looking just at figures like this.

You have to have a balance between income provided through a means-tested system that is targeted at those with the least resources and support for people with slightly higher incomes. As John said, there is this sort of resentment, and we certainly get it a lot with older people who say, "It was not worth me saving. I am no better off," or, "I am worse off." People do not always fully understand. Often they are a bit better off, but maybe not a lot better off from saving. Okay, it is too late for them to make savings decisions, but they turn round to their family-I was talking to somebody a couple of weeks ago, and she said, "I am no better off. If I was on Pension Credit, I would have been better off than had I saved and got a private pension. I tell my family, ‘Don’t save. Just spend your money.’"

So you will get that message across, and I think a general perception about whether it is worth saving and whether it is worth working is sometimes as important as the precise detail of how much income you will get. It is important. Benefits such as winter fuel payments are sometimes criticised because they go to people who are better off, but they also go to the groups who say, "I don’t get anything, because I worked and saved and because benefits are means-tested." So I think you do have to have a balance in provision in retirement.

Q10 Chris Heaton-Harris: On the same point, in the summary on page 6 of the Report it says that "Working Tax Credit awards were £3,173 per claimant in 2009-10. Providing this amount universally across the working-age population would cost £122 billion, more than 16 times current expenditure." There is essentially a finite pot of money that you can only spread so thinly in so many ways. So I guess we are trying to work out what the best way is of getting the right amount of money to the right people. Does that mean that you need both means-testing and universal-type credits?

Sally West: It is a balance, because if everything was means-tested, you would have a lot of disincentives for saving and for working and you would have a lot of people who felt that they were not getting any support from the system. As far as older people go, we would say that the means-tested system should provide a minimum adequate support. I guess that the Pension Credit level is around about what is generally accepted to be poverty levels. You do not have a luxurious lifestyle on Pension Credit, but generally speaking it should be enough to just about get by. We would not want the means-testing system to be made any less generous, but there are incentives for saving and issues around people feeling that they have contributed-you probably hear it a lot-they have worked or provided care or made all sorts of contributions throughout their lives and they expect support back from the country. That is why, if you just looked at means-testing support, you would get a lot more resentment from people.

Q11 Chair: John and Alison, do you want to comment on that?

Professor Hills: Successive Governments have all decided to run systems on the basis of a combination of universally available benefits-thinking widely, that is the National Health Service, state education and pensions, as well as things that replace people’s earnings when they are out of work for different reasons-and they have always ended up with a combination of the two, precisely because of the balance that you talk about. The difficulty is that, if you are taking decisions on a one-off basis, for any particular programme you think only of how you can focus your programme and how you can protect the most important beneficiaries of your programme. One of the most important recommendations in the NAO Report is that somebody, somewhere should be responsible for monitoring how all those things stack up together and what all those different systems will add up to. As I have said, that is particularly important as we move towards more localisation and more decentralised responsibility for designing some of these things.

Q12 Chair: Spell that out a little bit. Give us an example.

Professor Hills: Let me give you two examples of that. There is a consultation document at the moment from the Department for Work and Pensions on Council Tax Benefit reform, where there is an ambition to achieve a 10% reduction in spending on Council Tax Benefit, which is to be achieved while protecting pensioners, so it will be a much larger reduction for non-pensioners. Rather than setting out how that will be achieved through a national system, such as through a change in the 20% taper rate on income for people who apply for Council Tax Benefit, or by saying that you are no longer entitled to 100% Council Tax Benefit if you have a very low income, that decision has been left to each local authority.

To achieve that saving, each local authority will have to do different things, as well as having higher priorities. One local authority might decide that everyone has to make a minimum contribution, as we had at the time of the poll tax, while another local authority might decide to avoid that and have a much steeper taper on income to achieve the saving. Those rules will be different between each local authority. They will be different across boundaries. I think that runs the danger of undermining the clarity and the clear system that Universal Credit is supposed to achieve.

Another example of that is what is happening with university fees at the moment. As many universities, including my own, increase their fees towards-in our case-or to £9,000, most are designing their own bursary systems to protect the poorest students. They are doing that mostly, as far as I can see from a quick look, using the same income thresholds and measures as the Government use for their bursary system, which is, essentially, pre-tax gross income. There is then a little adjustment of that, and then you reduce. People who have incomes above £18,000 or £20,000 or £25,000 get less, and there are various cut-offs in the system. As far as I can see, each university is devising its own systems of cut-offs.

The London School of Economics, my university, has a system that goes down in steps. In combination with the Government’s bursary system for low-income students it adds around 27% to the marginal tax rates facing people in an income range of £18,000 and £42,000 a year, where an awful lot of earners are. That 27% comes on top of tax and national insurance, which is currently 32% or so and, for some people, will come on top of the Child Tax Credit withdrawal if they have younger children. You are soon into a situation where you are adding to 72% from tax, national insurance and Tax Credit withdrawal, or whatever it becomes within Universal Credit-it will be a little while before Universal Credit comes in. You are adding on these amounts. Some other universities just have a cut off. If you are below £25,000, you get a bursary. If you are £25,001, you will get nothing.

I looked at the University of Oxford, which advertises its very generous system. In the first year, if someone moved the parental income from £16,000 a year to £25,001, with the loss of the reduced fees, the loss of the Oxford bursary and the loss of the state bursary, they would be facing a 92% marginal tax rate, to which you add income tax and national insurance, by which time you are well over 100%. You can possibly add in Tax Credit withdrawal for people with younger children.

Each university is making that decision for very good reasons. We are all encouraged to prevent the new system from discouraging poorer students from entering, but if each university is inventing its new system and each local authority is inventing its new system, where they interact can pile up in what may be extremely alarming ways. Part of the purpose of decentralisation is to have those separate decisions, but it then becomes extremely important that somebody centrally is looking to see what parameters that should all be done within. Maybe in the response to the Council Tax Benefit consultation paper, the DWP will lay down some guidelines and we will discover to what extent these things can be varied and what limits can be put on them. It may be that someone at the Department for Business, Innovation and Skills is looking at what the effect of all the bursary systems will be.

Q13 Chris Heaton-Harris: I have a minor issue. Student tuition fees are paid for by the student. When the debt goes up over a period of years, it doesn’t affect the parents. You are adding two different figures together.

Professor Hills: No. There are two different parts of the system, with respect. One part of the system is that, where you pay a fee and you draw down a maintenance loan, you-as a student in your later life when you become a graduate-will pay things back. But, simultaneously within the system, is an increasingly elaborate system-it will become much more elaborate next October-whereby students whose parents currently have a lower income will get reduced fees or bursaries that will help them with their living costs. We already have some of those, but they will become much more important. They depend on the current year’s parental income.

At the moment, the crucial divide is between whether you are between £18,000 and £42,000. It is not such a large amount of money that rests on that. In future systems, a great deal will rest on the parental income. The parents may think that that is absolutely fine, but you may then have the danger that, at the end of the year, they turn round and say, "Hold on. We are actually worse off, having earned £25,001 than if we had reported an income of £24,999.

Q14 Chris Heaton-Harris: The parents’ tax does not change in the slightest.

Joseph Johnson: As a family.

Professor Hills: Yes. I am assuming that the parents take account of the welfare of their children.

Q15 Meg Hillier: I want to ask about going online. We have talked a lot about policy issues, but I want to ask about the cost. You might not want to comment on the cost of going online-that is perhaps for the officials to talk about-but do you think there are any pitfalls or good points about going online for the client groups that you have an interest in? The other thing that may be linked with that is the idea of "Tell us once", which was going around in Whitehall. It is where you provide your data once to the Government and, within data protection rules, it can be shared across the relevant Whitehall Departments, so that you do not have to fill in a lot of 20 or 30-page forms. Do you think that there is any benefit to that, and where have you got to?

Alison Garnham: I think "Tell us once" is a very good project. It was a very good idea and, in lots of areas, Jobcentre Plus and local authorities have been working very closely together to make sure that it can happen. One of the problems is the one raised by John that Universal Credit integrates lots of benefits so you should have far fewer Departments to deal with, but leaving Council Tax Benefit outside means that you still have the same number of Departments to deal with as before. That is still a big problem.

I have seen some of the early online work on Universal Credit and it is looking very good. The problem is that a significant proportion of the population, particularly poorer families, don’t have access to online facilities or are unable to use them, for whatever reason. Plus, even when you are using an onscreen method, the experience for claimants is still the same-it is a bit like looking at a form. It is the information that you need to put on it that is difficult. You don’t necessarily know what it is you are being asked. When someone says "capital", what does it mean? You have to go off, read the notes and find out what that is supposed to mean. What are your earnings? Are you having to average this over a period? What other income, such as maintenance, do you have coming into the household? You have to think about those things before you can complete the form. So the experience for claimants is not necessarily improved enormously, because the job of means-testing is still very tricky for them.

Q16 Nick Smith: I have a deprived constituency in south-east Wales with high levels of poverty, so I was pleased to hear that pensioner poverty has been reduced.

You have painted a picture of more means-testing in recent years and you make a good case for reducing complexity. Having said that, a third of pensioners do not claim their credits, and universal benefits for heating seem to have made a big difference. Can you say which benefits, universal or targeted, make the most difference in reducing pensioner poverty?

Sally West: If you are specifically looking at pensioner poverty, they probably are Pension Credit and the linked benefits, which are Housing Benefit and Council Tax Benefit, because they are targeted at people on low incomes. As long as you claim those, that is clearly going to make a difference.

In the longer term, the biggest source of income for most older people is still the state pension, which is a contributory benefit. It is not means-tested, but it is not universal in the sense that it goes to everyone over a certain age. Even with more generous contributory rules, you’ve still got to have paid or cared for 30 years. That is still the biggest source of income and acts as something on which you can build up your private income.

But for people who are already pensioners and who haven’t been able to build up enough income to give them a moderate quality of life, it is the means-tested benefits, as long as people claim. There is a lot of work being done by the Government, local authorities and voluntary organisations. We all spend a lot of time trying to encourage the claiming of benefits, but there is still a big problem with not being able to get everyone on to their entitlements.

I know that you have talked about value for money-that is more the sort of policy-but it is a problem for society if people are living on incomes lower than they need to be, because it is a difficulty. People are more likely to be in fuel poverty and are less likely to be able to afford good-quality food and housing. There is a risk for health and a knock-on effect on health and care services. People living on incomes lower than they need to be are a societal concern.

Professor Hills: Could I add to that point? Pensioners in the deepest poverty tend to be the ones who do not claim means-tested benefits and whose income comes entirely from the basic state pension. For them, the thing that makes the most difference is the level of the basic state pension. There is quite a significant group that, for one reason or another, either because of complexity or stigma, do not want to be seen claiming what some people see as charity, if it has a means-tested label on it.

I want to add one thing to Meg Hillier’s point. One of the valuable things in the NAO Report was its analysis of what had been done in the design of recent reforms of means-tested benefits. If you look at figure 10 on page 23, there is an alarming column on the right hand side of that, which looks at whether, in the design of each of the reforms, the Department had looked at the claimant burden of the reforms. Returning to the point that you were making, that was taken into account only in the case of Tax Credits.

In terms of thinking about the worry of what happens if these systems go wrong- it can be very expensive if they go wrong; they would have to be corrected or you end up paying out more money than necessary to smooth out the wrinkles-failing to take account of things that are going to put a great burden on claimants is clearly a problem, and I think identifying that issue is very important.

Q17 Fiona Mactaggart: You gave us a very dramatic description of how localised means tests could create some unintended chaos in the system. What I would be interested to know is whether you actually think it is possible to establish some central mechanism, which can make sure that means tests-let us just take in central Government, and not worry about things organised by individual universities-do not end up with unintended consequences, like people deliberately lowering their income, and so on. Do we actually know enough to be able to do that?

Professor Hills: I think we know enough to be able to model who is likely to be affected by different overlapping means tests. Indeed, there is a whole industry of people within Government itself. There is a model within Government that I believe used to be called IGOTM, the inter-governmental tax benefit model-the Institute for Fiscal Studies has a model of this kind and the University of Essex has a model of this kind-which tries to look at the circumstances of different households and work out all the different elements that they might be entitled to. It is actually quite hard to add in the things like the bus passes or some of the insulation measures you might be entitled to; but, yes, it is possible, and it does seem to me something that people should keep an eye on.

It is in response to that kind of calculation that we are currently going through the Universal Credit reforms-because somebody has done the calculation of how Housing Benefit interacts, and so on. My worry is that you start by doing that with the things you know about at the moment, and you produce a system that does try to avoid that, at a centralised level. Meanwhile, in another part of the wood, these things are growing again, and undermining your attempts to keep hold of it.

Q18 Fiona Mactaggart: You in effect have been saying that there ought to be a gateway for any such changes, which deals with this issue.

Professor Hills: I am not quite sure exactly what mechanism within Whitehall would achieve that, but it would seem to me to be helpful, if at any point somebody is carrying out a reform that says, "We will now restrict who is entitled to this; we will now passport this particular form of assistance or help that used to be universal, using this benefit," if somebody somewhere says, "Hold on; we have already piled a very large amount of stuff on entitlement to Pension Credit or entitlement to Child Tax Credit at the full rate"-or whatever it is-"and this has now produced a big spike in potential marginal withdrawal rates." I do not know whether that is described as a gateway or a health check, or something, but I would have thought somebody should be doing that.

Q19 Chair: Who do you feel to be in charge of means-tested benefit in Government? Who do you think is in charge?

Professor Hills: The people who do most of it are sitting behind us.

Q20 Chair: And which of those is in charge? Where does accountability lie?

Alison Garnham: It is probably the Treasury, isn’t it?

Professor Hills: It has to be Treasury, because these things are happening not just within the Department for Work and Pensions, but elsewhere.

Chair: It is interesting, what they tell us.

Q21 Jackie Doyle-Price: I just want to come back to what Sally was saying, because the issue of the low uptake by pensioners is of concern to me. You alluded in your opening remarks to levels of pensioner poverty having declined in the last two decades. To what extent is that due to more generous benefit provisions; or to what extent is it due to the fact that, increasingly, pensioners have their own private provision?

Sally West: I think there will be various factors, but when Income Support got rebranded for pensioners as minimum income guarantee, and made more generous-and then with the introduction of Pension Credit, which also introduced the Savings Credit: that certainly is one of the impacts. I am not sure I have ever seen any detailed analysis about how the different factors inter-relate, but I would say that the slightly higher means-tested benefits are certainly one of the factors, but for people retiring at the moment, private incomes have slightly increased.

Q22 Jackie Doyle-Price: What I am alluding to here is that increasingly people have become a bit more savvy about saving, notwithstanding that there are incentives for them not to-particularly the lower down the income scale they are. We seem to be building a class of haves, who have got their own provision, and have nots. We should really be targeting the help at them. The Report tells us that two thirds of those entitled to additional support through the Pension Credit are not applying for it, for whatever reason. Do we know more about what is stopping that group of people from claiming what they are entitled to?

Sally West: There is a whole range of interrelated factors. One is knowledge not only that a benefit exists but that it applies to an individual in their circumstances. People know about Pension Credit, but they might not think that they qualify, perhaps because they have a small private pension. The whole process also puts people off, having to fill in forms and give information. It is also about attitudes, such as a feeling of pride or "I can manage", even if it’s by turning off the heating. I think it is a combination of a range of factors. There are a lot of barriers, and people have to be able to get over all those barriers in order to go through the process of making a claim. Often, they need support to do that.

Q23 Jackie Doyle-Price: All the things that you have just painted substantiate my prejudice that it is the most needy who are the most proud and too disengaged from the system to put in a claim. To what extent would we be better off scrapping the means test and increasing the flat rate for the poorest people?

Sally West: I suppose the proposals that the Government have put forward for a single-tier pension of about £140 would mean fewer people would be entitled to means-tested benefits, but of course, as proposed, it will only affect future pensioners. If this is taken forward, I think that more older people will feel it is very unfair-certainly, a lot of people have already contacted us-but if you have that basis, it should make it easier for people to have a decent income in retirement, because all the people who would otherwise have had a lower amount will get a higher amount from the state. Of course, some people will get less, but there will be a simpler basis on which people can save for retirement. I do not think that the current means-tested benefits are too generous, so I would not want a reduction in those.

Q24 Jackie Doyle-Price: Not too generous, but if they are missing out two thirds of the people entitled, they are not working.

Sally West: There is a third missing out.

Q25 Jackie Doyle-Price: But the report says that 67% are not getting it, doesn’t it?

Amyas Morse: It would be quite interesting to hear what the DWP has to say about that when they come forward. They did some research. You may have done more recent research, Mr Devereux, but in 2006, they did some research which we are aware of giving reasons for not claiming: first, wrong assumptions of ineligibility; secondly, fear of loss or reduction of other benefits. They did not seem to find complexity a major factor. That is the latest that we were able to find, anyway, in doing our work, but we can no doubt ask for information.

Sally West: There is quite a lot of research, and it comes up with a range of things. Not knowing that they are entitled links in with complexity. That is probably one reason why.

Q26 Chair: We need to move on to our next session. Briefly, will each of you-starting again with you, Alison-say something to help us? Remember that value for money is our perspective on this.

Alison Garnham: Someone asked a question earlier about paragraph 6 and whether you spread out Working Tax Credit across the population. My immediate thought was that it depends what your objectives are. If your objectives are, for example, to reduce child poverty, you might want to pay that instead through something like Child Benefit, because that will hit the target of 100% of children. We know that child poverty has been reduced through that method and through Child Tax Credit over the past few years.

I wanted also to make a final point about the unintended consequences of the fiscal constraints that you were talking about. One is that, for example, the taper rate decided on for Universal Credit is 65%, even though the original idea was that it would be about 55%. That means that it has reduced work incentives, for example, for second earners. As people’s earnings increase, it makes it more difficult for second earners to move into work. That is probably an unintended consequence, but it does create problems for people’s work incentives, and therefore for ideas about whether the scheme is cost-effective.

Also, in order to encourage people to progress in work, there will therefore be a lot of in-work conditionality. Once you earn enough, you’re left alone, but if you just get into a very low-paid job, you will still be required to improve your earnings. For very many claimants, that changes the claimant experience. Even though they are working, it does not mean that they leave state provision and are on their own; they are still being pursued in terms of conditionality to improve their wages and so on. There are unintended consequences of the fiscal decisions that are being made.

Professor Hills: One issue that you as a Committee will probably want to keep your eyes on is the way in which Universal Credit evolves in terms of its ambitions to get it right instantaneously through the real-time system. The idea is that the computers of everybody’s employer will exchange information with the central Government computer towards the end of the month and tie together the information coming from couples. It will then send a message back to the system as to precisely what level of Universal Credit to pay out that month. I can see how that would work for people who are employed by organisations that are quite large and that run everything on a computerised basis. I find it a little hard to see how it will work for people working with small employers with much smaller systems and much more variable working circumstances-perhaps paid weekly, on and off, not in school holidays or on a rather informal basis. That is the issue that would be keeping me widest awake at night if I was trying to design that system within the DWP at the moment.

Sally West: Following on from that, in any kind of policy development, it is about effective delivery, and testing things out with the staff who have to use it and with the claimants. It is also about looking at the co-ordination. There are pensioners who pay tax and get benefits. We have different Departments means-testing people in different ways. If Government were able to have much more of an overview of all the different ways in which they assess individuals, we may be able to co-ordinate that and make it simpler and cheaper to administer and a lot less effort for the individuals.

Chair: Thank you all very much indeed. That was very helpful.

Examination of Witnesses

Witnesses: Robert Devereux, Permanent Secretary, Department for Work and Pensions, Steve Lamey, Director General, Benefits and Credits, HMRC, and Edward Troup, Director General, Tax and Welfare, HM Treasury, gave evidence.

Q27 Chair: Welcome. We found this a very interesting report. I don’t know what you guys thought about it. It is an enormous amount of money that the Government give out on a means-tested basis. It is incredibly complex and what you do in one area can really impact in all sorts of ways-intended and otherwise-on another. The main thing is who is in charge. Who looks across government? Do you, Mr Troup? You are probably the one I will start with. Are you in charge? Are you looking at some of these issues from our expert witnesses? Is it Mr Devereux or is it Mr Lamey? Who is in charge? Do you guys have regular conversations about means-tested benefits or did you just meet because you knew that you were coming to give evidence to us lot?

Edward Troup: We meet regularly. We have a lot to discuss. As you say, this is a very big area. Overall, benefits amount to £193 billion and means-tested benefits to £87 billion. As your report points out, they are quite widespread.

Q28 Chair: As time is tight, I would really love you just to answer the question. We have read the report.

Edward Troup: As Nick Macpherson said to this Committee last week, the Treasury cannot be everywhere. It is a small Department and it cannot second-guess everything.

Q29 Chair: Who’s in charge?

Edward Troup: It is a matter not of who is in charge, but of making sure that we get the right outcomes across the piece. A combination of the big Departments that have to deliver these things and the Treasury has oversight of public spending and needs to make sure that all the big chunks of public spending and talking to each other give the overall right result. The idea, however, that the Treasury could be in charge of all means-tested benefits, so the Athlete grants withdrawal rate is linked to x, y and z councils and to the Council Tax benefit withdrawal rate, and that is then linked to Universal Credit, is just not feasible when we have big delivery challenges on our hands.

Q30 Chair: Well, it sounds to me as though nobody is in charge. Who is looking across government at some impacts that our expert witnesses talked about? For example, who is looking at what is happening in HE and the impact that could have on incentives to work, or what is happening in Council Tax benefit and incentives to work, and why things are being simplified? Somebody has to look at that, otherwise we are not getting value for money out of £87 billion of expenditure.

Robert Devereux: That is a reasonable question. I would like to come back to the Committee on one thing, because last time I came here, or two or three times ago, your introductory remark was, "All the reforms you are about to do are going to be a car crash, Mr Devereux". I think that was your phrase.

Q31 Chair: I think John Hills almost said that at the end, but never mind.

Robert Devereux: I don’t think he did actually, and I protested reasonably vigorously at the turn of phrase.

Q32 Chair: Please answer the questions, just because we are on a tight schedule and there is a lot to cover.

Robert Devereux: Okay, but the reason I say that is because we are in the middle of the biggest single reform of welfare benefits for a long time. You could have written this report perfectly easily two years ago and I would have had no answer on welfare reform, but compared with now, I have an extraordinarily large undertaking, which you are rightly anxious about, and I am spending a lot of time doing.

The best answer to your question is that as we go along delivering what the Government have promised to do in the welfare space, we are trying our very hardest to ensure that concomitant things going on elsewhere are dealt with as best as possible. However, let us not pretend for one minute that I am about to revisit and redesign the entire means-tested system in one go. It is not practicable. It would be perfectly practical, however, having built a very good system, which is what I intend to do in Universal Credit, to ask, "When that is all done and safely delivered, is there anything that we can do to tidy up something else?"

Q33 Chair: In the meantime, while you are doing that, it is littered-I am not trying to take a partisan view, because you can argue about it over the last 15 years and into the next 15, no doubt. You look at one set of benefits and deal with the biggest, as does Mr Lamey in HMRC. I can see that your big day job is obviously trying to deliver the universal benefit, but if there isn’t co-ordination and at the same time something is happening out there on Council Tax Benefit or on HE funding, unbeknown to you, we will end up in trouble.

Robert Devereux: It is not unbeknown to me, nor is it unbeknown to those Departments.

Q34 Chair: Who co-ordinates it? Whose job is it? Who do we hold to account?

Robert Devereux: I think you should hold to account the people who are spending the money, and I am very happy to acknowledge that my teams are working with all the Departments you just talked about, to help them design it. The reality is that most of the money is coming out of my Department and HMRC at the moment and in future, it will be my Department. Lots of people then use the existence of an income-related benefit payment as a trigger for some other means test. It is very convenient for them and they need to think, in a world where I have reformed Universal Credit, fundamentally, for very good work-incentive reasons, what that does to their benefit. I am happy to help with that but I do not think you can say, "Could I be in charge of everything at the same time?"

Chair: I bet BIS is not thinking that what happens in HE funding could act as a disincentive to up your hours at work, which is what the evidence shows, so it takes you over the threshold and you lose all your-

Chris Heaton-Harris: That was a bad example. It was not correct.

Q35 Stephen Barclay: We did look at the issue on youth offending-the cross-departmental impact. I thought Sir Bob Kerslake had a working group set up looking at cross-departmental impact, where policy decisions in one area have a cost impact on another. Are you three part of that working group and does it cover the issues we are addressing here?

Robert Devereux: Sir Bob is doing some work, along with the Institute for Government, on the general subject of collaborative policy making. That covers a broad church of anything that goes between different Departments, within which space there are things such as means testing.

Q36 Stephen Barclay: Does this fall under that? I wrote to Sir Bob asking for terms of reference and I was surprised that there weren’t any. I do not know whether they have been produced now, but I am trying to clarify whether this work falls within it and if so, whether you have had any discussions with Sir Bob as part of that working group?

Robert Devereux: I have had discussions with Sir Bob about collaborative policy making, in my role as head of the policy profession, and we have not specifically touched on means testing.

Q37 Fiona Mactaggart: There is, for example, an inter-ministerial group that looks at domestic violence policy in different Government Departments. Why isn’t there an inter-ministerial group that looks at means-testing policy in different Government Departments?

Robert Devereux: I don’t know.

Q38 Fiona Mactaggart: Who is responsible for answering that question?

Robert Devereux: The Prime Minister.

Fiona Mactaggart: Right. I’ll ask him.

Chair: I think that that is passing the buck.

Q39 Mr Bacon: I fully accept what you say, Mr Troup, that the Treasury cannot be everywhere, although I have to say that in the years that I have been on this Committee the question that has most often come to my mind is, "Why wasn’t the Treasury on this earlier?" However, I take your point-you are a very small Department. But you almost made it sound like unintended consequences were unavoidable in something of this scale across so many Departments. Is that what you meant?

Edward Troup: No, it’s not. I am just looking at the numbers. Something like 80% of means-tested benefits are with Robert and his Department, or they will be once Universal Credit is in place. The Departments here do talk to each other very, very carefully and very closely. There is a huge programme of change going on, as Robert has described. It is not that we are completely blind to unintended consequences, but we need to worry about the 80% first. There is obviously some question as to whether higher education grants should be included in this consideration or not, but Council Tax Benefit is. It is a matter of ensuring that the way local authorities design their means test takes account of how Universal Credit will work.

Q40 Chair: And who is responsible for that?

Edward Troup: That is principally for CLG. It is part of the CLG process. Robert can probably say more about how that is joining up, because it is not for the Treasury to do that. It is for the Department-in this case CLG, which has Council Tax Benefit responsibility-to ensure that it is fitting in with the big 80% means-tested spend-

Chair: Lots of buck-passing back to the Prime Minister. Amyas.

Q41 Amyas Morse: I wanted to ask something similar, but perhaps in a different way. I suspect that all that the Committee really wants to know is whether this is integrated decision making of some kind. In other words, it is not so much, "Is there a group, or is there a body, or who is in charge?", but, "Is there an exchange of information?" I find it quite interesting to look at the international comparisons and see that in different countries the benefits fall in very different places. Is there some integrated approach that says that we want our combined benefit pattern to fall in this way? Is that sort of discussion taking place? I do not mind where it is going on, but is it going on? That is not intended to be a points-scoring question; I am just interested to know how the process works.

Robert Devereux: From my experience, there is not a single body that is worrying about means-testing in the round, but there are pretty vigorous bilateral conversations among those people who want a means test in one area-it could be education-and those people who have their own means test, who traditionally they have relied upon. Okay?

In a world in which Universal Credit exists, it simply isn’t the case that everybody on Universal Credit will be out of work. So, the question about what to do with free school meals is a real question for the Department of Education, with whom we are having conversations at the moment, and more generally-as your report acknowledges-Ministers have specifically asked the Social Security Advisory Committee to look at passported benefits in the round, to see where that might take us.

So there is not a single body; there is no one person in charge of the whole of means-testing. And right at the moment, I am not volunteering to do that for you either, because I do need to deliver all the things that you have already asked me to deliver.

Q42 Matthew Hancock: You are having conversations, for instance, with DFE about the free schools measure. That is an understandable process conversation that you need, because of the move to Universal Credit. That is different from looking at the impact of means-testing and the incentive effect of means-testing when you add in free school meals, because free school meals, although they are not a cash benefit, are a benefit in kind and therefore have an impact. Their removal would have an impact on work incentives, just the same as the removal of Housing Benefit would have.

Robert Devereux: When I say that conversations are going on, they are actually about, "How would you do free school meals in such a way as not to damage the work incentives?"

Q43 Matthew Hancock: So they would include the impact?

Robert Devereux: Yes.

Q44 Matthew Hancock: When you are thinking about your Universal Credit, you are also thinking about the impact of other means-tested benefits outside of your departmental control?

Robert Devereux: Exactly. As the Committee has already observed, I have quite a big task simply to bring the whole of Tax Credits and the whole of Housing Benefit into one system. While I am doing that, I am trying to look at all the other people who will necessarily have to change their systems because my system is changing and I am trying to work with them on what options are open to them to continue to fulfil their own obligations, while at the same time not sending something at right angles into the world that I am building. That is just ordinary collaborative, cross-Government policy work-

Q45 Chair: Can I just make the point that we will obviously be monitoring Universal Credit really carefully, but we all know that if you stick to your current timetable, which I am sure you’ll tell me you will, it will be 2017 by the time that is in? We have a lot of time between now and then. You are going to start introducing it in 2013-14 and it will be in by 2017. That is your end-date. You are looking at a cut-off.

Robert Devereux: Yes, but let’s be clear, this is starting for all new claims shortly after October 2013, so-

Q46 Chair: Yes, and it will be in by 2017.

Robert Devereux: There is a very substantial managed migration.

Q47 Chair: Fine. I am not attacking the timetable at all. I am just saying that there is a long period between 2011 and 2017, when a lot of money is here and we ought to be ensuring an integrated approach that provides best value for money out of the expenditure, in terms of work incentives, fraud, the cost of administration and all those issues. It is just a bit worrying that I bet it is our Report that brought that conversation to bear.

Robert Devereux: I disagree. I have tried to give you evidence that for each area for which other Departments have a responsibility, which are linked to my existing benefits, we are already in conversation about what would need to change. You raise quite an interesting point. Let us take free school meals. The reality is that I will have two underpinning systems going between 2013 and 2017-the old system, which can still run exactly as it did previously, if the Department for Education wants to do it, and a new system that might be something different. There is a complexity in that, but I can assure you that you have not precipitated that thought.

Q48 Meg Hillier: Is there a Cabinet Committee that looks at this currently?

Robert Devereux: Is there a Cabinet Committee? No. It is the same answer I gave you previously. The Cabinet Committee closest to what is going on in the welfare space is the Committee on Social Justice, which my Secretary of State chairs.

Q49 Meg Hillier: Sorry, who?

Robert Devereux: My Secretary of State-Iain Duncan Smith.

Chair: Austin has been waiting patiently. Fiona said she had something on this, but then I will go to Austin, then Meg.

Q50 Fiona Mactaggart: What I do not understand is that all these conversations you describe, Mr Devereux, are bilateral. The point is that it is not just between one end and the other that these things impact on the lives of the human beings in the middle. It is a kind of spider’s web. Your Council Tax Benefit is administered by a local authority and paid for by DWP within the context of a policy made by DCLG-your free school meals, Universal Credit and Pension Credit-and all of that affects people’s behaviour. Something that is designed beautifully on a drawing board has impacts, not only one way on another party, but on the rest of it. It is a matter of administration. There has to be a space within Government where there is a kind of impact assessment that looks at all that together. You are telling me that there is no space in Government where there is an impact assessment that looks at that together.

Robert Devereux: I am telling you that in terms of the critical behaviour-that is, the propensity to work and work incentives-I am worrying about that because the consequences of all the other things will have a bearing on work incentives, but they are not my responsibility.

Q51 Chair: But Mr Troup is saying that there is not.

Robert Devereux: Just to be clear, that is not the same as saying, "I have a group worrying about the totality of all possible behavioural effects of all possible means testing".

Q52 Chair: I think that is fair enough, but, Mr Troup, answer Fiona’s question, because I think the buck will stop much more with the Treasury.

Edward Troup: I think we need to step back a bit and see what we are talking about, because not only is Robert’s Department prospectively responsible for something like 80%, or slightly more, of all means-tested benefits, but the benefits that are not means tested include things such as Legal Aid. Means testing the right to Legal Aid is very important, but it is quite hard to see the impact on work incentives, given the nature of the population of Legal Aid claimants. I am certainly not saying that co-ordination is not important, but, particularly given the pressure on resources across Government and particularly on the DWP delivering Universal Credit, we have to look at which of the other means-tested benefits are priorities to join up, and I completely accept that Council Tax Benefit is one of those. A lot of work is going into that. I suspect that Athlete grants, which I had not heard of until I read this paper, is something for which the degree of joining up is probably a lower priority. Rather than saying that we need everything to be joined up, I think we should look for priorities-where means testing can affect outcomes and welfare, which is about work incentives first and foremost. That is where Robert is at.

Q53 Austin Mitchell: I don’t want to pursue the issue of work incentives because it is so emotive and so much prejudice is brought into any argument on it, but I want to pursue Fiona’s point about central co-ordination. The report says on page 35 that there is some machinery of co-ordination. We have the Child Poverty Unit. The Social Security Advisory Committee is undertaking an independent review of passported benefits. But the report goes on to say at paragraph 4.14 that understanding of the impact of benefit design on the wider system of benefits is limited. It concludes that "in our discussions with departments we found that there was a lack of clarity about who should be responsible for coordinating between means tests. This constitutes a risk to value for money in the benefits system." Surely if you have a central body that is defining benefits, examining the interactions and imposing a common approach, that must be conducive to less confusion, less cheating and greater value for money. Surely we need it if we are to operate the system effectively.

Robert Devereux: I think we’re going back over the same ground. It would be possible to say to my Department, "Please will you be in charge of everything?" I just don’t think that that’s a remotely efficient way of doing it. My team’s expertise in some of the things to do with the bursary example-

Q54 Austin Mitchell: This is passing the buck. How can you define a system of Universal Credit without this kind of body at the centre?

Robert Devereux: Simply by looking at the numbers involved. Edward has now read it out twice that by far the largest proportion of means-tested benefits will be the responsibility of the DWP under Universal Credit. It is fundamentally important that that 80% is well designed, even in its own right. That’s Tax Credits, Housing Benefit and existing DWP benefits. I will be delighted to come back to this Committee in 2013 and say, "We’ve done all that," and I’ll be delighted for you then to say to me, "Could you move on and do student support as well?" That would be a perfectly good question-but not now.

Q55 Chair: I thought this was quite telling. I hadn’t thought about HE until then. If you have a child going to university-excepting the student in the end phase, which is, I think, the point that Chris is making-and you know that if you increase your hours at work, the impact on the family income will be that you lose all the bursaries and discounts that you have for your child going to university, that’s a fantastic disincentive. I can completely get this from your point of view. You have a massive task. The last thing you want to do is think about something else, which Mr Troup tells us is only at the 20% level, but that’s a pretty important 20% in some instances. A massive disincentive to work could be provided.

Robert Devereux: Let me make one observation on that particular story. You’ve characterised this, as does the figure 17 example, as a household question. For all practical purposes, we have regarded the person going to university as starting their adult life. We can change that if you want and regard everybody at university as still part of their parental household, and leave the parents with the obligation to pay grants and loans.

Q56 Chair: But the whole point is that their eligibility for a bursary etc. comes from the parental income. What you say about the student may be right, but their eligibility comes from the parental income, doesn’t it?

Robert Devereux: There is a philosophical question here, and I’m not trying to dodge it. We need to decide whether we think my children at university are still part of my problem financially or whether they are out there on their own. At the moment, the entire benefits system rests on the proposition that, at 18, you are somebody else. You can disagree with it, but nobody disagreed in the previous-

Q57 Chair: No, it’s not a question of disagreeing or agreeing. The eligibility for the bursary is based on the family income, not the young person’s income.

Robert Devereux: The eligibility for the bursary is based, in this example, on the existence of either a full or a partial maintenance grant.

Q58 Chair: Which is based on the parental income.

Robert Devereux: Yes, and-

Q59 Chair: However you describe the student, if you are a parent in the real world and you know that doing another 10 hours-worth of work takes you over the limit and you lose the bursary and the maintenance grant, that will impact on your incentive to work.

Robert Devereux: All I am observing is that you are picking up that one Department regards that person as having been in different households but another Department does not. That is a perfectly fair question since it is manifestly inconsistent, but it is quite a big deal to change it right now.

Q60 Meg Hillier: How is online registration of benefits going, and have you looked at the experience of the claimant?

Robert Devereux: I only caught the end of your session with the previous witnesses, who talked about how potentially difficult online claiming will be. I have a couple of observations. I was in Warrington on Wednesday or Thursday of last week, because they have just completed the second leap of Universal Credit IT development. In that process, as I tried to describe last time I was here, we have had real customers and real staff looking at those screens to see whether it actually works. I can tell you that there is no question that says, "What is your capital?" We’re way beyond that already. We’re not bothered about questions such as, "What are your earnings?" I am not building a system that requires people to declare earnings. Some of the things you are hearing are based on presumptions about what I am building. As I tried to describe to you last time, the way we are going about this is to ensure that the screens, about which they were complimentary, are being tested by real people and by my staff, who will actually use them right from day one-in fact a day two years before day one.

Q61 Meg Hillier: Have you done an analysis of what costs you will save when you go online? You have an ambitious figure for Jobseeker’s Allowance, for example, haven’t you?

Robert Devereux: There are two things. In the long-term, steady-state costings of the Universal Credit, we have assumed that we will be able to remove a substantial amount of the administrative costs, of the order of half a billion a year.

Q62 Meg Hillier: But not immediately, because you’ll have to cover the cost of 40%.

Robert Devereux: Not immediately, because it will be 2017 until I can shut the old systems off.

Q63 Meg Hillier: Exactly. From 2017, though, you’ll make that half a billion?

Robert Devereux: They will progressively accrue. Halfway through this I will have a quarter, or something. I’m trying to guess as I go along.

As far as Jobseeker’s Allowance is concerned, we had this conversation when I was here two, three or four times ago. I cannot remember which. We set ourselves a target of getting to 50% of jobseekers’ claims online. We have told you that there are two ways of measuring that. At the moment, the one that says what proportion of completed claims began life online is, from memory, hovering around 20% or 25%. As I think I told you last time, the variation across the country is quite substantial. There are offices in Wales, which are on my list of places to go next, that are already at 50%. A lot of it is to do with whether staff are confident in encouraging claimants to do this. Most of our research on our own population shows that people are both willing and able to do this. Again, that is slightly contrary to the last lady speaker.

Q64 Meg Hillier: Can you tell me how "tell us once" is going? If I registered for Jobseeker’s Allowance online, is that enough? How much more information would I need to give for other benefits if I went for means-testing?

Robert Devereux: "Tell us once" has, so far, started only in the relatively simple space of telling us that somebody has died. After one phone call we will make sure that the rest of the public service knows about it. We will make sure that local government knows about it and that the Pension Service knows about it. That is the size of "tell us once" at the moment.

Q65 Meg Hillier: "Tell us once" was an ambitious project when it started, but I wonder where it is going because there are potential savings.

Robert Devereux: Yes. We were contemplating whether we could do "tell us once" for all changes of circumstance. Right now, we’ve put that on the backburner simply because there are a lot of other things to do.

Q66 Meg Hillier: It’s on the backburner, so is it on the shelf for when some future Minister might come and dust it off, or is it something you want actively to work on when you have the time?

Robert Devereux: It means that I am not currently putting resources into doing that because there are too many other things to do.

Q67 Meg Hillier: But do you have an idea that you will do it in the future? Do you have a date in mind?

Robert Devereux: We have been discussing that. It is not self-evident that DWP should be in charge of change of circumstance across government, and it is not self-evident that it should be run by the public service. The conversation we’ve been having has been with the Cabinet Office to say, "Look, in the space in which you are trying to do public service reform, is this a piece of work that you would be able to take forward?"

Q68 Meg Hillier: There are potential money savings, would you agree?

Robert Devereux: There are potential money savers in there. I would just caution you that the one thing we have to be really clear about is who is warranting that the data are correct and with what consequence. It is very easy to imagine that if I tell some central person that I have changed my circumstance it gets broadcast around. If there are any errors in that, who loses? Is it the taxpayer? Is it the customer? Is it the third party?

Q69 Meg Hillier: Have you done a cost-benefit analysis of going local on some benefits across Government? I am not sure which of you would be the best to answer that. There is normally a tendency in Whitehall to centralise but in this case we are hearing about some decentralisation of benefit decisions. Have you done a cost-benefit analysis of that?

Robert Devereux: I think the answer to that is probably no. What we have done, not unreasonably, is to think that, if you are going to run a national benefits system where people expect there to be some commonality across the country, it will be a national system. We have not gone about that in a national way. With the Work programme, for example, we were deliberately given regional contracts because that variation is perfectly straightforward.

Q70 Chair: Have you done a cost-benefit analysis, Mr Troup, on the decentralisation of Council Tax Benefit?

Edward Troup: No. We expect that to be done within CLG.

Q71 Chair: Are you looking at it? You seem terribly laid back about the Treasury role. This is a heck of a lot of money. It is £87 billion and we look to you in the same way as Parliament looks to us to make sure that we are getting proper value for this and you just pass the buck back to the Departments, as indeed you do in other areas.

Edward Troup: I am not being laid back about the role, but the local government spending team and the benefits team in Treasury would want to make sure-I am sorry that I don’t have the facts, but I assume the work has been done-that that cost-benefit analysis has been done and the Treasury itself would not do that.

Chair: I would have thought that if you come to give evidence to a Committee that is looking at value for money for means-tested benefits you would have looked at whether there is value for money out of this proposition. I leave that hanging in the air.

Q72Nick Smith: A quick question, Mr Devereux, about something I think you just said which is the percentage of people on Jobseeker’s Allowance able to make their claims online. I understand from the NAO report that it is 80%. I thought you just said that it was 50%.

Robert Devereux: There are two different targets. There is a target to get to 50% by March 2012 and 80% by October 2013.

Q73 Nick Smith: Okay. Two different targets. Thanks for that. I want to return to the point that Mr Hills was making in his evidence just before you swapped seats. He said that the thing that would stop him sleeping at night was this business about the link between small and medium enterprises and real-time financial information. Could you tell us a bit more about that, how it is going and whether you thought his comment about officials sleeping at night was fair?

Robert Devereux: I might ask my HMRC colleague to say a bit more in a moment. This is fundamentally in the first instance a reform which HMRC is making because it is a better way to collect tax. It obviously comes with the burden that the SMEs have then got to engage in this process monthly, but remember the counterfactual is that it just stays as it is: stuff happens and at the end of the year lots and lots of corrections are made. So in terms of running the tax system in a modern, efficient fashion, it is not obvious to me that this is not the right course of direction. As far as I am concerned, we are working extremely closely with HMRC on this. My senior responsible owner of Universal Credit sits on the board for RTI and vice versa because we are mutually dependent on this. I think that is as much as I want to say on that.

Q74 Nick Smith: Mr Lamey, Mr Devereux just gave you what we call the hospital pass.

Steve Lamey: Yes, I am familiar with hospital passes. I thought it had gone rather well to date. RTI is dramatically different and will be an absolute next step forward both for benefits and for tax. I think John Hills was probably needing to do some numbers on that. I think 80% of the staff employed in the UK are employed by 20% of the largest companies, which all will be using the most modern pay-as-you-earn systems. So when we did the analysis, 94% of all employees in the country are paid electronically these days. So if we are looking at numbers, 6% of people are paid non-electronically in weekly cash. That needs to be verified from our team on RTI. But the numbers we have are that circa 94% of all staff are paid electronically today. Now 6% is still a large number. Some 20-odd million are employed in the UK, so 6% of that would be 1.5 million. RTI is designed for those employers to let us know that those people are being paid in cash and how much they are being paid. So the system will pick out those people who are being paid in cash. It is a different solution to the main employees who, as we all probably do, get paid through electronic packs or EDI. The system will pick that up. The self-employed tend to be an issue because they don’t actually go through the pay-as-you-earn system. There is work under way to look at how we deal with pay-as-you-earn customers.

We believe that RTI is a big step. It is going to transform the accuracy of pay-as-you-earn which, I suspect, has been a topic at this Committee on many occasions. It will support the roll-out of Universal Credit in a way that we have not had before. Tax Credits will benefit from RTI in its last year of existence, 2014. We are expecting hundreds of millions of benefit by actually having that accurate income every month, aggregated from various employments. There are lots of detail yet still to be worked through: how do we aggregate all those different employments that people have into a total income for the period that we decide, be it monthly or quarterly? All those policy decisions are actually being looked at as we speak.

I have no doubts that RTI will actually take us to the next step forward. It is on time, contrary to what The Daily Telegraph would have us all believe. We are very optimistic that it will both transform the tax system and, as Robert was saying, make a real benefit to Universal Credit-and Tax Credits in the last year of their life. We think that it is going well.

Q75 Nick Smith: Good answer. I accept it in principle. You are, shall we say, working for a garage or a plumber. In those small businesses, they do their accounts or get someone in once a week or once a month to do their accounts. How will that be applied to the system?

Steve Lamey: Those will be the 6% we talked about. If we keep the proportion at 94%, there will be 6% that that applies to. The system is being set up so that those employers will be able to submit. Most of those people will have somebody doing their payroll. They won’t just be doing it themselves. There will be a payroll clerk who will be able to submit to us what they have paid the staff in that week. As for the detail of exactly how that will be done, will it be done through a web page? Will it be done by free software that the payroll suppliers are looking at developing for that purpose?

Q76 Mr Bacon: You said that most of those people don’t do it themselves; they have a payroll clerk. Which businesses are you talking about that have a payroll clerk?

Steve Lamey: I am talking about the 6%. We have 94% that have electronic means. There will be 6% doing gross calculations of how much to pay somebody. You then have to net that off to actually give them the net amount, following all the tax rules and other rules that we have. Somebody will typically do that in an organisation. It could be the owner of the garage; it could be the plumber/owner, but somebody in the organisation will be taking somebody’s gross income, turning it into net income and providing the information.

With RTI, it is important to keep the 94% and the 6% in context. They will have to find a way, which we are working through now, with the payroll providers and with the SME community at large as to how best that can be done. It is known to be one of the challenges. We are in progress with that, and I suspect that, over time, it will be absolutely clear how we do it.

Q77 Jackie Doyle-Price: Value for money is the degree to which we have a system that doesn’t encourage perverse incentives. If we look at paragraph 2.9 of the Report, it tells us that the Institute for Fiscal Studies estimated that, in 2010, 13% of workers faced effects of the marginal tax of above 70% and a further 2.6% of above 90%. That is a very significant proportion of people for whom work doesn’t pay. With the move over to Universal Credit, I want to know whether you have done any calculations to give us an indication that those high negative incentives to work are being addressed through reforms.

Robert Devereux: I could quote from any one of a number of documents. In the White Paper, you will find on page 54 a table that simply shows how many millions of people face what levels of marginal deduction under the current system and what that will be under Universal Credit. At the moment, 0.2 million people have a marginal deduction rate of over 80%. Under Universal Credit, that will be zero. In fact, the highest number for anybody under Universal Credit will be 76% if they are earning above the tax threshold. It will be 65% if they are earning underneath the tax threshold.

There are lots of ways of doing this, but both for single earners and second earners we have done the sums to demonstrate. It is a fundamental design principle of Universal Credit that this is to be true. In sweeping away lots of overlapping taper rates and putting in one and doing that consistently, you can get some extraordinarily big results, and that is what that is telling you.

Q78 Jackie Doyle-Price: That echoes what Mr Troup was saying about the key being the structural blocks on which we deliver our benefits, rather than co-ordination. It is a step-by-step approach. You are actually attacking the biggest sources of detriment and difficulty.

I want to ask Mr Lamey some questions. One of the facts in the report is that we have £9 billion worth of overpayments on Tax Credits. Obviously, that is inherent in the structure of the system because, by its very nature, it is retrospective. How much of that overpayment can you get back?

Steve Lamey: Not as much as we would like. The £9 billion is an accurate number that the NAO uses. The numbers that we have on our accounts are slightly different. They were £12.2 billion for a couple of years after, so I can use the £12.2 billion number that we are very familiar with. Of that £12 billion, £4 billion is recovered by what we call "cross-year recovery", so we are deducting directly the amount of overpayment from a previous award from people still in Tax Credit. We collect £2 billion through our debt collection purposes, so those involved are either not getting Tax Credits any more, but still owe us money for the overpayment or the award has been changed because of part-time circumstances. So typically, the level that we cannot recover will be about £3 billion or £4 billion when the people are insolvent, there are no assets to recover, we can’t find them or whatever.

Today, a balance of about £4.7 billion out of the £12 billion is still outstanding, and that is the work that we are currently doing this year. We get 40%-odd in a very efficient way; we get 20% in a more expensive way through debt collection and we struggle to collect the remainder. These people do not have the assets for us to recover the income from; they might be on different benefits because of not being able to work at all. It is a very difficult area. With a really good policy intent, overpayments have unintentional consequences of putting people in debt-sometimes, having done nothing wrong. The intent of the policy is to give them consistency over a period and certainty about how much they are getting, which is one of the valued areas that we discussed, but it does mean that we end up with debt in an area that we never thought would be that much. The reason why they get into overpayment is a legend in Tax Credits. We could all think of the reasons why people’s lives have changed to get them into Tax Credit debt, but never understand it fully.

Q79 Jackie Doyle-Price: It causes significant debt. You are finding it easier to get it back from people who are still claiming the Credit?

Steve Lamey: Yes.

Q80 Jackie Doyle-Price: But, by definition, their income falls quite significantly.

Steve Lamey: It could be because they have gone into full-time employment. With Tax Credits, you hope that they are moving up the employment ladder. Sometimes, they are more able to afford it, because they are now in more full-time employment than they were with Tax Credits but, given the economic situation at the moment, some will be back on Robert’s benefits.

We have a recovery method and can recover Tax Credits through the DWP benefits of about £10 a week maximum, which is effective. We are looking at pay-as-you-earn recovery, and are covering it through the pay-as-you-earn code as we go forward. When we piloted it, customers told us that it was hugely popular. Rather than have a big debt, we asked whether they wanted it recovered through their employment on pay-as-you-earn. Lots of our customers said, "Yes", and that is coming into operation in 2012, I think. That date is under advice.

Robert Devereux: May I add one thing? As we are having a conversation about a strategic level, the strategic answer to the question is to have real-time information.

Jackie Doyle-Price: Absolutely.

Robert Devereux: In practice, whatever you earn, whatever has been reported is the only number that I will take into account when adjusting. If it has overtime in it, good news, and I shall know that. It may have been mispaid or something else. The whole point of Universal Credit and RTI together is that a great deal of that nonsense disappears by definition.

Q81 Jackie Doyle-Price: Amen to that; it has been a long time coming.

Steve Lamey: We will have what we call "undeclared partners"-two people earning money together as a household. One of the big concepts for UC is how to join households together in real time, whether that is a possibility and how people should declare that they are living as a household, which would affect how UC is administered. We will take a huge amount out just by finding our disproportionate biggest loss in terms of income. That will make a real good added-value judgment. But we will then have others still to deal with. "Undeclared partners" is certainly an area that both of us are still concerned about to date.

Q82 Jackie Doyle-Price: That is a big area of perverse disincentive as well.

On to the £4.5 billion that I think you said was difficult to collect. Is there any evidence that overpayments of Tax Credit that result in big bills arriving at households discourage people from working at all?

Steve Lamey: I do not think that there is any evidence.

Q83 Chair: You should come to an MP’s surgery.

Steve Lamey: Now we can recover through DWP benefits, so if you are in work partially and receive Tax Credits to support you, we can recover the bill through that, and if you go into full-time work, we can recover it through that. It is not something that is on my desk on a daily basis-I get lots of MP letters, but that is not one I have seen before.

Q84 Jackie Doyle-Price: How are you not able to get that £4.5 billion back, then?

Steve Lamey: I guess for a number of reasons. The way we are tackling £4.5 billion has huge amounts of detail. A lot of that debt goes back to 2003, 2004 and 2005. Those in the Committee who remember when Tax Credits were introduced will remember that it was a very difficult birth. We have been working with the NAO, and in the debts that we looked at we could not find half the people, because they do not exist or do not admit to existing either in that name or in that family relationship. A lot of the debt that we are still carrying in the £4.75 billion is very old. So we are in the process of remitting that, which is not about writing it off and saying, "We’re not going to collect that debt cost-effectively." This Committee is about value for money, and we have looked at that £4.7 billion to see how much of it can be recovered in a value-for-money way. Probably about £1.5 billion of it will not be recovered cost-effectively.

We have a series of other campaigns in our error and fraud strategy, which is one thing that we are very proud of. We have stopped and changed the trend of error and fraud in Tax Credits. One of the key benefits of that is that it has reduced the number of overpayments. This year we will probably reduce the amount of overpayment by £1.7 billion-a lot of that would have gone on to debt, so one of the main things that we are doing is stopping the taps of debt flowing in by making sure that overpayments are grossly reduced.

We have other campaigns in which we write to people and say that they owe us money using debt collecting agencies. A range of different events will bring £700 million or £800 million back into the Exchequer this year. It is never the most popular thing I do. At every BCCG and social security committee I go to, I am asked, "Why do I have to chase people for debt that they didn’t mean to get into in the first place?"

Q85 Jackie Doyle-Price: But you put them in.

Steve Lamey: We put them in. The policy obviously gave people an opportunity to have a consistent income-

Q86 Mr Bacon: The policy also gave the accounting officer an opportunity to say to Ministers, "This is the sort of policy that will squander taxpayers’ money; it is not efficient, effective or economic, so you will need to issue direction to me if I am to carry it out." Did any HMRC accounting officer ever ask for such direction?

Steve Lamey: I knew that would come up today. I do not like passing bucks, and I would like to look at Robert or Edward but-

Q87 Mr Bacon: Mr Devereux is shaking his head-he is now looking at Mr Troup.

Chair: But no one is responsible.

Steve Lamey: The policy was implemented in 2003 and was being discussed in 2001. Thankfully, I was not in the civil service at that time, but my policy colleagues in the Treasury might have been involved in that conversation.

Edward Troup: It was a Treasury policy. The fact that there were going to be repayments was a conscious feature of the policy design, and, as Steve has explained, there was a good policy intention. It was inevitable that there would be repayments. I do not think that the scale of the issues that they gave rise to was understood at the time. HMRC did not have the engagement with that population-it did not understand that there were employments with people who moved around and whose circumstances changed, so the scale of the repayments was not understood. I am trying to remember who the accounting officer was at the time. I am sure that if he-it probably was a he-had felt that another direction was necessary, he would have sought one.

Chair: We doubt that in this Committee.

Q88 Mr Bacon: You were talking about the £4.7 billion. First, will you remind us how much new debt is being added on each year?

Steve Lamey: The new debt that we are forecasting this year-

Edward Troup: Sorry, to go back, I understand that, at that time, repayments was an Inland Revenue policy, so the then chairman of Inland Revenue would have-

Q89 Chair: You just said it was a Treasury policy.

Edward Troup: It is now. I have been reminded that it became a Treasury policy after 2005.

Chair: I do hate this. This is why we need somebody in charge.

Q90 Mr Bacon: You were about to tell me how much new debt is being added each year.

Steve Lamey: There is the figure of £4.7 billion, which we hope will end up at about £3.8 billion by the end of the year. In that difference between the two, there is about £1.5 billion of new debt that will be added.

Q91 Mr Bacon: That’s in one calendar year.

Steve Lamey: In one calendar year.

Q92 Mr Bacon: Then the calendar year after that. How much more new debt?

Steve Lamey: I haven’t got those numbers with me. We are trying to prevent, as I mentioned, error and fraud in the first place. That is what typically creates the debt. Our main target is to reduce the amount of debt incurred.

Q93 Mr Bacon: Perhaps the NAO will be able to enlighten us, but I seem to remember reading a figure recently that the expected debt-I think it was in HMRC accounts-was to go up from £4.7 billion to £7.5 billion. Is that right?

Steve Lamey: I have seen that number referred to. If I didn’t do anything, if I did not get my error and fraud numbers right, if I did not put the actions in place in our plan for this year, that is the level of debt that we would expect to see.

Q94 Mr Bacon: That is fairly scary, given that we have been at this for such a long time. I have one more quick question about the existing debt. Has HMRC done an assessment of how much of the existing debt is owed by people who are now overseas?

Steve Lamey: We would have had that data but I don’t know if I have got them. We did a very detailed analysis this year, with the NAO assuring us, as to how much of the outstanding legacy debt back to 2003, 2004-05 we could recover. We couldn’t find 50% of the people we were looking for and, of the other 50%, barely 10% acknowledged that they still had any Tax Credit debt. I really don’t know the answer: I suspect some.

Q95 Mr Bacon: Can you write to us on that? I am thinking specifically of the criminal fraud uncovered when an HMRC official wrote to me about it and I drew it to the attention of the NAO. People were coming here, getting low paid jobs, setting up a British bank account, claiming Tax Credits and then going home and getting the money out of a hole in the wall; then £5,000 or £6,000 later in parts of eastern Slovakia they could basically buy a house. That was only-I say "only"-about £300 million or £400 million. However, there were other frauds going on and other amounts from people who were intending to go back overseas or who had, for whatever reason. I am interested in the totality of the existing debt that, in HMRC’s assessment, is attributable to people no longer in this country.

Steve Lamey: I will certainly try to find that.

Q96 Chris Heaton-Harris: I want to know about the lessons learned form this. We have had Mr Troup describe it as a lack of understanding about policy, Mr Lamey saying it had a difficult birth and Mr Devereux saying it was nonsense. As we go forward looking at new means-tested benefits, will there be better policy understanding as a consequence and what lessons have been learned?

Robert Devereux: Maybe I can shoot into an open goal. If we go to figure 10, which Mr Hills talked about, you will see a chronology of means-testing assessment. Personally, I am quite proud to sit here and find that the one in 2011 called Universal Credit comes out with double ticks and triple ticks all the way across the piece.

Q97 Fiona Mactaggart: All of 2011 have no ticks at all under the impact on the individual.

Robert Devereux: You can raise that as a separate question; I am happy to do that. In terms of whether we have learned lessons from Tax Credits through to today, I would have thought that was a rather good illustration that we have.

Q98 Chris Heaton-Harris: I just want to know what lessons you have learned. I quite like the ticks, but I would like some details.

Robert Devereux: I am going to reprise myself. One lesson is, since people’s income and earnings do fluctuate wildly month-to-month, do not make six-monthly assessments. I will stick with that one for now.

Q99 Stephen Barclay: Can I come back to something you said earlier to Meg? Did you say that the "Tell us once" project has been put on ice?

Robert Devereux: What I said was that "Tell us once" has already been implemented in respect of bereavement notification. There were suggestions about doing other things as well, one of which was change of circumstance, which in the current climate I am not taking forward. That is the thing I am in conversation with the Cabinet Office about, whether it has resources to take it forward.

Q100 Stephen Barclay: So in terms of the original scope of the project, has it, in your view, been delivered? You can access it via Directgov.

Robert Devereux: I have not come prepared. Can I drop you a note on that?

Q101 Stephen Barclay: Sure. I just ask because, in January, the Department was telling The Guardian that, by November, you would be able to access it via Directgov. As we are now in November, I just wondered whether we were at that stage.

Robert Devereux: That may well simply be that you can access the thing that I have just described via that particular web channel. You can also already access it over the phone by talking to my staff.

Q102 Stephen Barclay: We are not talking about a-

Robert Devereux: Can I drop you a note?

Q103 Stephen Barclay: Absolutely, but the point that I was making was that the impression given was that this is something that, because of other understandable priorities, has been frozen. Your predecessor was describing this as a critically important project and there was huge ministerial backing. I am just keen to understand how much has been spent, where we have got to and why it is that something that was seen as so critical has suddenly been essentially put on hold. Could you spell that out?

Robert Devereux: I will happily do that. The answer to why Universal Credit was not part of the world which my predecessor was describing-

Chair: Okay. Matt.

Q104 Matthew Hancock: I want to come back to the central question of the plethora of means-tested benefits, their impact on work incentives and their cost. In figure 7, we find that "for a single-earner couple with two children increasing income from 33 per cent to 67 per cent of average wage in 2009" in the UK the effective tax rate is 84%, so they take home just 16% of their increase in income. Can you tell me how much of that is tackled by the Universal Credit and how much is not going to be tackled? You made the argument that you are focusing on the big things and there are other ones that are-I am putting words into your mouth-second order. I thought that the discussion at the start was one between somebody trying pragmatically to do their best to move in the right direction, and the vision and goal of where we might want to go if we could do everything we would like to in one go. I wanted to quantify that. Of that 84%, how much are you basically tackling and how much is left in systems that, while you have bilateral relations with them, are not part of this major reform?

Robert Devereux: My hunch, but you will have to ask the NAO, is that this chart is a very simple calculation of things to do with welfare payments, and it will have nothing at all to do with higher education and all these other things.

Q105 Matthew Hancock: All the extra ones. When you take all the means-tested payments-whether in kind or cash-based-do you have calculations of how much impact they have?

Robert Devereux: I do not. What I have already told you is that Universal Credit alone will move that 84% down to a maximum of 76%. There will be circumstances-we have had this conversation already-where you can then ask about what happens with maintenance or whatever.

Q106 Matthew Hancock: Exactly. I just wanted to try to quantify the Universal Credit versus all the other things we talked about earlier.

Robert Devereux: It will obviously depend upon which family we are talking about, and that is rather important. This is a chart for a very specific family with a very specific change in earnings. We could hypothesise that they have one or two children at university or two children under five. It is very difficult to give you a general answer, because of the question that says that people are all different.

Q107 Matthew Hancock: It depends upon the circumstances.

Robert Devereux: If you want to specify some circumstances, I will tell you what I think the best answer is, but there is no convenient ready-reckoner for this, because people’s circumstances differ: "Am I having maintenance from my partner? Do I have a part-time pension?" That is the real world, and it does not lend itself to an easy answer that says that I will go and take another 3% out.

Edward Troup: I was just going to add that, while we may not be co-ordinating everything, we are actually considerably improving the extent to which we are publishing analysis of the totality of Government policies on different sectors and the different quartiles and deciles of the population. We cannot do everything yet. The modelling is not sufficiently sophisticated, but at the last two Budgets we have published far more than previously. We are continually, with the DWP, attempting to incorporate into our modelling as many of the different benefits and receipts from Government as we can to give a good overall picture of what impact Government policies are having on different deciles. We are not there yet. It is a very difficult task.

Q108 Chair: Can we move on? We have to cover admin costs and the burden on claimants, and do a little more on fraud and error. Can I move you to admin costs? On page 19, figure 8, it shows that the costs of a new claimant Pension Credit is 351. The costs for income support are still high; it is 181. I remind you that the system will be place in 2017, so there is a long time to come. What is the justification for the difference, particularly as the Pension Credit has the highest level of fraud and error?

Robert Devereux: The short answer is that Pension Credit is, if anything, more complicated than Income Support. By and large, the customers who come for Income Support have relatively straightforward circumstances in terms of earnings or savings. They do not have to do calculations on lifetime pension awards or anything else that a pensioner might have. That is one point.

Secondly, the benefit itself is more complicated. I have mugged up for the Committee on what the savings credit element of Pension Credit does, but I am not sure that you want to hear it. It is almost unintelligible to most people. As a consequence, working that through to work out what the rules are and how it applies is more expensive. I have a benefit, which is more complicated by nature. It deals with people in more complicated circumstances just at the point when they are cashing in some form of finance asset and turning it into something else.

Q109 Chair: But, ironically, whose income is more stable probably.

Robert Devereux: Yes, but the rules of the benefit are complicated. Let us start again with the strategic answer to what we are doing about this. The reason why the Government have put forward proposals on the pension is that, at the moment, of the order of 40% of all pensioners are drawing means-tested benefits. A single-tier credit, albeit for new pensioners, will bring me out at 10%. The most important thing to do is think about the system as a whole. That is what the Government are doing, and that is what single-tier is all about.

You are bound to ask me whether I think that it is satisfactory; the answer is, obviously not. I shall give you examples. Unfortunately, I cannot do so for Pension Credit, but for Income Support new claim let me tell you something about productivity improvements. That is quite important because they go right to the heart of the unit costs. We have managed to make, even between the average in 2010-11 and in the first quarter of 2011-12, a 16% improvement in productivity in the actual processing of the claims. The numbers that you are seeing are absolutely all-up costs, so they have a share of my salary. They are not just the costs of the person whom they are literally working with on a Wednesday. At the level at which direct processing is going on, there are substantial productivity gains.

We are not satisfied that those numbers are right, but the reason that they have relativity is generally to do with the differences. To make my point again, if you then look at the existing claim costs, by and large, once you have got a pensioner on to Pension Credit, their circumstances are very stable at £47 a unit cost.

Q110 Chair: Except the fraud and error rate is highest. There is a figure somewhere on it. If you then look at Pension Credit as a proportion of the fraud in that particular benefit, it is the highest.

Robert Devereux: Yes. I have not brought with me the table that I had when we last spoke about fraud and error, but do you remember that we discussed whether we perceived fraud and error as existing at the start of the claim or being created during the claim? I am fairly sure that the figures on Pension Credit are to do with pre-existing fraud and error. It is not that pensioners’ circumstances change so much that they end up with fraud and error; it is problems existing at the start. Since I am not confident that that is right, can I drop you a line to ensure that it is correct?

Q111 Stephen Barclay: It is a remarkable thing that you said: some 40% of pensioners are getting Pension Credit. Taking someone on average earnings, at what age would it make no sense for them to start paying into a stakeholder pension, because they want to take themselves off Pension Credit?

Robert Devereux: I do not know. It is not as straightforward as you are implying for the 40%, simply because this savings credit thing, which I have just told you about and which is to do with our preparedness to taper away means-tested pension support, tapers away at a more generous rate if you have made some saving provision. For most people with no saving provision, it is taken away pound for pound with every extra pound of income that you earn.

Q112 Stephen Barclay: Sure. But there must be a median point or an average, whether it is at 50 or at 48 or at some point, where it makes no sense. The Government put huge impetus into stakeholder pensions as a product, yet designed a means-tested scheme that dis-incentivises people from saving.

Robert Devereux: Sorry, I am disagreeing, because the whole point of the savings credit, which is a top-up to the guaranteed cash benefit, which is what the basic Pension Credit is, is a top-up related to the amount of saving you have made. I will be more generous to you if you have more pension saving than if you have no pension saving.

Q113 Chair: Let us move on, because I am conscious that there may be a vote shortly. If we look at the burden on claimants, which is the one area we have not covered, we come to page 23. One of the arguments for universal benefits as opposed to means-tested benefits is all that stuff about complexity, cost and the stigma of applying. Presumably, if you want to meet the objective of your means-tested benefits, you want to overcome those barriers-[Interruption.] Is that the bell? Let me ask you quickly, if you look at that, claimant burden is not even looked at, except for Tax Credit.

Robert Devereux: Is it a quick answer and then we are finished, or is it a quick answer and then you are coming back?

Chair: Let us hear what the answer is.

Robert Devereux: I have here the chart that records whether the claimant burden was part of the formal impact assessment. I have looked at copies of the impact assessment, and it is correct. I have already told you at some length that I am going to a lot of trouble with real claimants in real places in Warrington, testing whether they can navigate their way through the online system. I am spending more time on that than any previous Administration have done, to ensure that claimants can navigate it. I will not simply say that I will make it as easy as possible, because part of you wants me to ensure that fraud is borne down on. We are giving away a lot of money. The question is: what is the reasonable amount of questions and testing I should do? I assert that the Universal Credit platform is far, far superior to anything else that you have seen, and I would be delighted to show it to you.

Q114 Meg Hillier: We’ll call you back in a few years.

Robert Devereux: A few years? I would be delighted if is a few years, but I somehow doubt that.

Chair: Thank you very much.

Prepared 23rd November 2011