Social Justice and Social Security Committee

Meeting date: Thursday, January 11, 2024

Official Report 544KB pdf

---

Budget Scrutiny 2024-25

Our next item of business is an evidence session on the Scottish Government’s budget for 2024-25 with the Scottish Fiscal Commission. This is an opportunity to consider the commission’s forecast spending on devolved social security benefits and how that is reflected in the social security budget for 2024-25. I welcome to the meeting Professor Graeme Roy, chair of the commission; Professor David Ulph, commissioner; and Michael Davidson, head of social security and devolved taxes. Thank you for accepting the committee’s invitation.

Before we move to questions, I invite Graeme Roy to make some opening remarks.

Good morning. Thank you for the opportunity to speak to the committee about our forecasts.

Overall, we estimate that the Scottish Government’s total budget next year is set to increase by £1.3 billion from the latest figure for 2023-24. That is a rise of 2.6 per cent in cash terms or 0.9 per cent after accounting for inflation.

Most of that increase in funding comes from income tax. A combination of the Scottish Government’s policy choices, high inflation, fixed tax thresholds in cash terms and relatively strong earnings growth in Scotland have contributed to a sharp increase in income tax funding. However, the Scottish Government’s spending commitments are also rising, with pressures from decisions to freeze council tax and increase public sector pay as well as our forecasts of increased spending on social security.

We forecast that social security spending will rise from £6.3 billion in 2024-25 to £8 billion in 2028-29. That is largely a result of payment rates being uprated by inflation and higher spending on disability payments, which reflects both a United Kingdom-wide increase in the number of people receiving disability payments and the effects of the Scottish Government’s policy and operational changes to disability payments.

We estimate that social security spending will be almost £1.1 billion more than the social security block grant adjustment funding that the Scottish Government will receive in 2024-25, with the gap rising to £1.5 billion in 2028-29. That gap is a result of Scottish Government choices on social security in the budget and previous budget decisions. That includes things like the introduction of new payments, such as the Scottish child payment, and the new approach to implementing social security benefits, such as adult disability payment. Those choices are designed to deliver a new approach to social security in Scotland, but they have consequences for the Scottish Government’s budget and future budgets, as we highlight in our report.

We are grateful to Social Security Scotland for the statistics that it has published on the Scottish child payment and the new disability payments as well as its on-going insights and intelligence gathering, which it shares. In particular, the statistics on adult disability payment have allowed us to begin to evaluate the assumptions and the effect on the overall budget of the changes that were introduced by the Scottish Government.

So far, there is no clear evidence from the data that we need to adjust our judgments on the level of successful applications or the average award level. However, we have modified our assumption around special rules for terminal illness, as there have ultimately been fewer cases than we had forecast. We will continue to monitor the data from Social Security Scotland and refine our forecasts as necessary.

Thanks very much, Professor Roy.

Much of the cost of uprating is covered by the block grant adjustment, as you indicated in your opening remarks. To what extent does this year’s uprating policy put pressure on Scottish Government budget choices.

That is a really good question. One of the things that we talk about in the report is the effects of uprating and inflation. Essentially, with high inflation, the key benefits get uprated by a higher number. Box 5.1 in our report shows that increase. The total effects of uprating by the end of the period are about £1 billion, after adjustments for higher inflation come through.

As the first part of your question mentioned, that uprating happens similarly across the UK, so the block grant adjustment protects a significant amount of that increased expenditure, because it comes from higher funding. As a rough rule of thumb, we think that about 85 per cent of that increase in uprating is matched by an equivalent block grant adjustment and the remaining 15 per cent is not covered by the BGA. That is where we would see the pressure coming through to the Scottish budget. Early on, 15 per cent is a relatively small number of increases but, as we get to the end of the forecast period, that number gets chunkier and bigger.

I will make an additional point. When you are uprating benefits for inflation, there is both a cumulative and a compounding effect of that uprating. If, for example, you uprate for inflation by 10 per cent in year 1 and by 6 per cent in year 2, that means that spending on benefits in year 2 will not be 16 per cent higher, it will be 16.6 per cent higher than it otherwise would have been, because you are uprating the first year’s inflation by the second year’s inflation. Long periods of persistent inflation can add up to very significant amounts of uprating in the future. That is why our forecasts further out for the effects of inflation tend to be very large. It is not just that there is large inflation further out but that the cumulative and compounding effect of earlier years of inflation is driving that.

How uncertain are the inflation forecasts?

That is another good question. There is a lot of uncertainty about the outlook for inflation, at least over the short to medium term. Over the longer term, we can be pretty confident that inflation will move back to the target of 2 per cent. By the end of the forecast horizon, inflation will be much closer to the target. The reason for that is that the Bank of England has a legal target to do that, so it will make that happen.

However, in the short to medium term, inflation prediction is really uncertain. In our report, we talk about the differences even in our forecasts over the past year, when we try to predict inflation. There are a couple of reasons for that. One is that a lot of the variability in inflation is because of things that are far outside the control of this place and even of the UK Government. There are global supply changes and pressures on energy prices, such as the war in Ukraine. Those are the big drivers that have been feeding through to inflation. Obviously, prediction is very uncertain, because we are trying to predict global events, so it is difficult. Therefore, in the short to medium term, there is uncertainty about the outlook for inflation. Over the longer term, we can be much more confident that it will come back down towards the target.

My final point comes back to your first question. To an extent, the block grant adjustment protects the Scottish budget from the uncertainty in inflation, because the funding flows through from that. Clearly, that is something that we have to monitor, and we update our forecasts as we get new inflation numbers.

There are two different aspects to the uncertainty about inflation. There is uncertainty not just about the level of inflation but about its persistence. The question is how far inflation is embedded in the system, through wage demands and pay increases, which tend to make it a more persistent level of inflation.

Roz McCall has a supplementary question on that.

Thanks very much. That is very helpful and informative.

The initial question was about the uprating policy putting pressure on the Scottish Government’s budget choices, and we have been very focused on inflation, which I understand. Is inflation the biggest concern when it comes to those pressures or is it the variance of the Scottish Government’s policy decisions?

There are a couple of things to say on that. If there were no differences in policies and, in essence, everything was exactly the same as what was happening in the rest of the UK, broadly speaking, the BGAs would be equivalent to spending here. The additional spending—where the additional money needs to be found within the totality of the budget—comes about, in essence, through the policy choices that the Scottish Government makes. We estimate that that figure will rise to £1.5 billion at the end of the forecast horizon. Roughly half of that is new social security spending that does not have a block grant adjustment—things such as the Scottish child payment—but the other half of that difference is where, in our judgment, the different approach in Scotland is likely to lead to higher spending relative to the equivalent in the rest of the UK.

A final thing that I should say about that is that we talk about that as a spending pressure and as a gap, from a public finance point of view, but it is clearly Government decisions prioritising certain elements of spend over others. Therefore, it is the nature of the funding that means that there is a gap there, but it is just about the choices that the Government is making.

If I may add to that, in addition to the pressures that are coming from inflation, we are seeing pressures coming from the rate of application for benefits. There is some interesting UK-wide data coming out that shows that the level of inflows into disability benefits, in particular, is far higher than the level of outflows. Therefore, the number of people getting benefits is increasing quite rapidly across the UK but also in Scotland.

I will continue the questions on inflation. I am also on the Finance and Public Administration Committee, so I realise that you might have answered this question before. Professor Roy, in your opening remarks, you said that the cash increase for the whole Scottish budget was 2.6 per cent and that that was a real-terms increase of 0.9 per cent. The difference is 1.7 per cent, so can you explain that figure of 1.7 per cent? We are putting benefits up by 6.7 per cent; can you confirm whether Westminster is doing the same?

Yes. My colleagues might want to come in on some of the specifics, but one thing to say is that I am talking about the total budget. One of the big things that are driving the change in the total budget is what has happened to capital budget, which is falling. The resource budget is going up by more than the total budget, so there is a cash increase of about 3.3 per cent on the revenue budget—the resource budget, which is where social security sits. The total budget is increasing by less because the capital budget is being cut significantly.

Another thing that might be helpful to consider—we have a table on this in the report, which I will try to find—is what happens to the total resource budget when we strip out social security and non-domestic rates income, because the remaining discretionary budget on resource is actually going down in real terms next year. That is the flipside of your question about what is happening to the total resource budget and what is happening to social security within that.

The table that I mentioned is figure 2.7 in the report. You can see that, between 2023-24 and 2024-25, taking out social security and non-domestic rates income, which are commitments that have to be paid for, the resource budget is falling in real terms next year. In essence, that is the effect of the squeeze on the overall budget of the commitments in social security.

The UK Government’s uprating for England and Wales is also 6.7 per cent—the same as in Scotland—so the same rates are being applied. That is why much of the increase in spending in Scotland is matched by the increase in the block grant adjustment.

Good morning, panel. Happy new year to you. You suggest that the cost of living crisis might be causing a short-term increase in applications for disability benefits across the UK. Will you explain further how that impact occurs and why you would consider it to be short term?

09:15  

To build on David Ulph’s point, at a UK level, there is a significant inflow of people into disability payments. We might come on to this, but the big driver of the overall increase in adult disability payments over the past couple of years relates to concerns about mental health and the prevalence of various other disabilities. That is matched by the BGA, because it is a UK trend.

We unpick part of that overall increase. Evidence shows that the cost of living crisis is potentially leading to an uplift in people claiming those benefits. That can be for a variety of reasons. It could be because the cost of living crisis is increasing concerns around disability and mental health and, therefore, pushing people into difficult positions, which means that they are more likely to claim those benefits. It could also be that, if people face difficult choices in their household incomes, they are more likely to claim a benefit that might be higher than another benefit that they might otherwise claim, so there might be an increase in people moving into disability payments rather than other benefits such as universal credit and jobseekers allowance.

To that extent, it can be a short-term impact, in the sense that people are making a short-term decision. We talk a bit about that effect, as does the Office for Budget Responsibility. In essence, it is a short-term economic cycle effect. I will make one caveat to that. Once people move into certain disability payments or move into a certain aspect of the benefits system, they tend not to come out quickly. There is a risk that, if more people move into claiming certain benefits, that can lead to a longer tail effect of any short-term effect that might come through from the inflow into disability payments.

I will elaborate on that point. We are not talking about a short-term effect on case loads. We are saying that the case load might increase but, as cost of living pressures ease, the pressures to apply for a disability benefit will fall back a bit. Therefore, the OBR says that case load growth rates will fall in the future as the cost of living crisis eases. It is not the level of case load that drops but the case load growth rate.

Thanks. I was trying to understand that. We are talking about non-means-tested benefits, so I was wondering why a cost of living crisis would drive up the cost of disability benefits. You also indicate that take-up of disability benefits through the form of ADP will be higher in Scotland. What is the case for that?

There is a different approach to delivering disability payments in Scotland. There is a combination of factors, such as the promotion of those payments, encouraging people to apply and a different approach to how the system engages with people—a more supportive environment for people, more help with the applications and less interrogation of the various requirements that people have. Overall, there is quite a different approach to delivering social security, which is the entire purpose of what the Government is trying to deliver.

From early on, our judgment has been that that will not only lead to a potential improvement in the service that people who are already claiming those benefits receive but make it more likely that people will apply who have been discouraged or put off from applying in the past or who have not even thought about applying. Therefore, we will see an increase in applications and less of an outflow at the end. That means that, all else remaining equal, we think that, for the adult disability payment, there will be a higher case load of people relative to the equivalent system that was in place before and that will lead to higher payments being made by the Government, which, in turn, means that there will be a higher public spending element.

To go back to my earlier point, it is a strategy—it is a policy decision to do it. It comes with a public finance cost, but then there is everything else that comes through, including improvements to tackling poverty and improvements to wellbeing more generally across society.

The lower rate of outflows that Graeme Roy was talking about is because of longer periods between case load reviews and lighter-touch reviews. Those reviews are not taking place yet; we are just seeing the additional applications for ADP. However, as that starts to come through, it will be another factor that is starting to drive the gap between ADP and personal independence payment numbers. That will be a factor in widening that gap more in the future.

Thanks for that. It is very helpful.

I invite Jeremy Balfour, who is joining us online, to ask a question.

Thank you, convener. I hope that you can hear me.

Good morning, panel. I am a wee bit confused about this, because the criteria are exactly the same for ADP and PIP—there is no difference between the two—so, surely, the awards in England match the awards in Scotland. Whether someone applies for PIP in England or for ADP in Scotland, the criteria are the same, so they should get the same decision. Are you noticing different decision making north and south of the border?

I think that I heard your question correctly. You are entirely right in that we are not saying that there is a change in the value of the payment or in the number of payments that people are getting. We are essentially capturing the fact that we think that the different approach to delivering social security in Scotland will encourage or support more people to apply for the same benefit than would otherwise have been the case.

Things such as promotion campaigns, encouragement to apply and a more accessible and supportive application process will mean that more people are successful in applying. Someone might start the process but then find it quite difficult or challenging and might give up on it and not follow through, but they will be more likely to be supported to complete that process in Scotland. There are also things such as changes to assessments. We know that people might be put off by a face-to-face physical assessment, and they are perhaps less likely to be put off if that is not a requirement.

In the report, we talk a bit about case load and we show the difference between ADP and PIP. We think that there will be an increase in the number of people applying and successfully receiving the benefit relative to what would otherwise have been the case. It is not that the benefit itself is changing but that there will be a higher case load of people applying and successfully securing that payment.

An interesting thing that we learned through talking to Social Security Scotland—it is one of the factors that Graeme Roy was referring to in talking about making it easier to apply for benefits—is that people can produce a wider range of evidence in support of their claim. We heard from Social Security Scotland that people think that the Government already has all of that information, so they turn up to apply assuming that the Government has a wide range of information available. They then find out that the Government does not have all the information, so they have to go off and acquire that information themselves. However, because they have made the initial application, that triggers the onset of the award. It may be that the award only comes along later, but it is backdated to the date at which they initiated the application. You might see a timing effect on payments because of that, but it will not affect the fact that the overall level of payments will be higher.

We have started to get the data through from Social Security Scotland on what is happening with regard to applications and approval rates for the adult disability payment. In figure 5.9 in our report, we show that, since adult disability payment was launched, looking on a comparative basis with England and Wales, even though, as we have discussed, there is a UK-wide effect—the number of applications in England and Wales is increasing—we can see, from the launch of ADP, that the number of applications increased far more quickly in Scotland. That supports our assumption that there will be more applications here.

We followed that through in the data, looking at approved applications as well, and, so far, we have not found any evidence that our assumptions about whether more people in Scotland will apply for PIP are wrong. Therefore, we are holding steady on those assumptions and we will continue to refine them as more data becomes available. At the moment, the assumption that more people will receive adult disability payment than would have received PIP seems to be holding.

That is really helpful.

I have a brief supplementary question to Mr Balfour’s question, which I thought was really important.

Clearly, ADP and PIP are two different things but they are being run similarly at the moment, with the same criteria. Professor Roy has outlined why we could have greater uptake here under the same criteria, depending on the culture and systems that we put in place to encourage applications. However, once the migration of PIP to ADP is complete, the Scottish Government will review the criteria, including, for example, the rule about a person’s ability to move 20m. The reason for not reviewing those criteria now is that, in Scotland, one qualifying household could be assessed under the criteria for PIP while another could be assessed under the criteria for ADP. Once all are assessed together under ADP, there will be new criteria.

Will the Scottish Fiscal Commission have a role in doing further modelling work to see what the additional benefit would be to disabled households in Scotland and what the additional costs would be in budgetary terms? Would you expect to be actively involved in that?

Yes, you are right. When we are trying to make a judgment about social security spending—for example, on ADP—there are two elements that drive it, aside from demographics and the stock of the population. First, there are the judgments that we make around eligibility, take-up and the number of successful applications. In essence, that is what we have been talking about so far—it will be the same system but delivered in quite a different way, which we think will lead to an increase in the number of people who successfully claim those benefits.

The next bit is about what would happen if policy around ADP were to change. You are right in saying that that could be about eligibility criteria; it could also be about the value or length of payment, for example. Once that starts to change, we will start to model that as part of the budget process and our policy costings. If the Government revised its policy—let us say that the 20m rule changed—we would estimate who would benefit from that and what the fiscal implication might be. That is where we would come in.

Thanks for putting that on the record.

It is not our role to make judgments about policy objectives and the extent to which those are being delivered; our role is simply to assess the fiscal consequences of whatever policies are implemented. However, we will certainly update our models and our thinking as soon as the Scottish Government reforms aspects of the ADP system.

That is useful clarification, Professor Ulph. I absolutely get that and would never try to draw you on the merits of policy intent, but politicians have to make informed decisions on costings, because there are lots of things that we would like to do but we always have to be able to afford them and budget for them—not just from year to year, but in the longer term. That is why this evidence session is particularly important.

On the baseline data, the Scottish Fiscal Commission notes with regard to the adult disability payment that

“the average award level for new applications has not yet stabilised”.

By “average award level”, do you mean the number of award applications that are successful, the level of award or a combination of both? More importantly, when would you expect that to stabilise? Would that become more baseline data on which you would advise the Scottish Government about modelling work for any future changes?

09:30  

You are right. It is difficult to explain and understand where we are. With the roll-out of the benefits, we are currently a wee bit in the dark, because we do not know what the long-term sustainable trends in the data will be. It is the first time that we are doing this, and we are impatiently waiting for the data to come through so that we can say whether our judgments are right or wrong and where we need to look at things differently.

As Michael Davidson said, we track the application rates and look at them to see what is happening. We look at changes in inflows and outflows and the like, and the average payment is crucial to what we do. Once we have a stable average payment, that becomes fundamental to pushing things forward. When we know what the average payment is and what the average inflows and outflows are, we can be pretty confident in our forecasts.

We do not yet have enough time series to be really confident about that average payment, however. What we know so far from the data is that the average payment that we have been tracking and the number of successful payments coming through have been quite volatile. Early on, when the system was rolled out, there was a significant increase in the average payment relative to what had gone before, but that figure has come down a bit—it is not as high as it was. We can speculate about why that might be the case. It might be that prioritisation by Social Security Scotland to get to the people who needed the most support early on meant that the average payment was higher. It is coming down now, and we just need a bit more time to see more data points before we can have more confidence in the average payment level, because it will not be as volatile as it is just now.

You have probably guessed the final part of my question. You say that you need a bit more time. The Scottish Government is committed to reviewing all of this in the round, but it cannot do that unless it is confident that it has robust baseline data. Quite rightly, you need a bit more time, but the committee needs to plan how we wish to scrutinise all of that. What do you think “a bit more time” looks like? When will things have stabilised so that we have that baseline data?

In many ways, it is the same as asking, “How long is a piece of string?” As we start to move into years 3 and 4, we should have a lot more data points than we have at the moment. There has also been a need to work through the backlog, and we have seen the progress that Social Security Scotland has made over the past few months in reducing that backlog. Again, that is really helpful, because it gives us more data points.

Over the next year, the average payment should start to stabilise, and, at that point, we can start to get more confident. I do not think that there will be an exact date when we can say, “Now, we are confident.” It will be a gradual process of having more data and being clearer.

I apologise, Professor Roy: I am almost treating you as though you are a politician, which you are absolutely not. If I had asked this question of a minister or a cabinet secretary, invariably my colleague would have said, “The financial year 2024-25.”

Once Social Security Scotland gives outturn figures for all of that and they can be verified, validated and analysed, are you good to go on this? If so, what is the timescale for that? We would ask that question of the politicians, but the politicians would have to ask others. I feel as though we are asking the experts directly.

Do you want to come in, Michael?

Things will be a bit more stable once the transfer of PIP cases from the Department for Work and Pensions is complete. The transfer should be completed in 2024-25, but there will still be a lag before the data is available. Social Security Scotland just published its latest statistics—I think that it was on the same day as the budget, on 19 December—and we currently have adult disability payment statistics up to the end of October. Social Security Scotland is actually quite up to date with the case load and application statistics, but we would still expect the data to be available only after the end of 2024-25, once things have settled down a bit more, with the completion of case transfer.

There are different trends going on. There is the trend that Graeme Roy referred to—that the early cases were perhaps the most clear cut for decision making and that those awards tended to be the highest, which is a factor that drives the average award down. We also have the issue of lighter-touch reviews and longer periods between reviews, which I referred to earlier. We will not start to see the impact of that on the average award for a while, but that could have the opposite effect—of driving up the average award—if the early awards were high and are prolonged. There are factors leading in different directions, which is partly why it is hard for us to say exactly when the average award will stabilise and settle down.

I do not have any more questions. That was really helpful.

I am looking at the overall picture of the additional spend of £1.092 billion that we have to find. I think that I am right in saying that roughly half of that is the Scottish child payment, which is completely ours because they do not have that in the UK, and that another big chunk is the extra money for ADP. Is that correct? Are there other factors in there as well?

At a macro level, in terms of the chunky numbers, that is pretty accurate. When we think about what is driving social security spending—what is driving the increase that we see happening over the next few years—there is the chunk of stuff that we have spoken about already—trends at the UK level such as inflation, the big rise across the UK of people inflowing into disability payments, the link to mental health, the legacies of Covid and the like—but that is all matched by the BGA. Those factors drive the increase in social security spending, but there is an inflow in funding.

Then we are left with the bit of the budget that is not matched by an inflow of BGA funding. Roughly speaking, you are correct in saying that there is a chunk of things for which there is no equivalent funding flowing in from the UK. That includes things such as the Scottish child payment, which is unique to Scotland—there is no equivalent benefit in the rest of the UK. The rule of thumb is that that is roughly half of the funding differential.

The remainder is the judgments, which we just spoke about with the deputy convener, about the additional approach that is being taken in Scotland to deliver benefits equivalent to those in the rest of the UK. Those judgments are leading to a higher inflow, which accounts for the remaining half.

The two big, chunky bits are, first, the Scottish child payment and, secondly, the adult disability payment. There are then some smaller elements to add, including child disability payments, which we think will run ahead of the BGA. There is also no BGA equivalent for things such as discretionary housing payments, so the stuff around the bedroom tax and carers allowance supplement require a small amount of money—typically in the region of £100 million rather than the £500 million or so required for the Scottish child payment.

So, they are kind of rolled into the Barnett formula.

Some of them are rolled in, but we separate them out—perhaps Michael Davidson could find the relevant table. Some social security payments, such as the Scottish child payment, have no equivalent funding at all, whether BGA or Barnett; some elements are rolled into the block grant, but they are very small in the scheme of things—I think that they account for about £100 million overall; and other bits are matched by BGA funding, which gives us the net position. In figure 5.8 in our report, we separate out the three different elements, which add up to £1.1 billion. About £368 million of that is the net position; £600 million is the new social security payments—including the Scottish child payment—that do not have a BGA equivalent; and about £100 million is those other things, some of which are wrapped up in Barnett.

As you said in answer to previous questions, you are assuming that there will be no change in policy and that both the UK and Scotland will carry on with their current policies. I presume that our figures could be affected by either of those changing. If the UK Government changes its policy, we will have more—is that right?

Yes—exactly. On policy and payments, it is a policy in Scotland to implement a different system, to encourage take-up and support people into it. It is a discretionary policy decision by the Scottish Government to implement a different type of social security system. Even if the payments and eligibility are not changing, it is still a different policy, because you will get more inflow into the system.

There is a UK policy and a Scottish policy, but you are assuming that both of them will carry on much as they are.

Exactly. Changes in policy at a UK level could, in time, feed through to the BGA. Changes to eligibility for universal credit could have a knock-on implication for eligibility for the Scottish child payment. If the UK Government were to reform PIP in the future, that would have an implication for the BGA, which would then feed through to how much funding was flowing into the budget. If the UK were to change its PIP policy, ADP policy would not change automatically. The funding through the BGA going into the overall budget would change, but it would then be for the Scottish Government to decide how to respond, or not respond, through its ADP policy.

It is often said that, for both the UK Government and the Scottish Government, this is a demand-led area and that we cannot control it very much.

I will let Professor Ulph come in in a second.

You can control it in the sense that you set the policy and the eligibility criteria, and you then rely on the demand through people flowing into the payments. You could, ex post, change the eligibility criteria and payments if you wanted to, but, once you set the policy, for the next months you are relying on the take-up of those benefits. That is quite different from traditional devolved spending such as in health, whereby you choose to give X amount of money to a health board and X amount of money flows to that health board. Here, you set a policy and eligibility criteria and you rely on the demand of people flowing into that. That is why it is quite a different way of managing public finances and spending relative to other normal devolved expenditures.

I will elaborate on that point. In other areas of spending, such as health and education, the Scottish Government and Parliament approve overall levels of funding and other people then split that up into, say, funding for universities and funding for schools. Every university and school gets a budget out of that, and they decide how that will be spent—whether they will fill a post by appointing a geography teacher or by appointing a physical education teacher. The real implications of that are felt at the end of a chain of decisions by a lot of individuals. Parliament sets an overall number and the implications flow through at a lower level.

It is almost the other way round with social security. You set the eligibility criteria and Social Security Scotland interprets those eligibility criteria to say who will get the award. You set the amount and say what evidence people will have to provide in order to get the award, and all of that gets translated into an overall spending number that emerges from a whole series of decisions. You do not choose that spending number; it emerges from a lot of decisions by other people. You can control that, because you can change some of the initial decisions, but in doing so you are changing things that are really conspicuous to individuals. You are saying, “We are changing your eligibility criteria,” “We are changing the amount of money you are going to get,” or, “We are changing this aspect of how you can apply.”

The implications for individuals are much more immediate when you try to control social security, whereas if you change health spending, people say that health spending has been cut but they do not see the implications for their own general practitioner visits for a while. Yes, it can be controlled, but, politically, you are facing a different set of issues because of the direction of flow between the spending of money and the ultimate implications.

09:45  

I certainly find it helpful to have had the issue laid out in that way.

This is my final question. We have looked at the split of the £1,092 billion. Would I be right in saying that the split will be changing a bit? At the moment, the Scottish child payment is about half the spend, but that will not be the case in future. Will you explain why that is the case?

Essentially, we think that, as we move toward the end of the forecast horizon, the effects, particularly of the different benefits in Scotland and the different inflows, will start to have a gradually greater impact, as the applications and case loads increase. Therefore, the block grant adjustment will take up a slightly bigger share of the gap.

On the Scottish child payment, we assume that the numbers will broadly remain relatively flat and will be relatively stable, with the payment being uprated in line with inflation. However, there will be growth in the numbers of those receiving adult disability payment and child disability payment, which will lead to higher levels of expenditure.

I invite Jeremy Balfour to come in.

I do not have a supplementary question to ask at this point, convener.

I believe that you were going to ask question 7 in our paper, Jeremy.

My apologies, convener. I do not have that in front of me. Could someone else ask it?

I invite Bob Doris to do so.

I know what the convener is hoping to get on the record from the Fiscal Commission. I will ask that question and then put my other one.

By 2028-29, the Scottish Government expects to spend around £1.5 billion more on social security than it receives in funding through the block grant adjustment. The whole forecast, of course, is based on the best evidence that you can get. How uncertain is that figure?

There are two aspects to that, and it builds on a bit of the conversation that I have just had with Mr Mason. We can be relatively confident about an element of the spending. For example, on the Scottish child payment, aside from future policy changes, we can be relatively confident about the amount of spend on that. That benefit does not have an equivalent BGA. Largely, the spend is to do with the number of eligible children who are eligible via universal credit, so it is relatively straightforward to forecast.

However, there is greater uncertainty and greater potential variability on the second element—the social security funding differential that we are talking about—as the Scottish Government is implementing a new approach on the adult disability payment. There is great uncertainty around that. We started to make a judgment on that in May 2022, I think, when we said that we thought that the different approach would lead to a difference in funding.

We are very confident—this can be seen in the data that is available—that there will be a gap relative to the funding that is flowing in. In many ways, that is entirely what you would expect. You would not implement a different system that was designed to encourage and support more people to claim the benefits, and to make the system more supportive, and then not expect more people to be successful in their applications. Therefore, there will be a gap. The uncertainty is about how big the gap could be.

Essentially, there is uncertainty in relation to about half the funding gap. As I said, there will be a gap, but—this goes back to your other question—we need to bit more data to see what that is.

As Michael Davidson said, we look at all those things as we go along. We thought that there would be a spike in spend when ADP was introduced, and that happened. We also thought that there would be higher levels of ADP relative to PIP after its national launch, and that has happened. The data that we are seeing is consistent with our judgment calls; we just need to wait for the exact number to determine whether the numbers need to be revised.

Does the Fiscal Commission build in a tolerance level in relation to that £1.5 billion? I absolutely get the point—your forecast could be dramatically blown off course by a significant policy change at a UK or a Scottish level, or by an unforeseen event. Does the Fiscal Commission build in a 5 per cent tolerance or a 10 per cent tolerance? I do not even know whether that is a thing, as this is not my area. If your forecast is £1.5 billion, would you say that that could be 10 per cent higher or 10 per cent lower, based on a range of judgments? If so, what is the tolerance level around that?

We do not have a specific tolerance of 5 per cent or 10 per cent. That is partly because the nature of what we are tasked with doing is such that it involves coming up with a specific number to help the Government to set its budget. However, our forecast evaluation report, which comes out every year, is crucial, because that is where we look back and unpick all the different elements. We set out which element we think will come through more successful applications, which element will come through terminal illness and which element will come through average payment. You can do that sort of work yourself. We then look back at where things have changed and where there might be variation in that. We do not have a rule of thumb, but we show our working so that anyone can look at the potential level of uncertainty.

I have one point to pick up on. We started to add in forecasts for the different approaches to disability benefits in the run-up to May 2022, but it was in May 2022 that we first did it for all benefits, including those for pensioners as well as the working-age adult benefit. The first time that we did that in full for adult disability payment was in August 2021, when we provided quite a lot of information along the lines that Graeme Roy mentioned, with the different breakdowns.

When we first produced our forecasts for ADP, there were 12 different areas of reform that the Scottish Government had introduced in relation to how ADP was going to be delivered. Because there was no evidence at all, we had to make judgments on each of those 12 elements. As Graeme Roy said, although we cannot say that the judgments that we made on the things that have come to pass so far were definitely correct, we have not seen anything that says that they were very far out, except in relation to terminal illness.

What we have not yet seen are the implications of our judgments on lighter-touch case reviews and longer periods between case reviews. The implications of that are still dependent on the initial judgments that we made two or three years ago. That is where the area of uncertainty still remains. It is a judgment-based forecast. However, that does not mean that we just made the numbers up. That judgment was based on lots of discussions with experts and other people in this area. That is why some of our earlier judgments have turned out not to be very far off the mark.

I love your comment that you do not just make the numbers up. I am glad that you put that on the record.

Convener, I had another question, as you know, but I realise that you need to manage the time. Perhaps I could come in if there is time at the end.

I am conscious of that. I think that you touched on the subject of Roz McCall’s next question. Would you still like to come in, Roz?

Yes, I would. Bob Doris keeps doing that to me. This question follows on from the one that he asked.

I am very interested in forecast error. My question touches on what you have said. I appreciate the information that you have provided. According to the information that you have given us, we are looking at a forecast error of 3 per cent, or £127 million. As has been mentioned, as time goes on and the bill increases, a 3 per cent variance becomes a lot of money. Can we continue to expect to have a 3 per cent forecasting error? Moving forward, we really need to have a plus or minus figure in our thought processes as we look at budget forecasting.

The 3 per cent figure relates to 2022-23. The one thing that drove that figure to be slightly higher than you would expect is that, when we talk about error, we talk about the difference between what we said at the time of the budget and what transpired, and that can change for error in our forecasting, but it can also change for error because policy changed. That was one of the big drivers of why that happened in 2022-23.

For this year—2023-24—we have not had the same in-year policy changes, and we have not had the legacy effects of Covid. At this moment, we think—touch wood—that that error will be smaller this year.

One of the things that I think we have to factor in is that, in time, we will get a better time series for the average forecast error. Again, we just have to get more data points. We have the 3 per cent figure for 2022-23 and we will soon have the figure for 2023-24. That will give us another data point that we can look at and interrogate.

More broadly—and this is a fundamental change with devolution, and it is the same conversation that we have with the Finance and Public Administration Committee on tax—this is about a fundamental difference in how the Scottish budget now operates. In the olden days, you had Barnett, you got that money and then you spent it on devolved elements of spending. You now have a forecast for income tax that will change and be volatile and the Government has to manage that. Social security is exactly the same. We make a forecast, but it is volatile, and the Government has to manage that risk.

To the extent that the risks are shared with the UK, the BGA will act as protection. If there are any differences in the risks between Scotland and the rest of the UK, those risks will now sit with the Scottish budget. That is just the nature of the really difficult change in how the Scottish budget will now work.

I will just make sure that I have got this correct in my mind. For every policy variance that the Scottish Government puts forward and every change that it wants to adopt, there needs to be cognisance of the insecurities that follow on from that as far as the budget process is concerned and we need to be aware that there will be an additional effect based on the Government’s policy decisions.

Essentially, there are two big risks. There is the risk of there being no differences in policy and Scotland just being different from the rest of the UK; secondly, the differences in the policy choices might lead to their own risks. That would come through as well.

That is excellent. Thank you very much indeed.

I am going to bring in Paul O’Kane, but, before I do, I have a quick question. Please indulge me here. It is more of a technical question about the international financial reporting standard 16, which is a new standard in relation to leases. Your report mentions that, at the moment,

“in the public sector, IFRS16 is being adopted progressively”

For our benefit, could you put in layman’s terms how IFRS16 impacts on resource funding and capital funding?

The IFRS is largely a technical accounting definition of the budgets and where things score and where things are reported in the budgets. It does not change the outflow of what the Government is spending; it is largely a way of reporting that. Essentially, the change is bringing the Scottish Government budgets in line with good process for how budgets are reported. In our report, we talk a bit about how the numbers would look if we used the new IFRS system and how the numbers would look if we used the old system, but it does not change the actual spend. It is about how it is reported.

That is great. Thank you very much for that.

I am interested in understanding more about the costings for new policies that will come in this year. However, my first question is about the impact that delays to new benefits have on forecasting. For example, the delay of carer support payment from spring to autumn has been widely discussed and we have spoken about it in the committee. To what extent does that have an impact on the forecast?

10:00  

It does not impact the totality of spend, because everything is backdated. It might have an impact at the margin if it changes year on year. If you had people who did not receive a payment in 2023-24 but got it in 2024-25 and it was backdated, that would uplift the expenditure in the year in which it took place. We have thoughts about this, and the one caution that I note is that, at this stage, we do not think that it is having a material impact on our public finance forecasts, because we are dealing with a relatively small number of inflows rather than the totality, and it is a small percentage. Given that it is a relatively small number of a much bigger number that drives the overall totality, it does not have an impact on our forecast. In addition, we have seen quite a drop in that over the past financial year, which in turn means that, from a public finance point of view, it does not really have a material impact on our forecasts.

That is helpful. There is a degree of uncertainty around the new benefits that are coming on stream this year, particularly the pension-age disability payment, which will replace the attendance allowance. You have said that costings remain highly uncertain in that space. To what extent has your experience of forecasting other disability payments informed your costings and assessment of the pension-age disability payment?

Perhaps Michael Davidson will come in on some of the detail about how the judgments that we use on things such as child disability payment and ADP and the data that we track feeds through into that, and perhaps he will talk a bit about the process that we go through with costing.

With the pension-age disability payment, our starting point is attendance allowance, which is the benefit that it is replacing. We forecast on the basis of the case load, which is the number of people getting the benefit and the average amount that each person gets. We have looked at the changes that have been made in promoting the benefit and encouraging more people to make applications, and we think that that will have an effect on the number of new people flowing into it. In our models, we have ways of forecasting what we call monthly inflows, which are the new people starting on the benefit each month, and we make adjustments upwards to reflect what our judgment is on those increases.

It is helpful that CDP and ADP were mentioned in that context, because we made similar judgments for CDP and ADP, and we have been starting to get the data through on those benefits to verify those judgments. So far, we have not seen evidence that we need to change our judgments for CDP and ADP. We are quite happy to make similar judgments for PADP, even though the client group is quite different, because it is a pension-age client group. We still think that, just because of the promotion and the help that people might receive in applying, there will be more applications and more people will be successful in receiving PADP compared with the situation if attendance allowance had been continued. That leads to a gap versus what we expect the BGA funding to be.

I have a point about the uptake of ADP and the increase in the number of people receiving the benefit. You have said that analysis is under way. Do you intend to share that more widely to inform this committee’s work?

In the report, we have already started to look at and talk in a bit more detail about that. In figure 5.9, we show the number of applications, but, obviously, that is the first stage in the process. Further down the line, because of the time of processing, we plan to continue to build up that picture in future reports. We will definitely build on that in May.

I believe that Marie McNair wants to come in, and then I will invite Jeremy Balfour.

Thank you for your indulgence, convener. Previously, I asked the Scottish Fiscal Commission about the difference that a more generous and quicker approach to terminal illness is making to the Scottish budget. Obviously, it is important to get that right for the claimant. You touched on that in your opening comments. Do you have any further observations that you can give to the committee?

Michael Davidson might want to go into some of the detail around that. The broader point that I was making about terminal illness is that we make an assessment of the number of claimants who might be coming through in relation to terminal illness. We made an assessment of the number but, looking at the data that we are getting so far from Social Security Scotland, we think that that was higher than the number that has actually come through, so we have adjusted our number accordingly. That feeds through to our forecasting.

I do not know whether Michael Davidson wants to explain a bit about the data we are tracking that shows that.

In published statistics, we can see the number coming through under the special rules for terminal illness. As Graeme Roy says, it is below the level that we had in our forecast previously. Based on the wider evidence, we still think that the level that we thought it would get to by the end of the forecast period will be the same, but the rate at which it will get there will be a bit slower—it will just take a bit longer to build up. That means that, in our forecast for 2024-2025, that reduction and adjustment in the profile has reduced the forecast by £27 million. The reduction reduces, if that makes sense, in later years, so it gradually builds up.

Are you aware of any reasons why that is the case? Is it to do with promotion?

This is more on the speculating side, because it is difficult to find hard data, but it could be the way in which the process interacts with people applying through the normal rules. It could be that people who might have become eligible under the terminal illness rules at the moment actually qualified earlier under normal rules. It is not a separate group of people coming in—that interacts with the rest of the case load. It is an area that we will continue to look at and try to understand, but it is difficult to see a specific reason at the moment.

Thank you.

I invite Jeremy Balfour to ask our very last question.

I have two questions. I would like to follow that up by asking whether the data that you are getting from Social Security Scotland is sufficient. Are there areas that you think we should be recording that have not been recorded as yet?

I will come in and then Michael Davidson might have something specific to say on that.

When we spoke to the committee maybe 18 months ago, we were saying that we needed much more data from Social Security Scotland and we had concerns about our ability to do the forecast. It is good to put on the record that the data and the support that we receive from Social Security Scotland are excellent. We have had a lot of the data that we are now seeing. To come back to the question from Bob Doris, we are now getting data points coming in and we are tracking applications and average payments. We just need to wait a bit longer for that. We are getting a lot of that data from Social Security Scotland.

We also get lots of really useful and helpful intelligence from Social Security Scotland. The agency is able to tell us about what has happened to backlog, what applications it is seeing and what it is doing on promotion. Again, that is helpful for us, because it means that we can take judgments about where we are on that.

I do not know whether there is anything specific that we are still waiting on, Michael.

No—it is more about looking to the next benefits that are being launched. In particular, the pilot has just started for the carer support payment. Obviously, we are keen and hopeful that Social Security Scotland will be able to provide the data on that in a timely manner. Following that, pension-age disability payment will be launched, so that is in the future.

The one area in relation to adult disability payment that we mention is the award review data, but that is because that part of the process is still in the early stages. We are hopeful that we might have data on that to inform the next forecast, which we expect to be in May.

Thank you. I will go on to my final question. What are the main elements of uncertainty in the costings for pension-age disability payment and pension-age winter heating payment? For the record, I am in receipt of the higher rate of PIP.

The pension-age winter heating payment is probably one of the points where there is much less uncertainty, because it is largely just about demographics and the number of people who are eligible. Because the benefit is universal, the forecast is basically about how many people are eligible.

There is more uncertainty around the pension-age disability payment because, as Michael Davidson said, we have to make a judgment about what the uptake might be and how many successful applications there might be. Again, I avoid the temptation to put a specific number on it, but it is likely to be less uncertain than ADP and CDP because of the nature of the demographics. There is a judgment that we have made in the report, but it is likely to be less uncertain than what we have done for some of the other benefits.

Thank you.

As no one else wants to come in, that concludes our evidence session and our public business for today. Several committee members are going to visit Social Security Scotland in Dundee on Monday, so I feel that we are now armed with a lot more information in order to ask pertinent questions. I thank the witnesses very much for your attendance today. We will now move into private session to consider the remaining items on the agenda.

10:11 Meeting continued in private until 10:34.