Public Accounts Committee 

Oral evidence: Progress in implementing Universal Credit, HC 458

Monday 11 March 2024             

Ordered by the House of Commons to be published on 11 March 2024.

Watch the meeting 

Members present: Dame Meg Hillier (Chair); Paula Barker; Olivia Blake; Sir Geoffrey Clifton-Brown; Peter Grant; Sarah Olney.

Work and Pensions Committee member also present: Sir Stephen Timms.

Gareth Davies, Comptroller & Auditor General, National Audit Office, Laura Brackwell, NAO Director, National Audit Office and David Fairbrother, Treasury Officer of Accounts, were in attendance.

Questions 100

Witnesses

I: Peter Schofield CB, Permanent Secretary, Department for Work and Pensions; Neil Couling CB CBE, Change and Resilience Director General and Senior Responsible Owner for Universal Credit, Department for Work and Pensions; Helga Swindenbank, Director of Disability Services, Working Age and Move to UC, Department for Work and Pensions.

Report by the Comptroller and Auditor General

Progress in implementing Universal Credit (HC 552)

Examination of witnesses

Witnesses: Peter Schofield, Neil Couling and Helga Swindenbank.

Chair: Welcome to the Public Accounts Committee on Monday 11 March 2024. Today, we are looking at the progress in implementing universal credit, which was established as a policy plan in 2010, with an original plan for delivery by 2017. Since various resets, the cost of implementing universal credit has increased by over £900 million, and its completion date has been pushed back by six years—not from that original date but six years to 2028.

DWP’s own evaluations show that universal credit claimants are more likely than legacy benefit claimants to be in work after making a claim, which is good news. It is by a couple of percentage points, but in work placement terms, that is very significant from the work that this Committee has done in the past, so that is good news, but those evaluations look only at the short-term impacts on claimants, so there will be work that needs to be done, particularly for those who claimed after 2018.

The other big issue for universal credit is the many people on legacy benefits who have not yet moved to universal credit. Some have had their benefits stopped and DWP does not really have a handle on why that is, so we want to ask some questions of our witnesses about that today. The migration is continuing and it should be in place by 2028. That is what we are hearing from the Report by the National Audit Office, so thank you to them for that. We are wanting to quiz our witnesses on how realistic that delivery date is.

Welcome to our witnesses. We have the permanent secretary from the Department for Work and Pensions, Peter Schofield. He is joined by a frequent flyer at this Committee, Neil Couling, who has, from the inception of universal credit, been in charge of the programme. To read out his full title, he is now the change and resilience director general and senior responsible owner for universal credit. Every time you come in front of us, you have more to do. That is a hint for you to ask for a pay rise from your colleague. I am delighted to welcome you.

Neil Couling: If you would like to make that a recommendation, Chair, I would welcome it.

Chair: Sorry, the Public Accounts Committee is not that generous. Helga Swindenbank is the director of disability services, working age and move to universal credit. Again, Mr Schofield, you seem to give everybody big jobs at the moment, but welcome to you, Ms Swindenbank.

Q1                Sir Geoffrey Clifton-Brown: Good afternoon, Permanent Secretary and other witnesses. Can I start by going to paragraph 1.14 on page 22 and asking you one or two questions about the business case for and the savings from moving these benefits on to universal credit? In particular, I want to ask you about value for money. You have assessed that, “The value for money case remained strong, with a net present value of £39.3 billion over the lifetime of the business case from 2022-23 to 2026-27.” You are claiming quite big savings from this programme. Can you give us any assurance that these figures are realistic, or is there a big optimist bias in those figures?

Peter Schofield: I am glad to, Sir Geoffrey. Figure 6 is the one that sets out the basis of those numbers. The starting point is the operational costs of running the service. With the implementation of a digital back office and all that goes with the automation, there are significant operational savings from the rollout of universal credit. Those are set out at the beginning—the difference between the £2.5 billion savings and the £2 billion of costs. There are further details on those savings further on in the Report.

If you look at figure 7, the NAO has found that we are on track to overachieve those. Savings outlined in paragraph 1.28 were significantly more than we set out to achieve. If you look at the overall operational savings over the period that we are talking about, before you get on to the other economic benefits, those already outperform the investment costs of £2.9 billion in the programme in its entirety. That is the first point.

The second thing, as the Chair said at the start of the hearing, is that there are then the economic benefits that come from getting more people into work and the effectiveness of that. The business case in 2018 made an assumption that, at any one time, once we had fully rolled out universal credit, we would have 200,000 more people in work as a result. As the Report sets out, it is now not possible to claim any of the legacy benefits, so there is no counterfactual, and so it is very difficult to produce an alternative universe where you can see a different outcome under legacy benefits.

We have been testing the last remaining cohorts of people who claimed under the old system before universal credit was fully rolled out for new claims at the end of 2018. We have looked at the impact of universal credit in terms of whether someone is more likely to be in work as a result. Of the two most recent studies that we did on that, the one that the Chair referred to showed that, for single people on JSA, we were 2% more likely to get people into work. The last one that we did was on the basis of lone parents, which showed a 5% addition into work.

Those are all good indicators that universal credit delivers more people into work. Converting that into the 200,000 is difficult, because they are chalk and cheese, in a way. When we roll it out fully, we will have 8 million people on universal credit. Maybe half of them would be in work; 5% of that would be 200,000. That is a back-solving way to get to the 200,000 from the basis of the lone parent number that we talked about before.

On top of that, you then have further economic benefits that come from the fact that, when someone is in work, they are spending more in the economy; those are wider economic benefits that come. All of those together give you the £11.1 billion of wider economic impact that is referred to in that table.

On top of that, you have two other things. One is fraud and error reductions, and I am sure that the Committee will want to get into that in more detail. We always talk about fraud and error when I appear before this Committee, understandably. Just to give you one sign of that, the use of RTI data means that we are able to, basically, eliminate fraud and error from earned income. Fraud and error from self-employed income was about £1 billion last year, and earned income is more in the economy than self-employed earnings. I imagine that just eliminating fraud and error from earned income gets you more than the £1.5 billion that is referred to here.

Then you have the negative on benefit payments, and there are two dimensions there. One is that universal credit will encourage more people to take up benefits, because it is one benefit that replaces six previous legacy benefits. You might have had someone who was claiming housing benefit but not JSA or working tax credit. By applying for one, you get all, and part of the business case showed an additional take-up of benefits. That is then offset by the fact that, by having more people in work, there are fewer benefits that you are paying out, so the two go together.

What you see there is the 2022 update of the 2018 report. It gives us a really strong sense that universal credit is absolutely driving value for money. A £2.9 billion full investment over a period of years gives you, according to our business case, £10.4 billion of annual savings. It is spectacular from a value for money perspective. I imagine that the Committee wants to delve in a bit deeper.

Chair: Yes, we certainly do.

Peter Schofield: That gives a bit of a sense of where we are.

Q2                Sir Geoffrey Clifton-Brown: That is very helpful. You usefully referred to figures 6 and 7, and I want to ask you questions on both of those. Part of this whopping £10.4 billion that you are claiming is to do with reductions in cost. You have helpfully referred the Committee to figure 7. From 2022-23, with actual costs of £1.288 billion, rising to £2.3 billion in 2026-27, that is a very big forecast of reduction in operating costs. You have referred to a reduction—almost a halving—in current back office costs over the next four years or so, which is really a very big saving. Can you give us any sense of how you are going to achieve that?

Peter Schofield: We are achieving it. It is the truth. If you look back at the years that we have had, you will see that in 2022-23 we were delivering £349 million of savings. I might get Neil to say a little more about this, but we look at the total costs of delivering the legacy benefits. Those are DWP costs relating to the legacy DWP benefits—income support, JSA and ESA. They also include the costs to local authorities of administering housing benefit for the working-age population, and also the delivery by HMRC of working tax credit and child tax credit. In establishing the costs of the legacy benefits, you have to add all of those costs together and see that system proceeding.

That then gets replaced by DWP’s delivery through a digital system. It takes very significant numbers of headcount out of the delivery—and very efficient delivery at that. We are seeing it already, and we continue to demonstrate that going forward.

Neil Couling: You can also see it, Sir Geoffrey, in figure 8, just a little bit on from figure 7, and the unit cost figures that the NAO has set out there. You can see that there is a fall there in unit costs sustained since 2018. In fact, if I ran the data back before 2018, you would see unit costs of over £1,000. We hit the break-even point early in 2020. Since then, as the costs have been falling, you see the administrative savings start to pile up. It is because of the efficiency of having only one organisation doing this rather than three, and the fact that it is digitally enabled, that allows humans to now be freed up to help people into work rather than working on administrative tasks.

Q3                Sir Geoffrey Clifton-Brown: I accept that, and the unit costs have gone down hugely, from £593 to £195 last April, but the big drop in those unit costs was because of the large increase in claimants during Covid, were they not? Since then, we have not seen anything like that amount of drop in those unit costs.

Neil Couling: There are two things going on in the big fall in unit costs. One is the nature of the caseload. I remember coming to this Committee and, indeed, the Work and Pensions Committee in 2018-19 and saying that the unit costs will fall because the caseload mix will change. As the NAO points out in its Report, that has happened. We have also been investing in this system over the years, so that human beings no longer have to do the things that they would have had to do in the legacy system. That is what is driving these costs down.

They do fluctuate a bit. If Ministers ask us to do a new policy—for instance, giving people extra help to get into work—that will push the costs up, but even with new policies coming in, you are seeing the overall unit costs fall here, and they will fall further. We are ahead of profile on unit costs falling. On the plan, they should be about £250 at this point in time, and they are tracking below £200 now.

Peter Schofield: The other crucial thing to say, if you look at the difference between universal credit and the legacy system, is that the move to UC is bringing more people across from a more expensive legacy system into universal credit, so you are applying these lower unit costs to more cases. When you look at the overall cost of the system, that also drives some of the reductions that you see in figure 7.

Q4                Sir Geoffrey Clifton-Brown: That is very helpful. I do not mind, Permanent Secretary, whether you or Mr Couling answer this final question from me. That very large and, if you achieve it, creditable figure of savings of £10.4 billion of net benefit a year once fully implemented in 2026-27, is in table 6, which you referred to. Can I take you to the wider economic impact figure of £11.1 billion? If you look at note 3, that includes the £6.1 billion, which is increased economic output of £4.7 billion and increased employment of £1.4 billion. How measurable are those two savings, given that previous Reports have said that trying to measure the economic benefits of extra people being in work is very difficult?

Neil Couling: What I cannot do is create a facsimile of the UK economy and run one with universal credit in it and one with universal credit out. We explained that to the NAO back in 2018, and they reported on that fact then. That remains true. That is true for any big infrastructure investment. I cannot un-invent the M25 or Terminal 5 and so forth.

Chair: Yes, we get the picture. We understand.

Neil Couling: We cannot measure that, but we have done these comparative studies to try to show how universal credit performs against the legacy system. The OECD regards jobseeker’s allowance as one of the most efficient labour market regimes in the world. Universal credit has outperformed that four times now, when we have run the tests on it. We have recently run it for lone parents, where we have had a 5% gain in employment, and that compares very favourably.

One of the great success stories—it is a cross-party story, so everybody should be happy with it—since the mid-1990s has been the increase in the lone parent employment rate, which is up from around 45% to around 66% now. There has been a layering of policies that have achieved that across the then Conservative Government, the Labour Government and the coalition, and now universal credit. You are seeing this upward drift, and it is a long success there.

The new deal for lone parents, tax credits and childcare changes estimated an 80,000 employment effect. The new deal for young people estimated a 30,000 employment effect. Working families tax credit was something like 25,000 to 59,000 in terms of employment effects. I cannot isolate them from that, but I can show you that the increase in the employment rate, and particularly the employment rate for lone parents, suggests that these policy interventions are working in terms of the magnitude in the business case.

The final thing about the business case that I want to get across to people is that, years and years ago, Iain Duncan Smith was trying to prove the investment case. Is it worth putting money into benefits to drive outcomes? The business case proves is that it is worth doing that.

Q5                Sir Geoffrey Clifton-Brown: There are quite a lot of estimates in there, and it is a very big figure that you are claiming. Is there an element of optimism bias in the wider economic impact? £11.1 billion is a big figure.

Neil Couling: If anything, there is pessimism bias in that figure, because I have not adjusted the employment estimates for the big reduction in the taper from 63% to 55%. I have not adjusted the employment effects for the big increases in the work allowances either. If anything, I have biased against that. The truth is that, if you get somebody into work, you get a double gain. You get reduced benefit expenditure and extra economic activity, which is how the £6.1 billion here breaks down.

Peter Schofield: The other dimension is that, when you have someone already in work, with the taper smoothed in the way that it has been and lowered in the way that Neil has described, there is more encouragement and incentive to do more hours and more work, which drives economic benefit. Going back to what the Chair was saying at the beginning, small employment effects drive very significant economic benefits.

Q6                Chair: You are measuring only for the six months in employment at the moment. What about the extent afterwards?

Neil Couling: That might be misunderstood. We measured the effects for a short period. There is a cohorting effect going on here. While you might see some decay in the numbers of people who went into work six, 12 or 18 months ago, they are being replaced by people in the labour market. There is a sustained uplift in the labour market as the cohorts come in behind each other and, ultimately, replace everybody on legacy benefits over a period of time. This is sustained and long term, but we have measured the effects on individuals only in the short term.

Q7                Sir Geoffrey Clifton-Brown: Maybe I have misunderstood what you just said, Mr Couling, so I want to come back on that. Surely, the whole object of this exercise is to reduce the number of people who are currently claiming benefits but can work, so you are not expecting to replace like with like. If anything, you want to reduce the number of people on benefits and increase the number of people in work, not just replace like for like.

Neil Couling: Yes, that is right. Imagine for a second that the population is static—obviously it is not. Somebody who has been on benefits for 10, 15 or 20 years has a likelihood of, say, 10%, of going into work under the legacy benefits. If the new benefit system under universal credit, as we bring it in, has a likelihood of 12%, or a two-percentage-point difference for unemployed people, as time runs on and people age, they would leave into pensions if they did not leave the benefit system. They are replaced by people with more likelihood of being in work, so this compounding effect goes on.

This is why you can ultimately claim the employment effects from things like the new deal for lone parents or for young people, or universal credit. They all have a positive effect here, and that is why, over time, although it bobbles around, the general employment rate in the United Kingdom has been up. The employment rate for lone parents is a stunning piece of history that hardly anybody ever focuses on, but it is one of the most profound changes that I have seen in my career. In the mid-1990s, we were one of the worst performers in Europe for lone parent employment. We are now doing really well.

Chair: We have the message loud and clear. We can digress, not that it is not important. I should welcome Sir Stephen Timms, Chair of the DWP Select Committee, who is guesting with us today.

Q8                Sir Stephen Timms: Thank you very much for your welcome. Can I ask a couple of questions, first of all, about the history? I have been looking back on my well-thumbed copy of the universal credit White Paper. It always intrigues me. On page 37, it refers to completing the transfer to universal credit by October 2017, but that was not what I was going to ask about. On page 51, it talks about the £2 billion cost of implementation. It is intriguing to me that, in 2018, the Department was still talking about a £2 billion cost, and we are now expecting it to be about £3 billion. Given the difficulties at the start of the project, and the reset and so on, it seems quite surprising to me that the cost had not increased between 2010 and 2018.

Neil Couling: If you read that very carefully, which I assume you have done, Sir Stephen, it says “in that spending review”.

Sir Stephen Timms: Yes, it does.

Neil Couling: I was not running the programme then. I took over the programme in October 2014. One of the conditions that I put on me accepting the appointment was that I could change the timetable, which was not feasible at that point. I do not know if it ever was feasible, but it was not at that point. It was obvious that I would be slowing expenditure. When we looked at the position in 2018, we thought then that we would ultimately be spending around £2 billion worth of investment, so we are comparing slightly different figures here.

Q9                Sir Stephen Timms: In terms of the £2 billion in the spending review at the start of the project, you would have thought that, by the time you got to 2018, they would be rather more than that, given that there was that abortive initiative.

Neil Couling: Some of the costs that we assumed for 2016 are being paid now. This was an estimate from 2010 to 2015 in the White Paper. We would have needed more money for the second spending review after 2015 to complete the programme. By the time that we get to the second spending review, we have elongated the programme. It had slowed down. It had come to a crash at one point. There is no great secret to that. We had to, effectively, re-plan out of that situation.

The truth is that doing the IT ourselves rather than contracting it out proved to be a lot cheaper. At 2018, I am looking at investment costs of about £2 billion, but, when we re-cost this for 2020, various things have changed. The programme has moved out over an even longer time. We may come on to this in a bit, but, in terms of what is happening to those costs, things that I assumed in 2017 would be in business as usual by then remain on the programme’s books. For example, just this afternoon, we announced a further extension of Help to Claim. That appears as a programme cost at the moment. If the programme did not exist, it would be in the Department’s normal running costs, so there is a bit of accounting jiggery-pokery going on here. It is not suggesting that there has been a massive explosion in costs.

Q10            Chair: You used the word “jiggery-pokery” in front of the Public Accounts Committee.

Peter Schofield: There is no new accounting. It is all very transparent and clear, and follows best practice.

Q11            Chair: But it is real cost to the programme, because it would not be there if it was not relevant to universal credit.

Peter Schofield: You have things like the operational costs of running the digital underpinning infrastructure, which sit in the programme. At some point, when we have unwound the programme, they will sit within our digital teams.

Q12            Sir Stephen Timms: They will move across. I am entirely persuaded of the advantages of universal credit over the legacy systems. I am not yet convinced that universal credit is any less prone to fraud than the previous system. Given that fraud is now a lot more than it was in the past, what is the evidence that universal credit is better for fraud?

Peter Schofield: I will have a go and then hand over to Neil. Part of that is picking up on the point that I was making to Sir Geoffrey earlier, but Neil will want to add to that.

We have talked in this Committee and, indeed, at the Work and Pensions Select Committee with you in January, about the underlying propensity for fraud and error, which would be an issue whatever system we had. If you look at the business case, both from 2018 and from 2022, we particularly focused on the infrastructure that we were putting in place using data to help us to manage the risk of fraud and error in a different way.

The particular example that I would give is looking at fraud and error from earnings from employment, which, in the latest figures in May last year, were virtually zero under universal credit, compared with fraud and error for those who are self-employed, which was about £1.1 billion.

If you look at the other big drivers of fraud and error in universal credit in the May stats, living together was a key part of that. There was also capital fraud, which relates to the controls that we have in universal credit but not in tax credits.

It is hard to see why any of those key drivers for fraud and error that we see in the May statistics would have been different under the legacy system, other than for earnings from employment, which have come down to virtually zero.

Q13            Sir Stephen Timms: How much of the fraud and error reduction of £1.3 billion is down to that?

Peter Schofield: There is a challenge in producing the counterfactual, but, if you think that fraud and error from earnings from self-employment were £1.1 billion, and that earnings from employment are a significantly higher proportion of overall earnings than earnings from self-employment, it would be that sort of order of magnitude.

The other big thing that has gone to virtually zero is fraud from childcare, which we have designed out of the system with universal credit in the way that we pay it. That, again, is virtually zero, and comes out as zero in the sample. Again, that would be some tens of millions of pounds.

Q14            Sir Stephen Timms: Given the difficulty in establishing the counterfactual, and for reasons that you have explained, I just wonder whether it is really valid to leave the £1.3 billion a year saving in figure 6, because we just do not really know, do we, what it is?

Peter Schofield: To get virtually zero fraud from earnings from employment is a very significant reform, which would be worth hundreds of millions of pounds, if not £1 billion, in its own right.

Q15            Sir Stephen Timms: How long will it be before the targeted case reviews start to generate savings?

Peter Schofield: They are already included in the AME savings target of £1.3 billion that we have set for the current year. At a relatively small scale, they build going forward. Neil, you might want to update on where we are in terms of delivering TCR.

Neil Couling: We might see savings of £100 million from TCR this year.

Q16            Sir Stephen Timms: That is in the current financial year.

Neil Couling: We are pushing quite hard at it. £100 million would be a good result.

Peter Schofield: Two things are crucial in the targeted case review. The first is demonstrating at a relatively small scale that it works, and that we are getting AME savings per hit above where we want to be. The second is building the capacity, and we have reported before on our plans to recruit into this capability. We are a bit below 3,000 people now working on targeted case review, which is slightly behind track, but we are looking to build that.

Q17            Sir Stephen Timms: Do you have an estimate for next year?

Neil Couling: It is around £400 million, but I might need to write back to the Committee, if that is not accurate.

Peter Schofield: There are details in the NAO’s Report on the annual report on accounts from last year. Those figures are in the public domain.

Q18            Sir Stephen Timms: A different thing that I particularly wanted to ask you about, though, is that we have learned from the NAO’s work that 20% of tax credit claimants are closing their claim rather than applying for universal credit, which seems a very large number. What do we know about why so many are not claiming universal credit?

Peter Schofield: We have been doing quite a lot of research on this, and we published the latest set of research on 29 February. Some of the research is already contained in the NAO’s Report. There are a number of things. It is something that we have been looking at very closely through the whole discovery process, and we `have changed some of the things that we have done as a result.

Early on, we discovered at a very small scale that some tax credit customers were a bit unsure around the fact that this was something coming from DWP, and they were used to hearing from HMRC, so we changed the migration notice so that it referred specifically to HMRC as well.

In the very small discovery numbers that we were looking at early on, it emerged that some people thought that, if they did nothing, they would automatically move on to universal credit and would not have to take any action, so we changed the migration notice on that. We made all of these changes ahead of then moving to scale from April last year.

Since then, we have done the analysis that we published a week or so ago, which gives a bit more detail. Some of it is that people’s circumstances have changed since they were last assessed for tax credits. If you recall, with tax credits, you are assessed on an annual basis, and people’s circumstances very regularly change. You get to the point where you have your migration notice. Some customers may not be on tax credits, but they may no longer be eligible. There is some evidence that people use those online benefit checkers and find that they would not be eligible and decide not to try.

Q19            Sir Stephen Timms: How much of the 20% might be in that category?

Peter Schofield: I will bring Neil in, but I will just do the other categories and then come to those. The second category is those who look at their circumstances and think, “I am only going to get a very small amount of universal credit.” Particularly for the self-employed, there are quite a lot of requirements around conditionality in universal credit, and they choose not to do so.

Q20            Sir Stephen Timms: The then Minister wrote to me last March to make the point about “robust governance arrangements to assess our readiness to scale the migration of tax credit claimants to ensure it is safe and secure to proceed.” What are the criteria that you have used to decide that it is safe and secure to proceed?

Peter Schofield: There are a set of readiness criteria that the programme board looked at. That is one thing, and I will get Neil to talk about those. The second thing is about what we are looking at going forward to check that we are on track and that we should be reassured by where we are. Neil, do you want to talk about the readiness criteria?

Neil Couling: Could I just say one thing on this no-claim rate? Paragraph 2.13 of the Report details the NAO’s exploration of this. We did a survey with Ipsos. We are planning a survey in April to try to contact everybody who has not claimed. The difficulty in answering your question, Sir Stephen, is that not very many people respond to that kind of inquiry. That is the reason why we asked Ipsos to do it, rather than ourselves, but we are going to try to contact everybody. I do not think that very many will respond. We are continuing to explore this. The thing that I draw some comfort from, though, is that we have a very low level of complaints. It is a tiny handful of complaints. HMRC is getting about one a month at the moment.

Q21            Sir Stephen Timms: Is this non-claim rate among those readiness criteria?

Neil Couling: No. We assess whether the non-claiming is because we have created barriers to people claiming. That is what we check: not that there is a magic number that we are trying to satisfy here, but whether we have created barriers to people claiming and whether those are stopping them from claiming. There is no evidence at all that it is the process. People are saying that they understand the process and the choices that they are making.

My initial reaction to this was that these seem strange choices—that one would turn down the opportunity to get money—but what is driving people to think that it is not worth claiming is the extra engagement with the state that universal credit brings, and the fact that the amounts of money are not high from their perspective, because they have relatively high earnings in their tax credit award. This has now gone on for 12 months. In every month that we have tried this, it is a similar sort of amount, so that may be the natural no-claim rate.

Q22            Chair: When you say “not much,” do you have an idea, from the work that you have done, of what figure they consider negligible or not worth the effort to claim?

Neil Couling: It is noticeable that those not claiming have median rates of entitlement of tax credit that are much lower than the ones who are claiming.

Q23            Chair: Can you give us a figure or a range?

Neil Couling: In response to a freedom of information request previously, we have said that the median for people not claiming is £3,200 a year, and the median tax credit award is £7,500. The 80% who are claiming tend to have higher tax credit awards, so they are porting over to universal credit.

Q24            Chair: That is a significant amount of money nevertheless.

Peter Schofield: What you do not know, of course, is whether at their next annual review they would still be entitled. Interestingly enough, the NAO Report says that 80% of people who move across from tax credits bring a debt with them, which would include people who had been overpaid in the previous year, from whom we would recover that, because their circumstances have improved.

Neil Couling: In terms of the volatility inside the tax credit load, just to remind the Committee, only 51% of tax credit claims at any renewal have not changed. Two-thirds of the other 49% are overpaid, and a third are underpaid, so you have real oscillation going on there.

Peter Schofield: There is a really important point that I want to get across to the Committee. Of the numbers that we have looked at in the Report, 99.5% are people who have been on tax credits. People are often in work and have an alternative source of income. I would be much more concerned about these statistics when it comes to people moving across from the DWP legacy benefits, which I am sure we will come on to. I have a very different view about those.

Q25            Sir Stephen Timms: When will the research that you are doing on this be finished? You said that it was starting in April.

Neil Couling: I do not know exactly when it will be finished, but, if your follow-up question is, “Will you be publishing it?” the answer is yes. You will have seen that I have published three sets of data now on move to UC and, as part of building confidence, as this Committee has said before, we should publish the information that we have. We also take the findings to our stakeholders and talk with them about what we have found, what they are finding and whether we are missing something.

Q26            Sir Stephen Timms: Could you send us a list of those readiness criteria, so that we know what they are?

Neil Couling: Yes. Should I quickly go through the five basic criteria, just to save me writing a letter?

Chair: Yes. We all like to save paper.

Neil Couling: We look at end-to-end service maturity. We look at affordability: can we and our partners afford to run the service? We look at the security of the system. We look at service stability: is doing the move to UC going to disrupt our ability to run new claims? We look at the legacy and people impacts: are claimants able to cope? Do they understand what is being asked of them? Can they do that? How much support do they need? You will see a difference—and we may talk about this—when we move from the tax-credit-only cases into the DWP benefits. We are changing our approach there.

Q27            Sir Stephen Timms: In that last criterion, this point about 20% not claiming is one of the things that you will look at. Is there a point at which you would say that that no-claim ratio is too high?

Neil Couling: If we were picking up complaints—if people were coming to their MP, for example, and saying, “Look at this. I have been left without support. I did not understand what was going on”—that would lead us to be concerned. This could be the natural rate of not claiming for tax credit customers. When we did the discoveries for all benefit types, we found that people on the DWP benefits who have no other income are practically all claiming. In our second discovery, the rates for jobseeker’s allowance was 100% claiming, and 99.99% for income support. You are seeing very high rates for claiming there, so this could be people making choices not to claim, which are not irrational.

Q28            Sir Stephen Timms: Can I just raise one final point? Advice agencies have said that they are worried that some of the universal credit leaflets are misleading and prompt some people to apply for universal credit before they receive their official migration notice to make sure that their benefit is not cut off. If they do that, they then lose out on transitional protection as a result. Is that something that you are aware of and might be able to address?

Neil Couling: When we have done previous leaflets, we have taken them to the stakeholder groups and discussed the risks. It is hinted at in the third chapter of the NAO Report, which says that about 51% of people currently on employment support allowance, who will have to wait until 2028 to come across to UC, could be better off if they made a move now. It is, for the reasons that we and the stakeholder groups know, quite a tricky process to prompt people to claim.

We have had isolated incidents and evidence of people who have been in this situation and who have claimed, mainly in tax credit cases, but we have not seen people respond by making voluntary claims in any great number, as far as we can tell from the data.

Q29            Sir Stephen Timms: Going back to the no-claim rate, some people who may not claim will be people who have savings of more than £16,000 and are entitled to tax credits but not to universal credit. Do you know what that proportion is?

Neil Couling: No. It is one of the things that we will try to find out from the research, but it relies on people telling us. I know of anecdotes. In fact, I listened in to a call with somebody who had £171,000 in the bank and said that they did not feel the need to claim. We still said to them, “You could get a year’s worth of universal credit here,” but they did not want to. I have nothing that I could share with the Committee or point out, but the research should try to bring that out.

On our estimates for the number of people with capital, I have said before in fraud discussions with this Committee that the family resources survey underestimates the number of people with capital. That is partly what is coming through in the fraud numbers as well. There may be more people who have capital and are in receipt of tax credits than we initially thought, but we do not know that, because tax credits do not need to collect capital information. We are collecting it only when people claim UC.

Q30            Chair: Mr Schofield, is this money that you are saving baked into your spending review for the current period?

Peter Schofield: Our spending review runs to March next year. If we are talking about the operational savings, the DEL savings are baked in, as well as the efficiency savings that our other transformation programmes are driving through, but that takes us to March 2025.

Q31            Chair: So any extra is a dividend to the Department, which, presumably, will be baked into the next spending review. At the moment, there is no additional benefit to you of achieving this savings target.

Peter Schofield: No. You can imagine that the Treasury has a great deal of access to our business case.

Chair: I see the Treasury Officer of Accounts grinning from ear to ear.

Peter Schofield: They bake that in and then we have a conversation.

Q32            Chair: I am just double-checking, because that is a good way of incentivising. On the other hand, as I say, the Treasury would be not doing their job properly.

Peter Schofield: It would be a great way of incentivising.

Neil Couling: Let us just be clear: we would like everybody to claim universal credit. This is not a source of pleasure.

Peter Schofield: In terms of operational savings, to the extent that people who are entitled to universal credit are not—

Chair: I was not talking about the underclaiming. It was about the operational savings.

Q33            Peter Grant: Good afternoon. Before I go to my questions, I want to follow up on a couple of the answers to previous questions. First of all, Mr Couling, in response to a number of questions from Sir Geoffrey earlier on, you were talking about how you know how much the additional economic impact has been from the introduction of universal credit, because that was a significant part of the anticipated savings. I am still not seeing how you can go from knowing how many additional people had a job at some point within a six-month period to knowing how much each of them contributed to the economy. Is it the case that you do not measure how long somebody stays in a job once they get a job? Is that correct?

Peter Schofield: The 2018 business case made an assumption that, at any one time, you would have 200,000 extra people in employment under universal credit than you would have had under the legacy benefits. As Neil said earlier, you cannot produce a counterfactual to demonstrate in an alternative universe that that is where you are, so you are constantly finding other ways in to look at that question.

If you look at lone parents, we are showing the data that was based on the 2018 cohort, which is the last time that we had people making new claims for legacy benefits, before universal credit was fully rolled out for new claims. Lone parents are 5% more likely to be in work than those who were on legacy benefits. Just to give you a bit of a sense of orders of magnitude, 8 million people will be on universal credit. Half of them, 4 million, will be in work. 5% of that is 200,000. That is a way of back-solving, and you can see how you get to that order of magnitude of 200,000 that is the basis of the core part of that £11 billion.

The remainder is based on other dimensions here, including the fact that when you have more people in work, not only are they contributing to the economy but their spending power is also driving further economic benefits. You can then add to that the fact that you will have more people who are in work and who will work more hours. You have got rid of the 16-hour rule, for example, under universal credit, which will encourage people who are just below the 16 hours to do more hours in work. That is not captured in the 5% number or the 2% number, but these are people who were in employment before and afterwards, but, because of the simple taper, they are more likely to take on those extra hours.

Q34            Peter Grant: Do you have statistics to show how many of the single parents who you mentioned end up with a secure job of 16 hours or more, for example?

Peter Schofield: The analysis that we did just showed whether they were more likely to be in employment.

Neil Couling: They got jobs.

Q35            Peter Grant: They could have a job for one hour a week on minimum wage. That would count as a job, from your point of view. Is that what you are saying?

Neil Couling: In our first study, we looked at the levels of earnings that people had. You can go on and look at this cohort in terms of the level of earnings that they were getting, and for what hours. The UC business case has two things. It talks about people into work and also about an increase in hours in work, both of which would bring economic benefits.

If you step away from individuals and look at the economy as a whole, even if person A moves out of work 18 months later, they are being replaced every six months by more people into work than the legacy system would have produced, so your net gain in the economy is an increase in the number of people employed. The business case does not say, “This is 200,000 people on this day who went into work, and they stay in work forever.” It is about the impact on the labour market.

You saw that recently in last week’s Budget, where there were three claims for labour market effects put forward by the Government, validated by the OBR, where they said that the national insurance change would produce another 98,000 people into work, the childcare changes 60,000, and the changes to the child benefit high-income charge 10,000. These are pretty standard things to do, and I mentioned that the new deals were all costed and organised on that basis as well.

Peter Schofield: The original 200,000 was based on analysis that had been done before, looking at things where we had simplified the tapers before, or where we had smoothed the administration of the system and how that had played in. It was all based on data from previous schemes in terms of what had been done. As a Department, we believe that this is robust data.

Q36            Peter Grant: When you talk about the data being robust, does that mean that, where something indicated a range of possible impacts, for example, you have gone for the mid-range one or for the bottom one every time? I have seen some reports that suggest that you have gone for the most optimistic impact every time in order to get to that 200,000 figure. Is that correct?

Neil Couling: No. I answered that earlier. If anything, we have underplayed our 200,000 estimate.

Peter Schofield: We would go for the mid-range, but then things have happened that make labour market incentives stronger under universal credit. In 2018, the taper rate was 63%, or was it 65%?

Neil Couling: It was 65% until 2018. In 2019, the work allowances were increased by £1,000. In 2021, the work allowances were increased by another £500, and then the taper rate dropped from 63% to 55%.

Q37            Peter Grant: Mr Schofield, I know that Sir Stephen has already asked some questions about levels of fraud and your confidence in achieving the savings at some point. You have had to introduce an additional cost of £415 million for targeted case review work. Why could that not be included in the original estimates for the cost of the project when it was set up?

Peter Schofield: We announced the targeted case review in December 2022 as part of our work to really drive out fraud and error and, in particular, to drive out the stock of fraudulent claims that had come into universal credit through the pandemic. It was an initiative where, very much in response to this Committee and the Work and Pensions Select Committee challenging us on the levels of fraud and error, we wondered what more we would do. We felt that this was a very important initiative that we could take to really stamp out the fraud that was already in the system. When the business case was done, it was pre-pandemic, and the world has changed very significantly, but the transparent thing is that these figures are included in the costs that we are anticipating for universal credit.

Q38            Peter Grant: The level of fraud and overpayment in universal credit was already way above the business case estimates before the pandemic was heard of. In 2019-20, it was up to 9.4%. The business case is based on it coming down to around 6.5%. Would it not be a fuller answer to say that part of the reason why you had to spend all that additional money on fraud prevention was because the estimates in the original business case for how quickly fraud would come down were too optimistic? Covid certainly made that worse, but the original estimates were never going to happen, were they?

Neil Couling: The NAO Report says at paragraph 1.22, “DWP’s assessment before the COVID-19 pandemic was that some of the expected impacts of UC on reducing fraud and error were being realised,” and that is where we were then. When I put together the business case in 2018, I thought that, in terms of the situation on fraud and error, we were tracking in the right direction.

We still have more work to do. Covid does, I am afraid, bring in new fraud and error. Underlying that fraud and error, we have this propensity problem going on, which we have spoken about in the Committee when we talk about fraud and error exclusively, and now some of that fraud and error is being masked by Covid. We were on a generalised rise, because that is what we have seen in other sectors of the economy. Fraud is up practically everywhere.

Q39            Chair: You have said this to us before and I am tempted to come in, because we cannot just say that fraud and error is going up because it is just going up in society. You and HMRC—and across Whitehall—have one of the best teams tackling fraud and error. We have private briefings about some of that work, which we would not want to play out in public. You cannot just say that it is rising and that your teams are not able to keep up with that, because they do know what they are doing. While not to the extent that we would love to see of it being reduced to nothing, they do foresee and stop serious fraud and error.

Neil Couling: During Covid, we stopped about half of the increase in fraud that came in. That was down to the team’s fast and rapid work there. What we are seeing, though, is a sea change in the propensity of people to commit fraud, which is continuing and going on. We are responding to that. Mr Grant asked why we need additional investment, and the answer is to tackle some of that.

Q40            Chair: More people are trying it on, so you have to throw more at it to catch them.

Neil Couling: More people are trying it on, yes.

Peter Schofield: One of the biggest challenges that we have is where changes of circumstance happen in people’s lives, and we rely on them to tell us.

Q41            Chair: For I do not know how many years, we have been discussing how you judge what someone’s main residence is, for instance. That is not necessarily a propensity for more fraud, but because people’s lives are more complex. Just to be clear, are we talking about more people trying it on or about more complex lifestyles?

Peter Schofield: It is clearly a combination of the two, although, when we classify something as fraud, it is very much in the former category. Just to give you a bit of a sense, the biggest drivers of fraud and error in universal credit are self-employed earnings, where we do not have the data feed from HMRC that we have for earned income. That is £1.1 billion on its own, so we are relying on people to accurately tell us what their earnings are.

Living together accounts for £1 billion. That is where you will have two people in a couple who then start living together, but will not tell us that that has happened in their relationship. Capital is £0.9 billion, which is, as we talked about before, about the need for additional powers to help us to get access to information around that. These key drivers rely on people to tell us things about their circumstances that we do not necessarily know.

Q42            Chair: There is this idea that it is a propensity to more fraud, but it is sometimes just ignorance. That is not an excuse, but what are you doing proactively to make sure that people declare?

Peter Schofield: It is an absolutely fair challenge. As we said before, we have done more to communicate to customers the need, for example, to tell us about savings. When we did quite a big communications exercise about that, we saw the fraud and error for capital come down, and so that is definitely part of it.

The key thing that we find is that so much fraud and error is due to changes of circumstance. When someone made a claim, they were accurate in what they said, but, when something has changed, they have not told us. The targeted case review is the opportunity for us then to look at a snapshot, and to go in and look at those cases and say, “The case is not what you currently tell us it is,” and it helps us to drive up the savings.

It is an important part of the armoury, but, ultimately, the way through is to have more data that enables us to prevent the fraud and error in the first place, and RTI is one element of that. In my answer to Sir Stephen, we have demonstrated that that is really effective at driving out fraud and error for that particular cohort.

Q43            Peter Grant: I want to look briefly at the decision that was taken in 2022 to delay the transfer of employment support allowance claimants to universal credit. First of all, Mr Schofield, just to confirm what the NAO has reported, that was something that was put forward by the DWP in response to a direct request from the Treasury to find a way of making extra savings on the benefits bill. Is that correct?

Peter Schofield: As the Report says, the circumstances in the autumn of 2022 were such that there was a need to find fiscal savings, and an interest in what could be done from the benefits bill to contribute towards that. The Report sets out that a series of options were put forward, and that the Government chose this as an option for making the fiscal savings that were necessary.

Q44            Peter Grant: When you say that a series of options were put forward, did the DWP put forward a number of options and did the Chancellor, through the Cabinet, make his decision?

Peter Schofield: It is not my practice to talk about how Government decisions are made in this Committee, but the Government made a decision and that was the decision that was made.

Q45            Peter Grant: From the figures quoted in the NAO Report, we are looking at about 600,000 people who would have been, on average, about £130 a month better off if they had transferred to universal credit. They are losing what would have been that entitlement every month until 2020. Is that correct? I can see you nodding, Mr Couling.

Peter Schofield: If you look at figure 14, you see the overall savings that come from the fact that some people who would have been better off on universal credit will not now be moved until later on in the decade.

Q46            Peter Grant: That is then, in fact, more than balanced up by the fact that there will be people who, if they transfer over once their transitional support has run out, will be worse off in the longer term. Crunching through the numbers in paragraph 3.7, I make it that there will be a longer-term difference or a reduction in benefit payments of £366 million per year at the 2022-23 estimates. Is that a figure that you would recognise?

Peter Schofield: Do you mean delaying those claimants who would be worse off on universal credit?

Q47            Peter Grant: I am looking specifically at 3.7, where the NAO has separated it out. It is not quite a 50-50 split. Some ESA claimants would get more if they transferred to universal credit; some would get less. If you work through the average benefit or the average loss, it is coming to a reduction in the benefits bill for ESA claimants of £366 million per year.

Neil Couling: The numbers at figure 14 are the savings.

Q48            Peter Grant: I am not talking about the impact of delaying the transition. I am talking now about what will happen once these people have moved over and have run out of their transitional support.

Peter Schofield: Transitional protection is not time-limited. That continues.

Q49            Peter Grant: It does not have a set time limit on it, but it does run out due to inflation.

Peter Schofield: Yes, it does erode.

Neil Couling: In April 2022, we re-forecast the gainers and those who would need transitional protection. For ESA, just going from memory, there were then about 1.2 million recipients in receipt of ESA, of whom 600,000 would gain, about 100,000 would have no change, and 500,000 would need transitional protection.

By the time you get to these figures, of course, as people age on ESA and they get to pension age, they will leave ESA. For example, if I had to move ESA claimants today, there would be about 800,000 I would need to move on to UC. By the time we get to 2028, that is now 600,000.

It is a bit confusing to follow all these numbers around, but, essentially, more people on ESA gain than have a lower notional entitlement, and the transitional protection bumps people up to what they were getting before.

Q50            Peter Grant: Mr Couling, we have a position here where about 600,000 people would be better off transferring voluntarily just now. Is there a way that somebody can get an absolutely certain estimate as to whether they would be better off and then apply for universal credit, or is it the same as it was for some other legacy benefits where, by the time they were far enough into the DWP giving them a calculation, they were already committed to transferring whether they wanted to or not?

Neil Couling: There are three benefit calculators out there, run by various groups and accessible via GOV.UK, which allow people to do these calculations. We have spent quite a lot of time looking at whether we could have a voluntary initiative here or a voluntary phase in and around the programme. Mainly, we looked before I developed all the things we need for managed migration. This was something we were looking at in 2021 and 2022 to see whether we could encourage people across.

When you talk to people on ESA, which we did extensively, despite over five in 10 of them being better off if they make the move, when you ask them about their own views on this, the number comes back as one in 10. You can understand this. People on ESA have often had the rough end of life. They are pessimistic and they do not believe that they would be in the gainer group. They all assume they would be in the loser group because they have had some hard knocks in their lives.

We then went to the lobby groups and stakeholders. We laid out these figures and said, “Do you think there is any way you could help us move people across?” The stakeholder groups were very worried, as indeed we were, that once you initiate this, if you get this wrong, you will be missing out on transitional protection, so we concluded that the best thing to do was to let people wait for the DWP to get to them and do the managed migration.

If you asked me today whether I would try to run a voluntary campaign to get people on to UC, I would say no. If that is what Ministers wanted, I would say, “Wait for managed migration.” Just bring managed migration forward from 2028, because that is the way I can ensure that everybody who is entitled to transitional protection gets transitional protection.

Q51            Peter Grant: Clearly, the estimate that you gave to the Treasury as to how much you could save by delaying the transition depended on most of those 600,000 people not finding a way to transfer voluntarily, did it not?

Neil Couling: Yes, because of this one-in-10 thing. People are very nervous about universal credit. They heard all the news about it five or 10 years ago, all that bad publicity. We know this. Maybe we will come on to how people are moving, but they move at the end of the three-month migration period and so forth.

Chair: We need to move on. We have got the point. Thank you. We are going to take a very short adjournment.

Sitting suspended.

On resuming—

Q52            Chair: Welcome back to the Public Accounts Committee on Monday 11 March 2024, where we are talking to DWP senior officials about universal credit.

I want to turn now, finally, to Helga Swidenbank, who has been waiting patiently while her colleagues have been talking about this issue, to ask a bit about migrating households on to benefits. We have talked a lot about tax credit migration today, but from April, of course, there will be the migration of other households. What assurance can you give us that you will be ready to do that work from April next year?

Helga Swidenbank: I have taken over the responsibility for the move to UC from December last year, and since that time I have been struck by a number of things. First of all, I have been struck by the motivation of my staff team. I have had an opportunity to go out and about and do some visits, including more recently to our team in Stockport. I have a really motivated staff team who are keen to move into the other benefits. I have also been struck by how well operations and the programme are working together.

Then the final thing that I have been struck by is the quality of the evidence and the way that the programme is doing the testing and learning, then rolling it out and building an evidence base. On that basis, I have a high level of confidence that we are ready for moving into those additional benefits.

Q53            Chair: Can you just give an example of that testing? You are basically piloting.

Helga Swidenbank: The discovery work we have talked about earlier certainly gives me a level of assurance that we have done some work with some of those non-tax credit benefits and really tried to understand what it is that customers want. We have worked closely with stakeholders and other groups to really understand what those customers want and how we can meet those needs.

Q54            Chair: Have you been piloting it or are you going to go straight out?

Helga Swidenbank: We piloted about 3,000 people through discovery.

Q55            Chair: What were the main lessons you learned from that?

Helga Swidenbank: Similar to what we have talked about earlier, it was about making sure we have clear, concise communication, and that our customers understand what it is that we are asking them to do.

Probably the big thing that we have learned, however, is how we make sure that we are working with the more vulnerable customers. We absolutely understand that that customer base is very different from the tax credit base, and so we have worked hard to develop an enhanced support package for those customers to make sure that we really are working with them, understanding what they need, and giving them extensions should they need them so that they transition as smoothly as possible.

Q56            Chair: Let us say I came to you as a claimant and I was nervous, perhaps because of what Mr Couling said. I had heard bad headlines from a few years ago, I thought that I might be losing benefits and it was all a bit confusing for me. What support would you put in place to make sure someone was comfortably going through that process?

Helga Swidenbank: Having sent out a migration notice, if we do not hear from a customer at week 11 or 12, we will write to them.

Q57            Chair: Week 11 or 12 of what?

Helga Swidenbank: Week 11 or 12 post the migration notice going out. We will write to them to let them know as a reminder, and also we will write to them and let them know that we will be phoning them to prompt and ask them about any concerns they might have.

Q58            Chair: This is after they have already migrated.

Helga Swidenbank: No, this is pre-migration.

Q59            Chair: This is 11 weeks after the notice.

Helga Swidenbank: This is 11 weeks after they have received a migration notice.

Q60            Chair: How far ahead is that of actual migration?

Helga Swidenbank: They have three months to migrate.

Chair: Of course they do, yes.

Peter Schofield: You have three months to migrate. As we do with tax credits, we would have already had a reminder earlier on. The new thing is what we do at week 11.

Helga Swidenbank: In week 11, we will write to them to let them know that we will be contacting them, and we will make three outbound calls to those customers. If we do not get a response from those outbound calls, we will write to them again to let them know that we will be arranging for a visiting officer to come and visit them to help support them through that process.

What we found through discovery is that people are often prompted by that letter about the visiting officer to contact us to then start the migration process. However, if we need to support customers through a visiting officer, we have visiting officers available. Again, we are recruiting additionally to make sure we have enough to be able to support that population.

Q61            Chair: We have seen this new contract again for supporting people. That is a telephone and online service. Are you using any third parties to support people and hold their hand? For instance, some people will have language or cultural issues that might make it more challenging.

Helga Swidenbank: We have a number of different channels. We have a telephony channel. Obviously, we have jobcentres. If we need to direct people into jobcentres, we can. We have a visiting officer cohort who will be able to go out and talk to people who need additional support.

Q62            Chair: You will have people in jobcentres able to support people as well.

Helga Swidenbank: Yes.

Peter Schofield: This is on top of the Help to Claim with Citizens Advice. This is what we do internally within DWP.

Q63            Chair: Citizens Advice is only online though, is it not? It is not face-to-face.

Helga Swidenbank: No.

Q64            Chair: We will come to that more in a moment, because we were interested to learn that when we were in our preparation.

It seems like you have learned quite a lot since the testing in September. If you were to look back at how you have handled this, are there any things you think you would have done differently? It is early days because you have not done the migration yet, but do you think that you prepared in the right timeline and have the time to deliver on the things that you have learned during the discovery phase?

Helga Swidenbank: We have. However, what I am also aware of about this programme is that we are quite agile. Should we need to make adjustments based on customer feedback or surveys, I am absolutely confident that we will be able to make adjustments and to meet those gaps we might have missed.

Q65            Chair: What if somebody, perhaps for various vulnerable reasons, does not claim at week 12 or does not answer their phone. It can be simply a number they do not know, they are not in when they visit or they do not answer the door. Will they just be cut off from benefits at that point?

Helga Swidenbank: No. When we come to make a decision around termination, my agents will review termination notices and termination decisions to make sure that we have covered absolutely every base and that we have considered everything prior to termination.

Neil Couling: We look back on the legacy systems, so we are checking for vulnerability; we are checking whether they have appointees or people who can speak for them. We have the power to extend the migration period as many times as we like.

Q66            Chair: Do you know what percentage of people you think will have a termination notice? Have you worked on that modelling at all?

Neil Couling: We think that the level of terminations will be around 2% to 3%.

Q67            Chair: That is quite a lot of people.

Neil Couling: It sounds like a lot of people, but that is the normal turnover in the caseload across the year. It looks like the terminations are people who are leaving the benefit naturally.

Q68            Chair: You hope the 2% is going to overlap completely.

Neil Couling: Yes. The other thing is that people ask why there are three months at the end of the programme. Why do we say we issue the final migration notices in September, finished by December, but the plan says March? It is because, when I look across the discoveries, the difficult cases are taking four to five months. It is just allowing ourselves a little bit of time to finish this and make sure we do not hurry. This is something this Committee said a while ago. Do not let the timetable dictate what you are doing. I have said that since I first came here in 2014, and I am not going to let the timetable drive us.

Q69            Chair: We could have a nice night out if we had a fiver for every time we have heard you say that.

Neil Couling: Yes, but I absolutely mean it. It is people at the back of this, not a timetable.

Q70            Chair: Exactly, and you are the Department for wonderful people, as your previous Secretary of State described you.

Ms Swidenbank, you paid tribute to your team. We certainly recognise the efforts put in, especially during the pandemic, but nevertheless you are dealing with people. If those people are not migrated and there are problems, are you confident you have every tool in the box you can to track people down and support them to continue to claim benefits, or that if there is a problem and it is picked up by, for example, their MP or a third party, you will be able to act quickly to put them on to the benefit that they deserve?

Helga Swidenbank: Yes, I am. We certainly have the volumes of people. We have learned through discovery, and we have a very highly motivated staff team who really do want to get this right.

Q71            Chair: The proof will be down the line, and we will be back in touch after that.

Peter Schofield: I had the privilege of going to Nottingham to visit some of Helga’s people and listening in to some of the calls. I was incredibly impressed, first of all, with how quickly people can get through, and, secondly, how helpful and constructive the team were. Maybe it was because they knew the permanent secretary was listening in.

Chair: Was the paint fresh as well?

Peter Schofield: Exactly, you always need to watch for that, but it is part of the culture that Helga’s team have.

Q72            Chair: We do not doubt that. You do not join the DWP unless you want to be helping people, one would hope, but the point is that a small percentage of errors can mean a lot of people in destitution. We have all had cases in our surgeries where people have come because their benefits have been cut off for one benefit or another at different times. They just have nothing for the next day and they have to have emergency loans. We know those cases arise. We are just trying to make sure that we are asking you the right questions.

Helga Swidenbank: We also have a two-tier assurance process. It is probably worth the Committee understanding this. Our tier 1 assurance is that my local line managers will assure themselves around the quality of calls and the quality of the work my teams are doing through listening to phone calls and watching how the teams are working.

Then we have a second level of assurance that is done through our corporate structure. Our customer experience directorate, independent of my operational line, will also make sure that they are testing the quality of what we are doing. Hopefully we have enough checks and balances in place to ensure that we are serving our customers well.

Q73            Paula Barker: We have touched on these with the Chair, but I just wanted to put some more meat on the bones. I am just referencing 2.14, which is on page 34 of the Report, where it speaks about leaflets being sent to claimants to understand the migration process and reduce barriers to claiming. Are they sent in different formats? Some people do not have English as a first language; some people require larger print, et cetera. I would be interested to understand that. I would also be interested to understand what the impact has been since those leaflets have been sent out.

Helga Swidenbank: We do have some of our customer base who require alternative formats, such as braille, for example. We have about 30,000 people who we are not activating the migration notices for yet, because we know we need to do some additional work to make sure we have the right type of communication for them to be able to understand and therefore respond.

I am not aware of the detail behind the other part of the question. Neil, is that something you are able to pick up on?

Neil Couling: The leaflets were about awareness rather than a call to action. It goes back to the question—I cannot remember if it was Sir Stephen or Mr Grant who asked it—about not triggering people too early. The material is available in those formats and, as Helga said, we have deferred some claimants at the moment until we have developed processes and products for them to be able to use. Braille is the biggest example.

Q74            Paula Barker: You have said the leaflets are more about awareness than triggering the application, for example, but surely you must get to a period of time where that would go up if the leaflets had been adequate. Has there been any impact assessment done on that?

Neil Couling: We are trying to get people to respond to the migration notice. We do not want them to respond before they get the migration notice, so it has taken some quite tricky communications planning here to try to get that balance right with people.

If you imagine this as a task of some very varied customers here with very varied circumstances, we are effectively sifting our way through here trying to pick out by type of cases, where we started with tax credit only to test those processes, and then to see what we need to do on top of that as we get into the customer groups that have more support needs.

Q75            Paula Barker: That leads seamlessly to my next question.

Chair: Mr Couling has been here often.

Paula Barker: If we look at page 36 of the Report, it says that you are reassured that the non-claim rate is not a cause for concern, as it has received few complaints. Mr Couling, you spoke before about whether there was an increase in claimants going to their MPs, for example, but 21% of tax credits did not claim, so they lost their benefits. Are you happy that the systems are right? It is not enough to say that, just because people are not complaining through your systems, everything is rosy in the garden.

Neil Couling: What have we done? This sort of non-claim rate has been pretty consistent through the entire testing and discovery process, and then as we have now been running it at volume. We have been trying to make enhancements and to see whether this works. Does this produce a higher claim rate or not?

Across this period, you have seen a rather rapid rise in earnings here, particularly at the lower end of the earnings distribution, as employers have had to compete for labour. Labour supply issues have been much talked about in politics beyond this Committee. There is a strong possibility that a chunk of those people are just because their earnings have gone up and they are no longer entitled, and they are making that choice themselves.

Then you are finding that there are two other things going on here. For the self-employed, there is a much greater rigour. We mentioned fraud before. We are trying to bear down on self-employed fraud here. We are probably driving some of that away from us, and people are deciding it is too much hassle to claim.

Then the amounts that people are entitled to are quite low for some people, and they are probably making a decision that that is a lot of engagement with DWP in order to get a reasonably small amount of money, so they think, “I am not going to bother.”

We are not sitting back. As I said, we are doing further research here, because I would be really pleased if more people were claiming. There is nothing in our research that suggests people are not claiming because they do not understand that we have built barriers to them claiming. It looks like they are making choices for themselves about what to claim and what not to claim.

Q76            Paula Barker: Your research does not include any follow-up contact, because it says that there are limited positive assurances that people are claiming benefits that they are entitled to, and no contact is made to ask why.

Neil Couling: I do not know how I would get positive assurance that people are not claiming what they are entitled to if the claimants will not talk to us about why they are not claiming. That is our difficulty at the moment. As I said, we are going to try to contact all the people who have not claimed. I suspect only a very small number will come back to us and tell us why, but if they do not tell us, I do not know how I get positive assurance.

It is a really difficult thing to do. There has been lots of work done on take-up over many years, talking about how to encourage and get that, and the only successful take-up campaign I have seen is quite recently around pension credit, where the late Len Goodman seemed to do the trick. I do not have a Len Goodman equivalent for tax credits, unfortunately. Everything at the moment suggests that people are making informed choices here, but we are going to do more work to try to find out.

Peter Schofield: I just want to emphasise what Neil has said, because I have said to this Committee before that it is as important to me that people who are not getting benefits and are entitled to those benefits are getting them as it is that we avoid giving benefits to people who are not entitled to it and eliminate overpayment, fraud and error. That is 31,500 people who so far have had their benefits terminated that we are going to try to ask, “Why didn’t you claim and what are the consequences of not claiming?”

We will work very closely with stakeholder groups, which may have more ability to find out about cases that we do. None of them has come forward with any cases where people have missed out and have not had those claims. Given that and the fact we have had very few complaints, we are assured, but we relentlessly keep looking to see whether we are missing anything. When we conduct this research, we will know a bit more.

Neil Couling: Also, there is one other thing that is worth bearing in mind. When your benefit is stopped, you then have a month to claim again. You still get transitional protection, and we will roll the claim back to the point where the claim is stopped. People are doing that, so around 4% to 5% of claims are coming round in that catching thing. Again, it is not a large volume of cases doing that. I thought that this may be people almost trying to stare us out a bit: “You will not switch it off.” For tax credits, there is not a large number of people reclaiming after we have stopped them.

Q77            Paula Barker: If we stay with transitional protection, though, stakeholders have said that they have concerns about how it is working and a lack of transparency. What work is being done? I know there have been workshops with stakeholders, for example, but what work is being done to help vulnerable people who may not go to one of the stakeholders that is in the Report? What is being done with the vulnerable people to help them understand the transitional arrangements?

Neil Couling: You asked Helga about some of the things we found in the first discovery. One of the things we found in the first discovery is that claimants did not understand transitional protection. It is a pretty simple concept, which is that you should get, if you move on to universal credit, at least what you got under the old system.

Achieving that is quite a complex policy. We found that if we tried to explain that policy to claimants, they were just mystified. I have to say that in the briefing for this session, I tried to explain it to my permanent secretary and obviously he got it brilliantly, but it is a very complex thing for particular cases.

In fact, the stakeholders asked us if we could produce a guide on all the complex cases, and we worked with them to do that. We are trying to turn that now into a form that more non-welfare advice experts could understand, but it is a complex area.

We have automated a lot of this so that calculations are done on the system by a tested algorithm that calculates people’s entitlement, but it relies on the right information coming into the system, which is the point that the welfare rights and stakeholders are making to the National Audit Office. We continue to work with them to help make sure that we put the right inputs into the system, but it is quite a hard area to do, because if I put all the rules about transitional protection out there with your notice of award, you would be reading it for weeks, with probably no greater understanding after that.

In your award, we say what the amount for transitional protection is, so what has gone into your calculation. If you are then not happy, you can ask us about it or you can go to welfare rights and get some advice there.

They are not picking up lots of cases where it is wrong. Where they are picking up cases where it is wrong, it is wrong on the inputting, so the wrong amount has been told to us. That is the thing we are looking at, and we are trying to make sure that the automated data feeds we get from the legacy systems give us the right amount, but it is a very complex calculation intended to try to achieve this measure of natural justice in the new system.

Q78            Paula Barker: You are effectively signposting individuals to welfare rights if they do not understand, et cetera. Welfare rights organisations, such as Citizens Advice, have all had their own challenges in terms of cuts. The Help to Claim service was originally for three years, and that had telephone, online and face-to-face, but since 2022 there has been no face-to-face.

I also noticed that there will be face-to-face support to vulnerable claimants in jobcentres and at home. As of September 2023, DWP had carried out 23 pre-claim home visits. That was against figures of receiving more than 30,000 calls between May 2022 and September 2023. It seems to be a very low bar that only 23 home visits were carried out.

You spoke earlier, when Dame Meg was talking, about recruiting new visiting officers. How many new visiting officers are you going to recruit? Is there any impact assessment being carried out about signposting to organisations who themselves are facing cuts? Is there any impact assessment being done on the taking away of the face-to-face contact in the Help to Claim service run by those welfare rights organisations?

Helga Swidenbank: We plan to have 55.5 FTE visiting officers by December, which is when we expect the peak of customers coming our way.

Q79            Paula Barker: Is that in addition to what you already have?

Helga Swidenbank: No, that is in total.

Q80            Paula Barker: How many is your total now?

Helga Swidenbank: In June, we are going to have 35 full-time. I could not tell you exactly how many we have today, but by June we will have 35. We have come to that volume, again, through discovery. We think about 10% of customers will need visiting officers, so that is how we have come to that number.

Again, as we have talked about, there is an agility to this. If we think that we need more, we can think about how we might respond to that based on customer feedback and feedback from stakeholders. Sorry, what was the second part of the question?

Q81            Paula Barker: It is about what impact assessment is being done with signposting to welfare rights organisations who themselves have faced cuts.

Helga Swidenbank: We have done some work looking at how customers get their support, and the vast majority of customers get their support from family and friends. Then the next layer down is through our own jobcentres and helplines. About 3% of our customers get support from Citizens Advice and other agencies. Again, we are really happy to hear what our stakeholders have to say about this and to think through how we might respond.

Neil Couling: I put some extra money in the Help to Claim contract that was announced earlier today for Citizens Advice to have extra resources to help with the move to UC.

Q82            Paula Barker: That does not include face-to-face.

Neil Couling: It does not include face-to-face, no.

Peter Schofield: We found that jobseeker’s allowance customers will have a relationship with the jobcentre already. This is coming out of the discovery work. They like to go and use those contacts and those relationships already. That is a very effective face-to-face channel that we have, on top of then the helpline that we run and the other support that we provide as well.

Q83            Paula Barker: You are assured that it is more cost-effective to the public purse to employ these additional staff up to 55.5 FTE, as opposed to putting it into these schemes with the Help to Claim services?

Neil Couling: The reassurance I would want as SRO is that we are not terminating somebody’s benefit whose benefit should not be terminated. I used to do visiting about 38 years ago, and I know that sometimes you have to go into the home to see what the situation is like for the individual. You could have a very distressed person there.

I like the fact that we have a high number of visiting officers here at Helga’s disposal to ensure that people claim, and that we encourage them to claim and support them in claiming. It is a vital thing that the Department does, and I have long been a fan of our visiting teams. I used to run them years and years ago as well.

Peter Schofield: The reassuring thing I would want to give to the Committee is that, particularly for this cohort that we are talking about—not the tax credit cohort, but this cohort that we are moving into in April—we are absolutely committed to making sure that they are able to make that claim.

We are agile and will respond as the data comes through, and we will constantly monitor the data in terms of claim rates going forward and what we are seeing in terms of people’s ability to get through on the helpline, people’s need for visiting officers and support in the home. We will work it through. With the scale of organisation that we are in DWP, we have the option of moving resource around if we need to.

At the end of the day, as Neil said, we can always extend a migration notice if we have not been able to provide the support.

Q84            Paula Barker: Could Ms Swidenbank write to the Committee and let us know how many FTEs you currently have as of today?

Peter Schofield: That is visiting officers.

Q85            Paula Barker: Visiting officers, yes. Sorry. Are you confident that the new visiting officers will be able to go through appropriate training in a timely fashion?

Helga Swidenbank: Yes, I am. I will write to you with the numbers.

Q86            Sarah Olney: HMRC estimates that 80% of the tax credit claimants who are migrating to universal credit are migrating with outstanding debt. That seems a very high number. Why is it such a large number of people?

Peter Schofield: A lot of it, Ms Olney, goes back to the way that the tax credit system works with an annual review of someone’s circumstances, particularly their earnings. For example, if someone gets a better job or a promotion or works more hours in the period after their annual assessment, then when they come to the point where they migrate to universal credit, they will have been overpaid by HMRC. That translates into an overpayment debt.

In the normal way, what HMRC would do for the following year of tax credits is that they would adjust down their tax credits to recover the fact that they were overpaid the year before, but obviously they are moving to universal credit, so that gets crystallised as a debt.

We then do something similar in that we make an adjustment to the payment of universal credit to deduct the overpayment that they had before. That is one of the biggest elements of the debt that comes over from HMRC.

Q87            Sarah Olney: In one of the previous reports from the Public Accounts Committee back in 2018, a recommendation was made about the amount of universal credit payment that could be reclaimed as a debt repayment as a percentage. It was 40% at the time; it is now 25%.

As part of that recommendation, it was revealed that the DWP has no real way of assessing financial hardship. In some cases, of course, even a 25% reclaiming is not going to be appropriate, because that will literally lead people to destitution. What is the DWP doing now to assess hardship and see whether it is appropriate to be recovering debt?

Peter Schofield: Obviously, everyone’s circumstances are different. We have support available through our debt management teams. From time to time, I listen in to those calls because I am keen to understand the nature of the support that we provide. Those teams are trained to be able to talk to customers about their circumstances. People often phone because they have looked at the debt repayment or the nature of the deduction that we are making from their benefits—not just universal credit, but particularly universal credit—and they are saying that that does not leave them enough to live on over the coming period.

Our teams then, in the light of that, will adjust down and put in place what is called a repayment plan that would be below the standard amount that we would recover, to give people longer to pay. Then we review that over two years.

It is something like 13%—the number is in the Report. About 360,000 customers are on a repayment arrangement. As I say, I listen in to make sure that we are treating our customers fairly, and we do. By and large, if someone phones up and says, “I can’t manage,” and they make a reasonable case and explain why, then our people will adjust downwards the repayment.

Q88            Sarah Olney: Are those transferring to universal credit with outstanding debts who may be facing the clawing back all given the details of the debt support helpline?

Peter Schofield: This came up in a recent hearing at the Work and Pensions Select Committee. I was asked about this about a year or so ago. One of the things that I was recommended to do was just to double‑check the nature of the letters that we send out when someone has a debt to make sure that they are aware of that. We have improved the communication to make sure that people are aware of the telephone number and what they are able to do by phoning up.

Neil Couling: HMRC tells the person moving across to universal credit that they are taking a debt with them.

Q89            Sarah Olney: You are confident that anyone who has a debt has access to the support they might need if they find that their debt reclaiming is putting them into hardship.

Peter Schofield: Yes, and I would want to use this opportunity again to publicise that, if you are a customer of universal credit and you have a debt that is leaving you in hardship through the deductions, please do phone that number. Phone our debt management teams, and they will be very happy to listen to the situation and make adjustments accordingly.

Q90            Sarah Olney: Mr Couling, the maximum deduction rate has reduced from 40% to 25%. What is the basis for setting the deduction rate at 25%?

Neil Couling: It is a choice made by Parliament in the end. It is all statutorily run. In fact, before in Work and Pensions Select Committees, I have been asked, “Why don’t you do what other organisations do, which is to do income assessments to decide how much debt is repaid?” We run a statutory approach for the social security system, so the maximum levels of a recovery are set out by Parliament, and then there are repayment plans under that where that proves to be too onerous for the claimant. It is because we follow a statutory system.

When we set it at 40%, we looked at the repayment rates across all the legacy benefits, and that was a 40% repayment rate in the legacy benefits. We set it at the same level for the new universal credit system. By about 2018, there was lots of debate about that. This Committee made a recommendation to this effect. Ministers listened to this and dropped the repayment rates first to 30% and subsequently to 25%. We are still lobbied about that figure. It is set out in legislation, so we would need to change the legislation before we could do anything.

Q91            Chair: Is it secondary legislation, presumably?

Neil Couling: It is secondary legislation, yes.

Q92            Sarah Olney: Are you tracking the number of claimants who have had their debt repayments suspended in terms of them not currently making repayments on their debt? Are you seeing that figure changing?

Peter Schofield: I am not sure we are seeing it changing. Obviously, if there is a short-term need for a customer to have that, then that is something that our debt teams would look at.

We tend to find that our customers want to be rid of the debt as soon as they can, so it is a balance, and our teams are very skilled at talking through with the customer what they can afford. They want to make progress, but obviously they have to manage their existing budgetary commitments. It is a fine balance, but our teams are well trained to be able to help the customer work that through.

Q93            Peter Grant: Can I come back to the question about tax credit claimants who do not continue their claim when they are being transferred to universal credit? We are talking about quite a lot of people. The original assumption was that 3% of them would not transfer; we are now looking at about 26%.

Neil Couling: Could I just come in on that original assumption? I did ask the NAO to put this in the Report, but it chose not to. It is the NAO’s Report, so that is fine; I am not complaining about that. I made an assumption in 2010. The Chair knows this. I was running the policy for universal credit back then. I was not the SRO. The only evidence I had was for DWP benefits, and it was for the move from incapacity benefit for ESA, and that was about 3%. As I explained to the Chair, that is the natural turnover rate. It was not that people were not claiming. I had no evidence on which to base the tax credits assessment.

At the time, public finances were under challenge and so forth. It was the first set of emergency budgets and the like. We decided to make a very safe assessment about what the no-claim rate was going to be. We put more money in the forecast than it turns out we needed to do, because we did not want to under-forecast the costs of universal credit.

I never made an assessment for tax credits back then. The only time I have had any evidence to make an assessment is since we did the discovery. At the discovery, as the Report correctly records, we have decided to change our assessment for tax-credit-only cases at 26%. It is 4% for all the other benefits, which include tax credits with housing benefit and the like.

Q94            Peter Grant: I am quite sure we would usually be quite happy if somebody made an assessment based on being pessimistic about how much money would be needed. We certainly would not criticise that.

Just to be clear, that is a 3% estimate, prediction or assumption—call it what you like. Was the business case based on the expectation that about 97% of tax credit claimants would end up on universal credit? Even though you thought the figure would be different from that, was that the figure that you based the costings on?

Neil Couling: The public expenditure forecast assumed, until we changed our assumption, that it would be 3% overall, but, as I say, we did not break them down into different benefit types.

Q95            Peter Grant: How much difference does it make financially to the total cost of the benefit if you are now looking at 26% not transferring over as opposed to a significantly lower percentage that the original case might have been based on?

Neil Couling: In last week’s economic and fiscal outlook, the OBR attempts to estimate that. Unfortunately, it has rolled it in with some other changes. As I mentioned, you have higher earnings going on that are driving reduced benefit expenditure. People have higher earnings because of the high rates of inflation and even higher rates of earnings from people at the lower end of the earnings distribution.

Q96            Chair: It is difficult to disaggregate.

Neil Couling: We have asked the OBR to disaggregate. I do not know what the answer to that question is, Mr Grant, at the moment, but you will get a disaggregated figure from the OBR when it has managed to disaggregate everything.

Q97            Peter Grant: Finally, Mr Schofield, we saw the statement about the award of a further two-year contract to Citizens Advice and Citizens Advice Scotland for the Help to Claim service. It is wonderful how there is always a statement of some kind from DWP the day that you are due to appear before the Public Accounts Committee. Are you able to write to us with details of the annual cost of that service since it was introduced?

Certainly, the numbers that are announced suggest that it is quite significantly less per year than it has been in previous years. If you do not have the numbers, it does not matter.

Peter Schofield: The numbers are in the Report. Is there a different set of numbers?

Chair: You have the most recent announcement. That is £38 million but over two years.

Peter Grant: Just for clarification, is the £38 million that has been announced today over the two years of the contract?

Chair: That is why we need you to write to us, unless you can readily answer that one.

Peter Schofield: That is for the two years from April 2024 to March 2026.

Q98            Chair: There is no bleed-over from previous years, so that is £38 million for two years.

Peter Schofield: Yes.

Q99            Peter Grant: When the announcement was made of the funding for 2022-23, for example, the figure quoted is £21.3 million for a single year. We are now looking at £19 million per year, and that is after a couple of years’ inflation on top of that. In real terms, the amount of money that is being offered is maybe 15% or 20% below what it was two years ago.

Peter Schofield: Yes. As Neil said, he has put a little bit extra in as part of the announcement today, but you just have to look at this in terms of the totality of support.

Q100       Chair: We have had a lot on the totality of support. It was really just about those figures, so we just need to make sure between us that we get them nailed down.

Neil Couling: It is lower because you have the maturation of universal credit going on, so you get to a point where you will have fewer new claims presented.

Chair: Perhaps we can discuss what we lay out in writing. We just want to get the figures nailed down because we pick up a lot of the issues when there are problems with advice. We are at the sharp end, so we see that very clearly, as Ms Barker highlighted.

Can I thank you very much indeed for your time? The transcript of this session will be available on the website uncorrected in the next couple of days. Given that we only have a couple of weeks until our Easter recess, the Report is likely to be published after that Easter recess, at some point after 15 April. Thank you very much indeed.