Finance and Public Administration Committee

Meeting date: Tuesday, January 11, 2022

Official Report 605KB pdf

---

Fiscal Framework (Independent Report)

Good morning, and welcome to the first meeting in 2022 of the Finance and Public Administration Committee. I wish everyone a happy new year. We are meeting remotely today.

The single item on our agenda involves taking evidence from two panels of witnesses to inform our consideration of the forthcoming independent report on Scotland’s fiscal framework. Although the terms of reference and authorship of the report have yet to be agreed by the United Kingdom and Scottish Governments, we know that it will focus on block grant adjustments.

First, we will hear from Professor David Bell, David Eiser and David Phillips. David Phillips will join us a little later, at about 10.15 am. Our second panel will focus on arrangements for the Welsh fiscal framework approach to block grant adjustments, and Dr Ed Poole and Guto Ifan from Cardiff University will join us.

I welcome Professor Bell and David Eiser to the meeting. We have about 70 minutes for the discussion. I remind members that broadcasting staff will operate their microphones, so please allow a few moments before speaking to ensure that you will be heard. Members who wish to ask a supplementary question and witnesses who wish to respond should type R in the chat function, and I will bring them in.

I will begin the questioning. What should the independent review cover, and what should be the terms of reference for the independent report? Perhaps Professor Bell can go first.

That is quite a big question to start with.

Take your time. [Laughter.]

The independent review clearly has to look at how the existing arrangements have worked, how they have been gradually introduced and the experience of the new fiscal arrangements. That comes down to a number of issues. It seems to me that a central focus will be the block grant adjustment, how it is dealt with and whether there is any possibility of finding alternative mechanisms that might be more satisfactory. That said, it is fair to say that people have racked their brains and not come up with anything that is satisfactory to both the UK and Scottish Governments and which meets the Smith commission principles. That is an empty set. The block grant adjustment and alternatives will certainly be a focus of the review.

Other issues include the transparency of the whole system. It appears to me that, although the Scottish Parliament and the Scottish Government are now in possession of very significant tax and welfare powers, that is not well understood by the Scottish people. It would be good to have transparency and a better understanding of how the whole system works. That said, it is very complicated—it has been said that, unless you are called David, you will have difficulty understanding it.

Another issue is how we improve the availability of relevant data to make forecasts of the block grant adjustment and of tax and welfare receipts. Another issue relates to intergovernmental negotiations on the fiscal framework and how they can take place more harmoniously and regularly.

Borrowing is another area in which difficulties have been encountered. There have obviously been difficulties associated with Covid, but there have also been difficulties relating to reconciliations and forecasts that potentially stretched Scotland’s borrowing capabilities. That needs to be looked at.

I do not know whether that is a long enough agenda for the review, but it would be quite an extensive agenda. Given that there is a very tight timetable if the review is to be completed by the end of the year, a lot of work will be required from all the relevant parties. It remains to be seen whether a satisfactory outcome can be arrived at within the time that is available.

Does David Eiser want to comment on the question that I asked? What should the independent review cover, and what should be the terms of reference for the independent report?

As you have said, the Government has agreed that the independent report will focus specifically on the block grant adjustments for devolved tax and social security. In that context, the independent report will need to articulate clearly the different roles that the block grant adjustments play. It will then need to identify a range of potential block grant adjustment mechanisms that could be used. Those include the index per capita method, which is used at the moment; the comparable method, for which the UK Government has always expressed a preference; the Welsh model, which adjusts the comparable method for different tax bands; and various other methods that have been mooted that control for demographics and other fiscal risks.

The report will then need to set out the risks and advantages of each of the different approaches to block grant adjustments. It will need to show, to both the UK and Scottish Governments, the risks and advantages of the different methods through clear illustrations and the use of scenario analysis.

The report is being commissioned jointly by the two Governments. I hope that there will be a publicly available interim report, and the final report should be publicly available, too. I hope that there is scope for input to the independent report from the public, experts and so on as the process goes ahead.

Of course, the report sits alongside a wider review of the fiscal framework, and David Bell has picked up on many of the issues that the wider review will need to consider. At a minimum, it will need to consider the adequacy of the Scottish Government’s existing tools to manage forecast error risk, the forecast error borrowing powers and the Scotland reserve drawdown limits. It will need to cover the capital borrowing limit, and it is important that issues relating to dispute resolution are considered. It will also need to consider the extent to which the Scottish Government should have additional budget management tools beyond those that it currently has. There could be scope for borrowing powers to deal with not only forecast error but other budget uncertainties.

I reiterate what David Bell said: there is a lot to be covered in the review as a whole. We know that the independent report will focus on the block grant adjustment mechanisms, and there is a lot to cover just in that report.

Yes, there is. Before I touch on the block grant, I will follow up what you said. You talked about the adequacy—that was the word that you used—of the fiscal tools. Do we have adequate fiscal tools in terms of borrowing limits, for example? Should the limit be raised? Should it be linked to inflation? Where do the limits need to be increased to give more flexibility for Scotland?

Currently, the Scottish Government has different sorts of limits for different sorts of borrowing. One of the obvious issues is that the capital borrowing limit is fixed in cash terms. There is no obvious reason why that should be so. As a minimum, the review should consider how that is indexed over time. Should it be indexed to some measure of inflation or the size of the Scottish budget? However, there is then the question of the extent to which the capital borrowing limit itself is adequate.

On forecast error borrowing, the current limit is £300 million a year. In the short space of time that we have had the new fiscal framework, we have already seen one year in which the reconciliation amount was more than £300 million. That said, it is early in the process so it is difficult to know whether that was a fluke one-off or whether reconciliations will regularly exceed £300 million.

There is some analysis by the Scottish Government and the Scottish Fiscal Commission that suggests that forecast error borrowing might need to exceed £300 million fairly regularly. On the basis of that analysis, it seems that there is probably a pretty strong case for that forecast error borrowing limit to be extended. How much it should be extended is a tricky question. At the moment, it is temporarily extended to £600 million anyway because, in 2021, a Scotland-specific economic shock was triggered. Over the next three years, that temporarily raises the limit to £600 million. That amount probably feels adequate at the moment and there is perhaps a good case for saying that it should remain permanent rather than being a temporary limit.

There is a third issue, which is the extent to which the Scottish Government should have the ability to borrow to fund discretionary spending. It cannot do that at the moment. In a recent report that David Phillips, David Bell and I produced—“Options for reforming the devolved fiscal frameworks post-pandemic”—we argued that there was a good case for the Scottish Government to have some ability to borrow modestly to fund discretionary spending to deal with unforeseen events that arise during the year. It would be sensible if the fiscal framework review covered that as well.

I echo what David Bell and David Eiser said about the scope of the independent report and the review.

On the degree of flexibility for the Scottish Government in capital borrowing and discretionary resource borrowing, it is important to recognise that, at the moment, there is no specifically English borrowing in the system. If the UK Government borrows, it borrows to fund UK-wide spending on, for example, defence, pensions or debt interest payments or to fund English measures that lead to Barnett consequentials for the Scottish Government and the Welsh Government. That means that any borrowing by the UK Government automatically gives money to Scotland.

To the extent to which additional Scottish borrowing powers are used, that is on top of UK borrowing. Of course, the Scottish Government pays that borrowing back, so it is not free money, but the extent to which those powers are made use of suggests that they are seen as worthwhile and valuable.

10:15  

One of the reasons why we say in our report that any powers on, for example, discretionary resource borrowing should be modest is that making them substantial or moving towards, for example, a prudential borrowing regime for capital would mean that we would need to think about the equity issues across the UK, given the fact that those powers would not exist at an England level. We could say that that is not a Scottish issue but something for the UK Government to decide for England. However, how such extensions would play out against the fact that there is no England-only borrowing regime at the moment is one of the factors that we thought about in our report.

David Bell, you and the other two Davids produced a paper in November 2015, in which you said:

“it is impossible to design a block grant adjustment system that satisfies the spirit of the ‘no detriment from the decision to devolve’ principle at the same time as fully achieving the ‘taxpayer fairness’ principle: at least while the Barnett Formula remains in place.”

We have moved on more than six years since that paper was published. What are your thoughts now on trying to develop a fair block grant adjustment system? Has it become easier or more difficult and more complex?

My view is that those sentiments have not changed, and I think that the same is true of the other two Davids. It is still difficult, if not impossible, simultaneously to satisfy all the Smith commission principles. The eventual decision around the comparable method was a compromise between the Scottish Government’s position and the UK Government’s position.

None of the methods will benefit the Scottish budget overall if there continues to be a relatively long-run decline in income tax revenues. The key element is income tax. Although the other BGAs are important, they are less important quantitatively.

The key issue for Scotland is the indexation method that is used. There are significant differences. The difference is in economic performance, which has distributional issues. It is not just about overall economic growth; the really important point is how tax revenues grow in different parts of the income distribution. There will continue to be issues around BGAs that are somewhat larger than the tax revenues that Scotland generates and, in consequence, a downward pressure on the Scottish budget.

Again, I agree with what David Bell said. It might make sense to explain a little bit why we thought the principles were in conflict and then say a little bit about what that might imply looking forwards.

The no detriment principle was the idea that there should be neither cost nor benefit to Scotland from the act of devolving tax revenues to Scotland. At the point of devolution, you devolve £10 billion-worth of revenues, and you take £10 billion off the block grant so that there is no net effect—just devolution. What matters is the performance after devolution.

It was our view, and I think also the view of the Scottish Government, that a reasonable benchmark after that point would have been to say that if Scottish revenues per capita kept pace with those in the rest of the UK, there should be no detriment from that either. You would get the same percentage growth rate in revenues per capita. That is a good benchmark against which to make an assessment. If we do better than that, we gain; if we do worse, we lose. However, if we can keep up, we should not lose out.

We said that that was in conflict with the taxpayer fairness principle because the Scottish population is growing less quickly than that in England. If you say that we will ensure that Scotland does not lose out from devolution as long as its revenues keep up per capita, but the population is growing less quickly, in order to achieve that you need to transfer some revenues from England to make up for Scotland’s slightly slower population growth. That would mean that you would not be satisfying the taxpayer fairness principle, which says that, once taxes are devolved, the revenues should stay in the countries in which they are raised. That is why we said that there is a fundamental conflict between the different principles.

That means that we need to think about which principles will be priorities. My view is that, if there is a fiscal union between Scotland and England, it makes sense to have some degree of fiscal transfer from areas of faster revenue growth to areas of slower revenue growth. I would therefore probably de-prioritise the taxpayer fairness principle and would prioritise the no detriment principle, which is what currently happens. While that conflict exists, because a fiscal union will involve transfers between different parts of the country, the taxpayer fairness principle is less important and the no detriment principle is potentially more important.

That can become slightly difficult. The more that you argue for further fiscal devolution and further fiscal powers, the further you move away from a fiscal union. The taxpayer fairness principle might then become more important because there is less of a rationale for transfers between parts of the country when there is less of a fiscal union.

That also leads to one of the issues that we touch on in our latest report. If tax revenues grow less quickly in Scotland than they do in the rest of the UK or in England, as has been happening with income tax in recent years, should the Scottish Government bear all the risk of relatively slower growth in per capita revenue, or should there be some form of insurance against that? That is one of the questions that David Eiser said should be discussed in the review of the block grant adjustment mechanism.

Time is marching on and I have six colleagues who want to ask questions, so I will ask just one more. It is about the issue of transparency that David Bell touched on, so he might want to answer.

The Institute of Chartered Accountants of Scotland highlighted the example of a lack of clarity around the role of block grant adjustments in relation to the land and buildings transaction tax additional dwelling supplement. The institute noted that some had suggested that the tax was introduced in Scotland only because, without the introduction of a similar tax to England’s, the corresponding block grant adjustment would be such that there would be a reduction in the block grant. It commented that that

“demonstrates that either the BGA is wrong in principle or that there is great confusion over it”.

Will you comment on that, Professor Bell?

This is an important point. The design of the fiscal framework is such that there is a block grant adjustment for each tax that has been devolved, and for each welfare benefit and so on. That means that it is difficult to change the structures radically. It is difficult to make even modest changes in the structures of taxes, and there is even less scope to consider changes in the way that welfare benefits are structured.

Once you start to try to do that, it becomes more and more difficult to decide what the appropriate block grant adjustment is. If you no longer have a particular benefit or have changed it to something that does not resemble it in any significant way, how on earth do you work out a block grant adjustment? You have to assume that the previous instrument continued to exist, but estimation of the appropriate block grant adjustment would be very difficult. I wonder whether the whole idea of block grant adjustments to some extent restricts the Scottish Government’s ability to redesign parts of its tax and welfare systems. I think that ICAS has picked up that point.

On the specific point that ICAS makes, I note that what is going on here involves a basic principle. It can sound irrational, but you have to bear in mind the interaction with the Barnett formula. When the UK Government introduced the second home levy or whatever it was called, that was in effect a tax rise, as an increase in stamp duty land tax in England. A tax rise will generate additional spending, and additional spending will generate a Barnett consequential for the Scottish Government. In this case, however, that tax rise did not apply in Scotland.

The fact that the block grant adjustment increases makes sense in that, in effect, it ensures that the Scottish budget does not benefit from a tax increase that applies in England but not in Scotland. It is then up to the Scottish Government to decide whether to introduce an equivalent tax, or to raise tax equivalently, in Scotland. In that particular case, it decided to follow suit and introduce the additional dwelling supplement in Scotland, and it subsequently decided to increase the supplement to 4 per cent in Scotland although the equivalent tax remained at 3 per cent in England for a while.

There is a rationale to what goes on. I think that, sometimes, the rationale is lost because it is complicated and you have to think about the interaction with the Barnett formula, but it is that interaction that makes the block grant adjustments seem a bit more logical. There was a logical principle behind what went on. It was not just that the Scottish block grant was being cut and, therefore, the Scottish Government had to respond to maintain its budget at a particular level.

Thank you. I open up the meeting for questions from colleagues.

My first question is for David Phillips. I was very interested in—[Inaudible.]—all the different principles behind the Smith commission and that, because of that, choices might have to be made about priorities. Obviously, those are political decisions for the Scottish and UK Governments. Is it your view that it would be possible for the independent review to flag up the costs and benefits of choosing different priorities? I know that it will not be the job of the independent review to recommend policy, but should it be part of the review to look at the costs and benefits of the different priorities that could be chosen?

I will echo one of the comments that David Eiser made in response to an earlier question. It would be feasible, as part of the independent report and potentially part of the wider review, to have a conceptual look at what the different principles that guide the current framework are and how they fit into the different potential conceptions of the union’s purpose.

As I said, the taxpayer fairness principle, under which money stays in the country in which it is raised, makes more sense if we are moving towards a looser union. The no-detriment principle makes more sense if we see a stronger fiscal union, as does some sort of insurance against what will happen if there is a decline or an increase in revenue per capita. Revenue sharing makes more sense in a stronger fiscal union. There is a qualitative, principles-based discussion about the benefits, costs and trade-offs of the different options.

10:30  

There is scope for quantitative modelling of the different options, using historical data to show what would have happened in the past if the different options had been in place. In your next evidence session, you might hear about some of the work that Guto Ifan and Ed Poole have done at Cardiff University in looking at the different options for Scotland.

We can also look at future scenarios. When we had our initial look in 2015-16, we did two things. We considered the effects and what would happen if revenues grew more quickly or more slowly and if populations grew more quickly or more slowly. Such analysis is very informative. We can see the different options leading to hundreds of millions of pounds of difference in the Scottish Government’s budget over the space of just a few years.

Just to be clear, you are recommending that that should be done in the independent report about the—[Inaudible.]—and also in the independent review of the wider issue.

Yes, because I think that that links up to the other aspects of the fiscal framework. For example, if we have different trajectories of revenues in the short term and there is less insurance via the block grant adjustment against revenues growing more quickly or slowly in Scotland, you might want to have more borrowing powers so that there can be self-insurance via borrowing and reserves. That is why I would have that as part of the focused work that looks at the block grant adjustments and ensure that the interactions between that and the wider review of things such as borrowing powers, reserves and dispute resolution are focused on as well.

That is helpful. That is important for the overall principle behind exactly what we are trying to achieve—namely, the best outcome for Scotland and the United Kingdom.

There is probably growing consensus on the political spectrum that there is a case for examining the issue of forecast errors. People who have been in front of the committee fairly recently have had concerns about our ability to forecast well, the time delays between forecasts, and whether we get things right.

I am not worried about who answers these questions, but are there aspects that we need to address that are not just to do with the timing of different forecasts, particularly Scottish Fiscal Commission forecasts against Office for Budget Responsibility ones? Is there extra data that we should be working on to improve our forecast ability?

It seems to me that the Scottish Fiscal Commission is trying to put together as much data as is currently available. The real-time income tax data is helping quite a lot, but we do not have a long run of data yet, and we have had very unusual circumstances over the past five years, particularly because of the impact of the pandemic and, to some extent, Brexit. It seems to me that that has made it very difficult to develop over that time a stable model whose forecasting capability we can rely on.

I will add a point that is relevant to that question and to what David Phillips has just said. An issue that increasingly occurs as we add to the number of fiscal instruments and their BGAs, as has been the case over the past five years, is how they interact. Do they move together or do some of them move against one another? The overall borrowing that the Scottish Government might require is partly dependent on that because, if one tax goes up and the other goes down, the need to borrow will be much less than it will be if they both go in the same direction. That further complicates the already difficult issue that the Scottish Fiscal Commission faces as regards forecasting.

Thank you. That is—

I think that Liz Smith’s feed has frozen.

I do not know whether you have seen the submission from the witnesses on our next panel, which compares the arrangements in Wales with those in Scotland, but it seems that, on the whole, the Welsh are looking positively at their financial outputs, whereas we are looking negatively at ours. It seems that, under all the scenarios that our next witnesses look at, regardless of whether the comparable method or the IPC method is used, our budget will go down relative to that of Wales in the coming years.

Does there need to be a change in the current arrangements so that the Scottish budget does not continue to suffer? Specifically, should we look at some of the things that the Welsh have done, such as splitting different tax rates for different block grant adjustments? The Welsh also have a 5 per cent bonus on the Barnett formula because, as I understand it, the formula is designed to cut the needs-based Welsh and Scottish spending over time. Should we be following the Welsh model? I do not mind who answers that, but perhaps we can start with David Phillips.

I have a couple of points to make in response to that question. First, the relative decline in Scottish revenues compared with the block grant adjustment over the forecast period is partly to do with the fact that the block grant adjustment is based on UK revenues. It includes freezes in the higher rate threshold that have not yet been built into the Scottish forecasts because the Scottish Government has not yet made policy for income tax beyond 2022.

Based on current forecasts, if the Scottish Government was also to freeze its higher rate threshold, as is being done in England, Northern Ireland and Wales, I would expect the gap to be somewhat smaller. It is worth bearing in mind that the differences between the forecasts reflect not only differences in underlying performance, but differences in policies that have been pencilled in for future years.

I turn to the more substantive question of which options Scotland should go for as regards its fiscal framework and the block grant adjustments. The work of Ed Poole and Guto Ifan shows that there would be a benefit to the Scottish Government if there were separate adjustments by income tax band, as happens in Wales, but if, in order to secure that, it was necessary to move from the index per capita method to the comparable method, that would more than offset the gains. That is because, in Scotland’s case—historically, at least—the protection from slower population growth has been more important than the protection from the fact that more of the tax revenue in Scotland comes from the basic rate, which means that it is more exposed to basic rate income tax trends and less exposed to the higher rate income tax trends.

If you had to choose between the two methods, using the index per capita method would make more sense than using the method in the Welsh model. If you could combine the two, that might be seen to offer the best of both worlds, but whether the Treasury would concede that is another matter.

The Barnett floor that is put in place for Wales—the 5 per cent uplift on top of the population increments that Wales gets—was introduced in the context of the work of the Holtham commission. It suggested that, in the early 2010s, of the three devolved nations of Scotland, Wales and Northern Ireland, Wales was the one that was potentially underfunded relative to what a needs assessment would suggest. That does not hold for Scotland or Northern Ireland. On any sort of formula basis, it would probably turn out that Scotland is relatively overfunded relative to the rest of the UK, although not necessarily absolutely. Therefore, although it would benefit Scotland to get a 5 per cent uplift, it might be harder for Scotland to make that case to the Treasury than it was for Wales.

That is helpful. We are pressed for time, so I will move on and perhaps the other two witnesses can come in on that subject later.

Recommendation 1 on page 5 of your report says that funding guarantees should not continue. I understand the logic behind that, but I wonder how we deal with it. To me, the problem is that, when Westminster announces expenditure, we do not know whether it is new money or existing money. It would be better if Westminster would just tell us how much of it is new money. For example, if the UK Government is spending £5 billion on London crossrail and we are going to get £500 million, if it would at least tell us that £250 million of that is definitely new money we could then bank on that, even though it is not technically a guarantee. Is there a way round that issue? Otherwise, it takes us ages to find out whether we are getting new money.

That is for Professor—

I—

I was going to say that it was for Professor Bell, but maybe David Phillips wants to come in.

David Phillips can go if he is ready.

Sorry.

I was going to say two things on that. First, I completely agree with John Mason that we need better information from the Treasury on what is new money and what is not new money. We produced a short report on that back in 2020, when some of the money for the plan for growth and jobs was actually just rebadged old money but that was not clear from the statement on the plan. I definitely agree on that point. We need better information on what is new money and what is from existing budgets.

The second point is that the report that we are discussing today also suggests that the Scottish Government should have some discretionary borrowing powers. One reason for that is that it would allow the Scottish Government to respond to not only unforeseen events but issues such as the one about new and old money. We have suggested the package as a whole. We are saying that we need better information and some discretionary borrowing powers and that, if there are any late cuts to budgets at the supplementary estimates stage, the Scottish Government should have the ability to borrow to carry that forward so that it does not have to make cuts straight away. That would give the flexibility or ability to plan that is provided by the guarantees without the potential unfairness to England of the guarantees being more than Scotland should get. That is what led us to that conclusion.

I get the point about the unfairness.

I have one final point. On pages 45 and 46, your report is quite positive about the idea that we should have more flexibility in relation to transferring capital budgets to resource. We are allowed to put resource into capital, but we had not been doing so. Especially during the pandemic, the ability to transfer capital to resource would have been very useful. Am I right in saying that you feel that that would not damage the UK in any way and that it would give a bit more flexibility to the Scottish Parliament?

David Eiser might want to answer that question.

In essence, the answer is yes. Our report makes the point that there is a case for the Scottish Government having greater ability to flex its budget during the year in response to unforeseen events. That would be done partly through the modest additional discretionary borrowing that we have talked about and partly through increased limits for drawdown from the reserve. It would also be partly about the ability to transfer budgets from capital to resource.

10:45  

The first two of those things probably give better scope for flexibility. If you were to request additional flexibilities, the first two might be better options to start with. In essence, however, our report makes the case that the Scottish Government should have better ability to flex its budget during the year through those mechanisms. At the sort of scale that we are talking about, the changes would provide useful additional flexibilities to the Scottish Government but would not pose any major fiscal risks to the UK Government, so there is no fundamental reason why the UK Government should object to those sorts of extensions to flexibility.

Thanks very much.

Liz Smith was cut off in her prime earlier, so I will allow her back in for a question before we move on.

Thank you, convener. I am sorry about that—there was a slight blip in the broadcasting.

The third issue that I want to raise is exogenous shocks. Obviously, we have experienced one as a result of Covid. Your report is clear that, in such circumstances, there might have to be minimum funding guarantees. I think that everybody agrees that they have been a good thing, but you also say that they cannot continue on a long-term basis, because that would be inherently unfair. You set out that, in an exogenous shock situation, there is the potential to have enhanced borrowing powers. How exactly do you see that working for the period of an exogenous shock? Would there be a time period for the enhanced borrowing?

We have discussed that quite a bit. There is a lot of difficulty in identifying what we would call an asymmetric shock rather than an exogenous shock. An asymmetric shock is one that hits Scotland or Wales while the economy in the rest of the UK carries on unperturbed. There is a question about timing. For example, we could say that the decline in the North Sea industry is a sort of slow burn rather than an immediate shock such as the pandemic. There is a question about how to recognise the differences between those.

Funding guarantees worked well during the pandemic, but that was partly because it was not an asymmetric shock. That is the point that we make in the report. Covid affected Wales, Scotland, Northern Ireland and England to pretty much the same extent, and therefore the Barnett consequentials were at about the right levels. We do not have a methodology for recognising an asymmetric shock, which we would have to be able to do first, before deciding what borrowing regime to bring in to deal with it. That is another issue that the review could be asked to look at.

It is important to try to establish critically the purpose and effect of each of the measures around the block grant adjustments, as that is where the bulk of the discussion and debate is likely to be. I want to clarify a few things. In essence, the comparable method seems to involve isolating the ability of a Scottish taxpayer to pay tax in comparison to the average UK taxpayer. On top of that, in the index per capita method, we have a factor to offset the ability of Scotland to grow its population relative to the rest of the UK. Wales has a system that isolates each tax band, which, in a sense, offsets its ability to change the make-up of its tax base. Is that a fair summary of what the three different methods do?

In your report, you say that the review of block grant adjustments is a political decision, not a technical one. I think that I understand that. Therefore, is the political decision the extent to which the devolved Governments are able to influence the size of their population, the extent to which their citizens are able to pay tax and the structure of that tax? Is that a fair summary of what the political drivers of the models might be?

Yes, that is quite a good summary of what the different options do. You are right to say that the comparable model provides protection when starting off with lower revenues per capita, because, instead of having to get the same pounds-per-person increase, there needs to be the same percentage increase. The IPC method gives protection on top of that if population growth is slower. The Welsh method gives protection if growth rates are expected to be different because of the structure of the income tax base. That is a good summary of the different approaches.

There are two aspects to the politics. There is the question of what is influenceable by the devolved Governments—the extent to which they can have an impact on the performance of the economy and on population growth. There is also a question about the degree to which there should be risk sharing and redistribution across the UK. That is a separate political issue about the function of the union. If one thinks that the union is about redistribution according to need and about insurance against shocks that hit different parts of the country, one will be pushed towards a system that might insure devolved Governments against more of the different kinds of risks. Those who believe that there should be a move towards a system of fuller fiscal autonomy, with more things devolved, might think that there should be protection against fewer of those risks. That is why we discussed it as a political issue.

Another reason why we highlight the politics when thinking about the block grant adjustment relates to what happens if one starts to put in more mechanisms to insure against different risks. For example, even if there is protection against population growth and the different starting tax distributions, what happens if revenues per capita grow less quickly in Scotland because the economy grows less quickly? Should there be protection against that? If we try to provide that protection, some ad hoc adjustments will be needed, so politics will get involved. If it seems to be a zero-sum game, the issue can become quite politically contentious. Some culture change in how devolved and central Governments interact on fiscal payment issues will be needed to avoid such issues becoming a constant source of tension between the Governments. That is why we think that the politics is just as important as the technical side of things when thinking about block grant adjustments.

Does either of the other Davids want to come in?

I think that I typed R at the same time as David Phillips, who has said what I would have said. This is about principles and politics. An example of that relates to when we talk about whether the indexation method should be the IPC method or the comparable method. As David Phillips said, the IPC method, which is currently in operation in Scotland, protects the Scottish budget against the risk that the Scottish population grows more slowly than the population in the rest of the UK. The politics is that the UK Government will say, “Hang on a minute. When it comes to the Barnett formula side of the equation, that doesn’t fully adjust for population. Scotland can benefit—and, indeed, has benefited over the years—from relatively slower population growth than there has been in England, so why should we protect the Scottish budget from slower population growth on the tax side if Scotland can benefit from slower population growth on the Barnett formula side?”

There are principles and there are politics, but your assessment of the different methods is correct. The IPC method protects the Scottish budget against the risk that the population grows more slowly in Scotland. The comparable method does not provide that protection. The IPC and comparable methods both protect against the lower tax capacity of the Scottish and Welsh economies, but the Welsh model, which has different adjustments by band, also protects against the different initial distribution of the tax base.

I reiterate that, as David Phillips said, if you introduce more different protections into the mechanism, you not only complicate the system a bit more but introduce questions about whether, if you want more protections on the tax side, that is inconsistent with what is happening on the spending side and, therefore, whether we should introduce more assessments of need on the spending side.

There are lots of different trade-offs to make. Those decisions are partly about the principles of which risks the Scottish budget should be exposed to, but there is a lot of politics in the matter as well.

My next question follows on from the answers to some of John Mason’s questions. Although applying the Welsh model in Wales projects that the Welsh budget will increase, it does not do the same thing for the Scottish budget if we apply it in Scotland. It means that the decline is not as severe, but we would still be worse off than if the current set-up had not been put in place.

When the Scottish Fiscal Commission’s report came out, prior to the budget, everyone was taken by surprise at how significant the lag in income tax growth was in Scotland compared to in England. That is the fundamental driver. Why is that the case? Why is income tax growing more slowly in Scotland than it is in not just the UK as a whole but pretty much every other region in the UK, including Wales? What levers are available to the Scottish Government to address that?

The exogenous shock that Liz Smith talked about is coming through the gradual decline in North Sea oil. That industry, along with financial services, was among the few industries in Scotland that were producing additional-rate and higher-rate taxpayers. The top 1 per cent of the taxpaying population pays 13 per cent of the total revenues. The slow growth in tax revenues is partly distributional and partly to do with the point that, in the Scottish Fiscal Commission’s view, economic growth in Scotland post-2022-23 will be very modest, which will also generate relatively smaller growth in tax revenues.

Is the fundamental point not that, regardless of what model you choose, if we have a devolved taxation system, Scotland needs to grow its income tax faster than the UK average in order to benefit? In a sense, it does not matter which model we use; that fundamental truth is the overarching one.

11:00  

That is critical. Although the indexing method does make a difference, as we showed in our 2016 paper, in the long term, it is economic growth that is key to the evolution of Scotland’s total revenues.

I have one final question, which is about the long-term sustainability of the method. If we are basing it on tax and welfare decisions in Scotland compared to policy positions for the UK as a whole in 2016, that becomes more difficult as time passes. It is feasible, because the UK has, by and large, pursued the same overall policies on the method and size of taxation and welfare spending compared to 2016. However, if the UK Government diverged significantly from that, either by increasing or decreasing tax or welfare spending, projecting what would have happened from that 2016 position becomes more and more difficult, if not impossible. Will we have to have a far more fundamental rethink in five or 10 years’ time?

On the first point, the block grant adjustment is based on actual revenues in England and Northern Ireland at the moment—and potentially only in England in future if taxation is devolved to Northern Ireland. Because it is based on actual revenues, you can still make future block grant adjustments. As David Eiser said, it is of fundamental importance to the fairness of the system that, if income tax is substantially increased in England, leading to more spending either on UK-wide functions or via the Barnett formula, you would need a bigger block grant adjustment to offset the money that Scotland gets via the Barnett formula. It is fundamentally important still to link the block grant adjustment to what happens to revenues in England.

That system can continue working in future. However, if we were to see an on-going trend of relatively slower—or faster—growth of revenues in Scotland over five, 10, 15 or 20 years, should the block grant adjustments be updated to account for that and to share risk across the UK? That is one of the questions that we look at in our report. It comes down to what one thinks is the role of the union. If one thinks that it is a risk-sharing union of redistribution from richer to poorer areas, one probably would argue that there should be such reassessments of the block grant adjustment in the longer term to redistribute revenue from faster growing to slower growing areas. Someone who thinks that there should be a loose union, or no union, would find that position harder to argue.

We talk about that a bit in the report. It is technically difficult to do that, which means that there is a lot of scope for arguments about what elasticities to assume or what might happen. That is why there is probably a role for independent institutions in that aspect of the system. Australia has the Commonwealth Grants Commission to have an independent voice in what could look like a zero-sum game.

[Inaudible.]

Are you still alive, Daniel?

I do not have any—[Inaudible.]—but I would be interested to hear whether David Eiser has any comments—[Inaudible.]

David Phillips’s response was exactly right. Indexing the block grant adjustments to growth in rest-of-UK revenues serves a number of important purposes. It protects the Scottish budget from UK-wide shocks or declines in revenues. As we saw in 2020, falls in Scottish revenues because of Covid were offset by equivalent falls in RUK revenues. It also ensures that we get taxpayer fairness, with Scottish taxpayers not benefiting from increases in tax in RUK and vice versa.

However, Daniel Johnson is right that, over time, if Scottish revenues were to fall because of a long-term decline—as a result of what is happening to the offshore sector, for example—that would not feel particularly fair. If, as a result of tax devolution and trends that neither the Scottish Government nor any other Government can do a great deal about, the Scottish budget was disadvantaged significantly over time, that would not feel like a particularly reasonable outcome of devolution. There would, as David Phillips said, be a case for looking at resetting the clock periodically.

However, that does not in itself imply that the block grant adjustments and the basic principles on which they are calculated are fundamentally wrong, because they serve a number of important purposes.

I will go back to the question of risks. From reading the report, it seems that the risks for Scotland and the risks for Wales are completely different. I am trying to understand that. Is it down to differing appetites for risk between the two Governments? Is it essentially a political decision that has been made by each Government? That question may be for David Eiser first.

We have touched on that. David Phillips made the point that the demographic risk—the risk from the distribution of income tax payers, if you like—in Wales is slightly different from the risk in Scotland. The population of Wales is projected to grow somewhat more slowly than the English population, but the difference is not as marked as it is for Scotland. Scotland has a different distribution of income tax payers, but the difference is not as marked as it is for Wales.

The calculations in the two countries are a bit different. If Scotland had to choose whether it wanted to move to the Welsh model, which would expose the Scottish budget to differences in population growth but protect it from a difference in the distribution of taxpayers, as opposed to the method that it currently has, that would be an interesting calculation. I think that Scotland would rather stick with what it has than go for the Welsh model. Looking at projections and doing some scenario modelling is important in informing such decisions.

Would I be right in thinking that our current model was selected probably because the Scottish Government had more appetite for risk than the Welsh Government did when it was negotiating its block grant adjustments?

I do not know that it was about having more appetite for risk—the Scottish Government simply made a different calculation of the risks that it faced. In a sense, the Welsh Government was prepared to accept the comparable method as part of the negotiation that also saw it secure the adjustment to the Barnett formula to ensure that there was a sort of floor below which Wales’s allocation could not drop. At the moment, that is not an issue that the Scottish Government is likely to be looking at, because its allocation under the Barnett formula is way above what a needs assessment would suggest.

This follows on from Daniel Johnson’s question about the fact that, to keep level, our economy would have to grow by the same amount as the rest of the UK. Is that the same for Wales? When we look at the graph, we see that the net effect of its tax devolution is positive, but ours is negative. Is Wales growing its economy at the same or higher rate than the rest of the UK? Is the rate higher than Scotland’s?

That is an important point. There are different mechanisms for indexing block grant adjustment. They are important, because they influence which risks the devolved budgets are exposed to. Fundamentally, if the tax base grows much less quickly in Scotland or Wales than it does in the rest of the UK, there is not a block grant adjustment that will compensate for that.

In a way, it would be odd if there was. You have to think about the way that the interaction with the Barnett formula works. We have talked about the implications of that. It would be somewhat strange if we had a devolved tax system whereby the devolved tax base and tax revenues grew less slowly, yet the budget was not exposed to the impacts of that.

I am just trying to understand the issue. Looking at the graphs, we see that there has been a positive impact for Wales. Does that mean that its economy is growing faster than Scotland’s? What does Scotland need to do to try to halt the decline of the benefits of tax devolution?

Maybe one of the other Davids would like to answer.

I can briefly touch on that. One thing to bear in mind is that income tax devolution has really only just started in Wales, and it started the year before Covid. Because more of the economy in Wales is public sector, during the Covid crisis, the Welsh employment and income figures held up largely better than the rest of the UK’s. It looks like one factor underlying the largely better performance of income tax revenues in Wales is that there has been this slightly stronger short-term performance during the Covid pandemic.

Looking beyond that, it is not fully clear to me what factors would be driving the forecasted continued small improvements in Welsh tax revenues compared to those in England and Northern Ireland, given that the OBR has the same assumptions about growth. Unlike the SFC, the OBR does not do a separate forecast for Wales’s economic growth; it is basically the same forecast that it is doing for the UK as a whole. It could be to do with factors about the distribution of income and the starting points. There have been some factors that have led to slightly better performance on income and employment in Wales.

What is going on in the forecasts in the longer term is a bit more uncertain. On your question about whether it is fundamentally the same model if Wales’s economy is growing faster and doing better, yes, it is fundamentally the same model.

11:15  

I have a question for all three of the Davids, but perhaps Professor David Bell could start. We have focused on Scotland and Wales, but I am interested in what helpful precedents operate elsewhere in the world, particularly those that deal with issues around fiscal transfers and divergence over time. I feel as though we are dancing on the head of a pin, particularly in relation to indexation methods, and some of those problems are not unique. I would appreciate some thoughts from Professor Bell on that in relation to the review.

The UK system is rather unique. Most countries have a more systematic or legalistic approach to the funding of federal Governments and provinces or states. That is not to say that they all protect from risk—in fact, some of them perhaps provide less protection from risk. There are transfers, tax equalisation methods and spending equalisation methods, and different countries do it in different ways. We have mentioned the Commonwealth Grants Commission, which is an independent body that decides on levels of support for different parts of Australia. In Canada, the provinces are pretty independent—for example, they have their own borrowing powers and the federal Government does not really intervene in that borrowing.

What we have been discussing is relatively unusual; one of the key things is that it works without a formal legal framework around the arrangements that we have discussed. Whether that is a good or a bad thing can be debated. We argued in our recent paper that we managed the model through the pandemic because the Treasury offered funding guarantees. That was a completely new idea that suddenly appeared but which was helpful to the devolved Governments over that period. The arrangements in the UK are not by any means typical of the relationship between federal and state or province level, or whatever you call other levels of government.

To what extent does the Barnett formula embed structural imbalances in the UK? You mentioned financial services, which is an industry that I was involved in for many years. Over the course of my career, head office functions moved to London, bar a few noteworthy examples such as Standard Life, although things have changed for it as well.

The Barnett formula is not based on need. In relation to equalisation mechanisms, most countries look at the relative levels of deprivation, unemployment and so on, but that is not how the Barnett formula works. The UK economy has very significant regional differences between the richest and poorest parts of the country, which the EU regularly pointed out.

To an extent, that might be down to the way in which Government funding is done, but it might also be down to more fundamental economic processes that have led to the relative decline of different parts of the UK. It is not just the Celtic nations that we are talking about; it is also the north of England relative to the south-east.

Earlier, we talked about how one could grow the tax base—in other words, the working-age population—in Scotland. Daniel Johnson mentioned that, the committee has talked about it a great deal, and I would like to understand the issue a bit more.

Professor Bell, in your view, what are the limitations on the fiscal levers that the Scottish Government currently has to influence and grow the tax base, regardless of the indexation method? Of course, we are really looking at the working-age population. I would appreciate hearing your thoughts on the current limitations in the area that we are discussing.

I will mention an issue that could have come up in answer to previous questions, which relates to the difficulty of making forecasts and to the issues with growing the population. Scotland’s population was in decline for a large chunk of the last century; it has grown only since the beginning of this century, and that has happened largely because of migration rather than because of what we call natural increase, which is the difference between births and deaths.

The most recent forecasts for the Scottish population are from 2018, so they are a bit out of date, which could affect the projections that we have been talking about at length during this morning’s session. We are also in a new migration situation as far as the whole of the UK is concerned, and it is not clear whether that will have a positive or a negative effect on population growth in the medium to long term. That, in turn, will be reflected in all the metrics that we have been discussing this morning as regards the way in which the current risks are shared across the different parts of the UK as a result of the fiscal frameworks that have been agreed with Scotland and Wales.

I do not know whether the other guys want to add anything to that.

I can see that David Phillips wants to come in. He has put an R in the chat box.

I have two brief points to make, the first of which goes back to Ms Thomson’s first question about what happens in other countries.

One of the big differences in European systems is that there is often equalisation not just at the point of devolution as we have in the UK, but also on an on-going basis for differences in trends in income tax or other devolved tax growth rates. For example, in Germany, any Land or state that gets less than 99.5 per cent of the average tax revenue has more than three quarters of the shortfall corrected. That is a different approach from the one that we have in the UK. In the UK, the figure is 100 per cent of the shortfall at the point of devolution, but 0 per cent for any further shortfall that grows after that point. Other countries tend to treat shortfalls more symmetrically. Germany says that, with anything below 99.5 per cent, three quarters of the shortfall will be compensated for.

There is stuff to learn from other countries. To be very frank, I think that what we have ended up with in Scotland was driven, ultimately, by the four-week Smith commission, and it takes more than four weeks to properly consider such issues. Although the review needs to take place relatively urgently, I hope that it is not too rushed.

Secondly, on the question about what levers are devolved and what powers Scotland has, the important thing to note is that, as well as population and participation, the other, most fundamental, driver of economic growth is productivity, as was set out by the sustainable growth commission and in various Government documents.

Large parts of capital investment are devolved to the Scottish Government, so the capital budget is actually the most generous aspect of the Scottish budget, relative to that in the rest of the UK. Government capital investment per person is about 50 per cent higher in Scotland than it is in the rest of the UK. There is therefore scope to think about the use of capital investment.

The other devolved area that is very important for productivity is education and skills policy. I do not deny that there are areas that are not devolved that are very important for devolution, such as regulation policy and overall macro policy, and the UK shared prosperity fund will be much more UK driven than the EU scheme was. However, some things that are devolved are key levers, particularly when it comes to capital investment and education and skills policy.

Thank you for that—although my question was actually about what the limitations of the current arrangements are, rather than the benefits. Perhaps you would like to fill in on that. What are the limitations of the fiscal levers that the Scottish Government has in influencing the tax base? That was my question.

A number of areas are reserved to the UK Government: most regulations around product markets and labour markets; the international migration regime; and overall macro fiscal policy. Those areas are not under the Scottish Government’s control.

That brings us to the fundamental question, which would involve considering the additional benefits that Scotland could have if it had access to additional levers, versus the potential costs from reduced redistribution and equalisation in the UK. However, that goes far beyond the scope of your question.

Thank you for mentioning those fundamentals.

I will let you move on now, convener.

The floor is yours, Ross.

Thank you, convener. I have one question on the process, and then a couple on policy if we have time to get through them.

First, on the process itself, there has been a bit of confusion in the public discourse on the independent report and the review. I say “public discourse”; it is not as though a huge number of people have been engaging in this conversation beyond those of us who are participating in this meeting, but some have. Some folk are mixing up their terms when they reference the independent report rather than the review.

The report itself will not make recommendations. To an extent, it is simply an evidence-gathering exercise. I would be interested to hear the witnesses’ thoughts on exactly what they think the most desirable outcome is for the independent report. What purpose is it trying to serve, given that it is not its purpose to make recommendations?

I direct that to David Bell in the first instance.

I am not entirely sure what the clear answer to that is. As we said earlier, the report should go through the various options, do some backcasting and some forecasting, present those findings to the review and have them taken forward. To an extent, it will be a fairly automatic exercise, which will bring in some of the issues that we have discussed about forecasting errors, population forecasts, sensitivity of the results to variations and all those kinds of parameters.

That is my expectation, although I am not privy to the thinking of the UK and Scottish Governments on the matter at present.

11:30  

Thank you—that is useful. The whole exercise is highly politicised, of course, but given those confines, the report is a relatively technical part of informing what will be a much more politicised review.

David Eiser mentioned forecast error borrowing. Before we get into a debate about how we decide on the limit for that—whether it should be a cash percentage or whatever—which I presume will come with the review, should we ask what the rationale is for having a limit on forecast error borrowing at all? It is less about a divergence in policy choice and more about correcting for a divergence in technical exercises; it is about correcting for error rather than for a divergence in choice. What purpose does a limit serve when the issue is simply to do with forecast error corrections?

I am quite sympathetic to that view. As you say, the Scottish Government does not set that independently; it is a result of the outcomes of the two independent forecasters operating together. Even limits that were significantly higher than what we currently have would not pose any fiscal risk to the UK Government.

I will move on to my other question on borrowing—I direct it to David Phillips, as he mentioned this issue earlier in referring to the report that your institutions recently published on discretionary resource spending borrowing powers. The report recommended that the Scottish Government should be given some limited powers for discretionary borrowing, and said that the rationale for those powers being limited is an equity issue, as there is no England-only borrowing regime.

Can you expand a little on that rationale? The UK Government is de facto the English Government when it comes to areas such as health and local government, and it has an unfettered ability to borrow and spend in those areas if it wishes to do so. What is the rationale for granting the Scottish Government a discretionary borrowing power but having it limited for the purposes of equity within the union?

That is a good question. I will go over the rationale first and then address the second part of your question. As I said earlier, in effect, the rationale is that, as it stands, any borrowing by the UK Government either automatically generates spending on things that are UK wide, such as defence or social security, or paying down the national debt, or is spent on public services in England and therefore generates Barnett consequentials for Scotland, Wales and Northern Ireland. Any borrowing by the UK Government is UK wide.

Scottish or Welsh borrowing comes on top of that, and it is paid back by the residents of Scotland or Wales, at least unless there is any sort of bailout. There is value in that, otherwise people would not want the borrowing powers. That is what led to our conclusion in the report that one needs to think about the fairness of all this to England.

On your second point, while it is true that there are no England-only borrowing powers, the UK Government can, if it wants to borrow more for England, simply make that decision and go ahead with it. There is therefore an argument that, although there is financial unfairness to England if we have significant Scottish and Welsh borrowing on top of UK borrowing, there is a power-dynamic unfairness in the current system. Borrowing for England can be done at the drop of a hat if the UK Government decides that that is needed, whereas if the Scottish Government decides that we need more borrowing for Scotland, it does not have that power.

There is a power issue involved, and I am not sure of the extent to which that can be avoided. Even if there was a separate English Government with its own borrowing powers, England is 83 or 84 per cent of the union, so it will have a large political say in the union. The power issue is potentially inherent in a union that has one very large area.

Of course, one could say that there are regions within England that have different preferences. The fundamental way to think about the issue is to understand that we potentially need to move towards a system in which the UK Government makes decisions on behalf of the UK as a whole but takes into account the needs of each different area of the UK. That is why we say in our report, in looking at what happens if there is an asymmetric shock that hits one of the devolved nations—or, potentially, one of the English regions—harder than the rest of the UK, that there need to be ways to bypass the normal system and give additional direct funding or additional borrowing powers to that nation or region.

My own view is that, in more normal times, it is not unreasonable in a unitary state for overall fiscal policy to be determined by the central Government. The UK is a slightly unusual unitary state, thinking about it nation by nation, given that one nation makes up more than eight tenths of the overall size of the state.

Yes, I think that we are a deeply asymmetric unitary state.

I will leave it there, convener.

I thank our three witnesses. I had hoped to ask a question about behavioural responses—I know that that issue is close to David Bell’s heart—but time is sadly against us, as we are well over time already, so I will call a halt to the session.

I hope to see all the witnesses again before too long; I am sure that we have a lot more to discuss on this particular issue. Once again, I thank the three Davids.

11:37 Meeting suspended.  

11:45 On resuming—  

For our second evidence session, which will focus on the Welsh fiscal framework, we have been joined by Dr Ed Poole and Guto Ifan from Cardiff University. I welcome them both to our meeting. Members have received a written submission from our witnesses.

We will move straight to questions. I remind members and witnesses that our broadcasting team will operate their microphones, so you should pause for a few seconds before speaking to ensure that you are heard. We have about an hour for this session.

I will start by asking about the three separate block grant adjustments for income tax that are a critical element of the Welsh fiscal framework. I understand that they insulate the Welsh Government from UK-wide factors that disproportionately affect one part of income distribution, thereby allowing a fair system to operate in Wales despite its markedly weaker tax base.

We have your submission, but for the record and for people who are listening in, will you talk us through how the system works to benefit Wales?

Good morning, convener and committee members. Thank you very much for having us. Bore da. Diolch yn fawr iawn i chi gyd.

As we say in our submission, the most important part of the discussions that we had in Wales before the fiscal framework was agreed was about how to deal with Wales’s poorer tax base. We had the advantage of the Scottish fiscal framework negotiations having happened earlier in the same year, as well as the great deal of work that had been done by parliamentary committees, the three panellists that you had in your previous evidence session and a large number of individuals in Scotland on how the block grant adjustments would work in the Scottish case.

However, when we looked at the numbers and how they would apply in Wales, we saw that although population risk was an element that would be of concern, the bigger risk would be the great dependence on Wales’s lower-rate taxpayers in our tax base—in particular, on taxes that are earned from the basic rate of income tax.

You will remember that at that time—in the first decade of the coalition Government, or the first five, six or seven years of the previous decade—there were very rapid increases in the personal allowance, or the amount of income that is tax free. That was, if you like, cannibalising more of the Welsh tax base than was the case across the UK as a whole, and was therefore likely to have the impact of faster block grant adjustments than Wales could earn even in a reasonable scenario of tax adjustment and tax growth.

As part of the negotiations—which involved David Phillips, who was on the previous panel—we looked for a way in which the weaker tax base in Wales could be accounted for in a way that was fair to both the Welsh side and the UK side. The way to do that was to have three block grant adjustments for income tax, which means that Wales would not be overly penalised by impacts on the tax base hitting Wales disproportionately. Instead of comparing the whole Welsh tax base with that of the UK as a whole, we compare the basic rate in Wales with the basic rate in the rest of the UK. That gives fairer protection. The impact of the rapidly growing personal allowance would be felt by Wales and the UK, so we considered only the basic rate of income tax on both sides of the border.

Likewise, only a very small number of earners in Wales pay income tax at the additional rate—the top rate—which is much more important at the UK level. Again, we needed a separate block grant adjustment to account for that massive disproportionality.

Thank you—that was very helpful. Your submission is excellent and very readable, so you have already answered many of my questions, but I want to take the discussion to a wider audience of people who will not have seen your submission.

The information that you provide in figures 3.2 and 3.3 and in paragraphs 3.11 and 3.12 is compelling. At the end of paragraph 3.11, you say:

“the overall impact of tax devolution on the Welsh budget is forecast to be a surplus of £252 million a year by 2026-27.”

At the end of paragraph 3.12, you say that in Scotland, by contrast, the

“negative projected effect of tax devolution is estimated to reach £355 million a year by 2026-27.”

You go on to say that if separate block grant adjustments for different income tax bands were implemented in Scotland, although there would still be a deficit, it would be only £271 million. Although that figure is still high, it is £84 million lower than is currently projected.

Are those figures based solely on the type of block grant adjustment that is currently made or do they take account of the impact of economic growth? Are you advocating that Scotland should go down the road of lobbying for block grant adjustment powers for separate tax rates?

I will briefly answer that, before bringing into the conversation my colleague Guto Ifan, who has been working on the forecasts, too.

The key point to make is that those are all forecasts. That is a really important point, given that we are living through extraordinary times as far as the fiscal arrangements are concerned. As the previous panel said, income tax powers are very new in Wales—the Welsh rate of income tax came in only in April 2019—so we are dealing with data from a very short period.

In Wales’s case, the Office for Budget Responsibility does the Welsh tax forecasts and the UK BGA forecasts—in other words, the same organisation does the forecasts for both sides. It is not particularly clear from the OBR’s devolved tax report what is driving the forecast that Wales will do better on income tax. There is little indication of whether it is to do with growth in the Welsh economy relative to the position in the English economy.

However, we know that the UK Government’s personal allowance freeze will help the Welsh tax base. As I mentioned, we are very dependent on incomes at the lower end of the scale. Although the personal allowance freeze will not help taxpayers, it will help the tax base. In addition, as the previous panel mentioned, an element of protection is provided by the high percentage of the tax base that comes from public sector employment in Wales, which has been relatively shielded—“relatively” being the operative word—during the pandemic.

At the moment, there seems to be an indication that Wales will do better on income tax in the next few years. In addition, our replacement for stamp duty is performing above estimates. There is healthy growth in that because prices are recovering more quickly.

Guto Ifan might want to come in on some of those forecasting points.

I reiterate that our projection is based on forecasts by the OBR. As Ed Poole said, it is not entirely clear what is driving the better forecasts for Wales as regards the block grant adjustment, relative to the position of England and Northern Ireland.

Of course, it is early days in Wales from the point of view of tax devolution. We have outturn data for only the first year of devolution—2019-20—whereas Scotland has three years of data. It is also important to mention that the relative position reflects tax policy changes by the Scottish Government and the fact that the Scottish Fiscal Commission is making assumptions that are different from those of the OBR.

I think that the OBR is set to publish more analysis of its forecasts for the Welsh share of income tax revenues in future years. It will do that soon, which should shed more light on the assumptions that the OBR is making about relative growth in Wales.

Your second question was about whether Scotland should be advocating for separate block grant adjustments. Our projections in the different models are based on outturn data and on forecasts. As you said, the projection is still for a negative effect, even if you account for relative population growth and separate block grant adjustments. The situation improves slightly if you introduce those, but the trend is still downward. That probably reflects Scotland-specific factors—mainly what is happening in the oil and gas sector in the north-east.

The question then becomes what would be a fair amount for the block grant adjustment. There is another way of thinking about the block grant adjustment and what it should capture, which is to consider the revenues that are being forgone by the Treasury because of tax devolution. That is knowable after the point of devolution because of tax rate changes. The block grant adjustment has probably been growing faster than the revenues that the Treasury has foregone. Having a separate block grant adjustment might be a way of bringing you closer to the revenues that have been foregone by the Treasury in recent years.

In other devolved regions around the world, the baseline is reset every five, seven or 10 years. The previous panel touched on that. That might take into account major structural changes such as—as you and others have mentioned—in the oil and gas industry in the north-east of Scotland. Should a reset be built into any future agreement between Scotland and England and, indeed, Wales and England—or, I should say, between Wales and the UK and Scotland and the UK?

I will bring in Guto on that point.

There is another difference between Wales and Scotland. If you look at the data for the past 20 years, there has not really been any major divergence or convergence of Welsh economic performance and that of England and Northern Ireland. That might be less of an issue for Wales; there are other factors in Scotland. The output per person in Wales is probably closer to the UK average, but Scotland had a trend of relative convergence during the early years of devolution but now has a trend of relative divergence from the UK average.

If factors are not in the control of the Scottish Government—the different make-up of the tax base, for instance—there is a case for limiting the effect of that on the budget. That can work both ways. If tax devolution had happened earlier in the devolution process, the Scottish budget might have benefited and a reset would have taken money away from Scotland. If we are thinking about the Scottish budget being exposed to risks that are outside the Scottish Government’s control, there could be a reset after a period of time to limit the divergence in the Scottish budget. As the previous panel said, it is difficult to know what the reset should be and what the extent of any changes after the fact should be.

A reset must be based on fairness to both sides. There cannot be a situation in which Scotland—or Wales—would always benefit. The reset would have to be based on specific criteria. That should be looked at.

Wales’s powers are different from Scotland’s. What impact has that had on the block grant adjustments? Wales has more limited powers. You said that property prices have gone up more than was anticipated and that that has helped with land taxes and stamp duty. What impact has that had? Has the Scottish experience encouraged or discouraged Wales from seeking more fiscal powers?

12:00  

In general, Wales has been in Scotland’s slipstream on fiscal developments. Certainly, the approach that Wales took to the fiscal framework negotiations was similar to that which was taken by Scotland. We learned a lot from work that was done in Scotland going right back to the Calman commission. It is mentioned in the report that the longest-lasting legacy of the Calman commission is, ironically, probably the income tax system that is currently operating in Wales. We learned a great deal from the Scottish experience.

When we looked at the numbers in Wales, whether we used indexed per capita tax or a comparable model—whichever models worked for the BGAs in Scotland—we were concerned about the impact at the bottom end of earned incomes, which was why the Welsh fiscal framework had quite a divergent outcome from—even though it was inspired by—the Scottish experience.

We often criticise the way in which the UK Government conducts its negotiations on a bilateral basis rather than involving all the Governments of the UK multilaterally, but on this occasion the ability to have Wales-specific elements in the framework was important in the outcome that you see so far being projected for tax devolution in Wales, compared with what is projected for Scotland.

Wales might have learned from Scotland, but it is time that Scotland learned from Wales. In terms of the block grant adjustment, what has happened with regard to separate tax rates has been a real eye opener.

Our inquiry will go on for some months; I hope that we will be able to speak to both of you further down the line. However, I have a final question for you, before I bring in other members. If the Scottish Government could wave a magic wand and make one change to Scotland’s block grant adjustment, what should it be?

I will pass that to Guto Ifan.

Well body-swerved.

It would be helpful to have some recognition in the block grant adjustment that there are risks that are outwith the control of the Scottish Government. Since the start of tax devolution in Scotland, we have seen the effect of the sectoral decline of the oil and gas industry in recent years. It would be good to have some adjustment for that, such as the introduction of a separate block grant adjustment methodology. I think that the key ask from the Scottish Government will be for acknowledgment that some factors are outside the Scottish Government’s control.

Thanks.

I was interested to read in our papers about the extra 5 per cent that Wales gets if there are changes because of the Barnett formula. Correct me if I am wrong, but I believe that the Barnett formula reduces the needs-based element when extra funding comes up, but Wales has managed to counter that with the mechanism of having an extra 5 per cent. Can you explain a little of the history of that? I think that suggestion was that the figure should be 15 per cent, so can you tell me how you ended up at 5 per cent?

Certainly. We mention in the report that, about 12 years ago, the Welsh Government set up a review commission that was chaired by Gerry Holtham. It looked at operation of the Barnett formula in Wales and across the UK and found that Wales was underfunded relative to how it would be funded if it were to be treated as a region of England. In other words, if the formula allocations through which funding in England is distributed by the UK Treasury were used in relation to Wales, Wales would receive more funding.

That was correct at the time because, during the first decade of this century, there was a large increase in spending that accelerated the so-called Barnett squeeze and the convergence of spending per capita in Scotland, Wales and Northern Ireland down to per-capita levels in England. That had, if you like, the boosters under it in the first decade of devolution.

At that time, there was a great deal of concern across the parties in Wales that Wales was underfunded as a result of the Barnett formula, which was not the case in the other parts of the UK, including England, with non-formula spending, and Scotland and Northern Ireland, with Barnett formula spending. There has been a long-standing cross-party concern in Wales for at least the past 15 years on how to address the relative weakness in Wales’s spending through the Barnett formula.

The 5 per cent multiplier was, if you like, the carrot for the Welsh Government to accept the stick of the comparable model in the fiscal framework negotiation and agreement on that. In technical terms, in relation to the agreement “needs-based multiplier” is a misnomer, because there is no needs calculation at all—it is simply a number that is inserted into the fiscal framework. It is a zero-sum bargain between the Governments.

The figure is set at 115 per cent, so every single Barnett consequential would be topped up by 15 per cent. However, that kicks in only once Wales’s spending per capita converges to below 115 per cent of the level in England. We think that that could take years—the figure is currently 120 per cent. For the period in which it does not drop below that level, the multiplier is 5 per cent. That is lower than the multiplier that is expected in the long term, but it is still a considerable boost to the Welsh budget, particularly given the extraordinary amounts of money that have been pumped into the block grants as a result of Covid-19 funding.

That is helpful.

The figure of 5 per cent sounds a little arbitrary. I think that you said that there was a deal between the two Governments. We sometimes have the problem in Scotland that the two Governments do a deal, but the Parliament does not get a look in. In Wales, it would be the Senedd. Was that broadly what happened in this case? What was the reaction of the Senedd? Did it feel that Wales should have pushed for more than 5 per cent?

As in the Scottish case, it was a Government-to-Government agreement, the results of which came out when a report was published. There was relatively little parliamentary input to that. Because it was such a departure from the Scottish agreement, it was somewhat unusual. There was no baseline from which to analyse the 5 per cent needs factor, because that was a totally new innovation compared with the Scottish agreement. There was not a huge amount of discussion in the aftermath about what that would mean. We now know that it means quite a significant amount of resource going into the Welsh block grant compared with a situation in which tax devolution had not happened and the fiscal framework had not been agreed.

That is great—thank you.

I will move on to a slightly different area. In the previous evidence session, there was quite a lot of discussion about the amount that we can put into and take out of reserves, and how much flexibility we have at year end. In Scotland, we are certainly feeling constrained by all those things. During Covid, we were not allowed to switch any capital spending into resource, even though that might have made sense. Are those issues relevant in Wales, too?

Guto Ifan might want to come in on that.

Yes, that has been a concern in Wales, especially during the past two years with the massive in-year changes to the Welsh budget. Especially before the funding guarantee last year and this year, the Welsh Government did not have a heads-up on how much it had to spend. From our perspective, in scrutinising the Welsh Government’s response to Covid, it was very difficult to understand the Welsh Government’s fiscal firepower, which made it difficult to assess whether the Government was doing enough and was fully utilising what it had. To a large degree, the Government did not know that, which meant that it perhaps had to keep funding in reserve to allocate later in the financial year.

Leaving aside Covid-19 over the past two years, a few years ago, we submitted evidence to the working group for the Scottish fiscal framework review in which we compared the reserve powers and drawdown levels, and the current borrowing powers, including capital borrowing powers, between Wales and Scotland. We found that, in absolute terms, the amounts were smaller in Wales. However, compared to the size of the devolved revenues, the Welsh limits seemed slightly higher, or significantly higher, than what Scotland had. If you compare the limits to the total revenues, and if you think that the budget management tools and borrowing should reflect the devolved taxes that you have, there is certainly a case for increasing the Scottish limits, based on the Welsh fiscal framework agreement.

The different fiscal framework agreements happened at different times and with different chancellors, who had different fiscal rules and different focuses on borrowing. There has certainly been a massive change in the Treasury’s fiscal rules from 2016 to now. Maybe that will be reflected in the review as well.

That is helpful. I will leave it at that.

I, too, thank the witnesses for their written submission, which is extremely interesting. In particular, I am interested in figure 3.4, which shows the impact of applying the different models to Scotland. I wonder whether the Scottish Government needs to be careful about what it wishes for. In discussions on the fiscal framework and the Welsh example, the assumption often is that indexation for separate bands would be in addition to the existing IPC model, which would of course reduce the negative consequences. However, the grey dotted line in the graph in figure 3.4 shows that applying that model but using the comparable method would actually result in our being significantly worse off.

The points that you have just discussed with John Mason also need to be added into the mix. The issue should not necessarily be seen in isolation from the Barnett formula and the block grant. In a sense, John Mason was talking about the endpoint that the Barnett formula works towards for Wales but, if we look at what the Barnett formula currently delivers for the nations, we see that, in essence, Scotland ends up being the best off from that side of the equation.

Should we be careful about what we wish for, given that, with the application of indexation of separate bands, we could end up having the comparable method? Is there a risk for Scotland in a re-examination of Barnett and how it applies to us?

Yes. The choice of the block grant adjustment method is the most important issue for the whole upcoming review negotiations. That applies at macro political level as well as to technical aspects that we can forecast in reports such as our written submission. The interesting thing that strikes me in that graph and in the analysis of the different methods is that they show the relative degree of risk on population versus the risk on the composition of the tax base. Although Wales has a relatively slower growing population than that in England, historically, that slower growth has not been as pronounced as it has been in Scotland. In contrast, Wales has a weaker tax base that is more dependent on the basic rate than is the case in Scotland.

Therefore, there is a different balance in thinking about where the most important adjustment to the block grant adjustment is. If you were to negotiate the continuation of index per capita and add to it the three separate BGAs for income tax that exist in the Welsh model, that would of course be the most preferable outcome for the Scottish Government and the Scottish budgetary perspective.

In any negotiation, the Treasury will ask about the differential treatment of population—that certainly came up in the Welsh negotiations. That is one of the reasons why the comparable model was so important to the Treasury. As a result, adjustments were made that benefited the Welsh side of the ledger in those negotiations.

12:15  

That is helpful. In the interests of time, I will not ask any other questions, but Guto Ifan might want to add something.

On a wider level, it would be beneficial if reforms happened on a UK-wide basis with a from-first-principles approach to the spending and revenue sides. However, that was not done. Pretty much ad hoc arrangements were bilaterally negotiated across the Scottish Government, the UK Government, the Welsh Government; in Northern Ireland, the Fiscal Commission is now going down the same route. There is no UK-wide thinking through of how needs should be assessed and what spending levels the block grant should reflect.

The needs-based factor in the Barnett formula for Wales was almost an admission by the UK Government that needs vary across the country, but the UK Government did not reform the Barnett formula for Wales in a way that made sense across the board. There is an arbitrary figure of 5 per cent for a transitional period that could well last for decades. The needs-based assessment figure of 115 per cent was based on the Holtham commission’s report, which used data from the 2001 census. There has certainly not been an update of Welsh need.

To go back to the point about the Scottish Government being careful about what it wishes for, one of the Smith commission’s principles was that the Barnett formula should be kept. In an ideal world, that would be looked at in the round but, because of the vow around the Smith commission principle of keeping the Barnett formula, you are stuck with it, in a way. That inhibits the discussion of block grant arrangements on underspending and tax.

I thank Ed Poole and Guto Ifan for their submission, which is really helpful. Figures 3.2 and 3.3 drew my attention. What is happening in Wales is almost a mirror image of what is happening in Scotland. You have mentioned that many risks that Scotland has are outwith our control. The decline in oil and gas has been mentioned, but I guess that there could be other factors that would draw the figures down. Is that right?

Yes. Currently, what is driving the income tax growth in the out years in the Welsh forecast is relatively unclear. That is partly to do with the relative protection of the public sector in the pandemic—“relative” is the operative word—and partly to do with the policy changes that are not yet built into the Scottish projection, because the Scottish Government has not given a long-term forecast of where the thresholds and policy changes might lie. The higher rate threshold and the personal allowance threshold have been frozen, and Wales’s thresholds match those in England, of course.

Relative to Scotland, there is forecast growth in the out years, but the OBR is relatively unclear about what is driving that. One of the benefits of having an organisation such as the Scottish Fiscal Commission is that it can do a lot of bottom-up work in analysing the tax base at a granular level. The OBR is much more of a top-down forecaster. It considers the Welsh forecast, but that is less bottom up and granular than the Scottish case.

The Welsh Government has talked about doing a detailed investigation of the trends in the future—the out years—because, as you can see in the graph in figure 3.2, they are important in driving a projected benefit from tax devolution.

There is more certainty at the moment about positive change on the land transaction tax, which is our replacement for stamp duty. There is faster house price growth in Wales partly because of the pandemic—a lot of people want to buy homes out of the big metropolitan areas of England—and partly because of policy changes that are driving quite a lot of the growth relative to England and relative to a situation in which the taxes have not been devolved.

It is good to hear that the Welsh Government is doing a detailed analysis to try to understand those figures. I guess that the Scottish Government should do something similar to try to understand why the figures here are going in a negative direction.

Sure, but you have the Scottish Fiscal Commission, which we do not have in Wales. The detailed, granular work is done in Scotland by the Scottish Fiscal Commission. The Welsh Government does not have that resource, so the Treasury has to be involved in such analyses. Detailed analysis is important for understanding the impact on your budget not just in the current couple of years but in 2025, 2026 and 2027.

That might also inform policy changes to try to stop the reduction in budget that we will receive over the next few years.

Sure. Some elements are outside the Scottish Government’s control and some are within its control. The question is whether those are policy areas that lend themselves to long-term planning, such as skills, education and migration.

I thank Ed Poole and Guto Ifan for their submission. I echo the sentiments of everyone else on the committee: it really is excellent.

How have the different BGAs per income tax band been beneficial for Wales, given what you have outlined about the difference in your tax base? What are the primary benefits of that system going forward in the light of post-Brexit immigration restrictions?

That is quite a general question. It is fine for whoever wants to answer to do it.

Perhaps we can both give our views on that point, because it is really interesting.

As we mention in our submission, the three block grant adjustments for income tax are fundamental. They are the only reason why we see positive growth in our income tax relative to the block grant adjustments.

Income at the basic rate is important in Wales, so the block grant adjustments are a way to protect Wales’s budget from the impact of changes that have a disproportionate effect and that are outwith the control of the Welsh Government, because the thresholds are the UK Government’s responsibility. A UK Government decision could have a direct negative effect on the Welsh tax base were it not for the three block grant adjustments that take into account the different make-up of the tax base. That is important because, as Wales, like Northern Ireland, is a relatively poorer part of the UK, there was a lot of concern about what more reliance on our own-source taxes would mean for the budget. The BGAs are a way to ensure that tax devolution can work fairly in Wales.

Migration is very important for the overall pattern of the tax base. Free movement of people often resulted in young people moving to the big metropolitan cities of England. Migration to Wales was lower in percentage terms than that to the big metropolitan cities of England, so there might have been a short-term benefit relative to England as a result of free movement issues. However, that will not last into the future. Migration policies will be very important for the skills base, education base and tax base.

One of the main reasons why Wales has a poorer tax base is its lower productivity relative to the rest of the UK. Reducing the productivity gap and raising wages is the key to raising Wales’s tax base in the long term, and I am sure that that is true in Scotland, as well. Although we have relatively small short-term adjustments when we look at the impact on the block grant adjustments to what we are raising in revenues in Wales, migration and the retention of skills—[Interruption.]—very important in the long term.

Who has not fed their dog? Is that you, Michelle? It is Michelle’s dog—I wondered why she had disappeared from view. Are you back, Michelle? Are you joining us?

[Inaudible.]—and everybody for the barking that came out of nowhere.

Thank you for that, Ed. Do you have anything to add, Guto? That is my only question.

On the benefits of the separate block grant adjustments, it is a bit too early to say for Wales, especially because we have had only one year of outturn data. We can say that we modelled what the effect would have been if the measure had been in place since about 2010—it would have cushioned a lot of the blow to the Welsh tax base had tax devolution been in place in previous years.

It is interesting that, since full tax devolution in Scotland in 2016-17, the relative growth in the additional rate in England and Northern Ireland has been a lot higher than in the rest of the tax bases. That growth at the very high end has been driving a lot of the growth in the overall tax base in England and Northern Ireland. Having separate block grant adjustments would have cushioned the Scottish budget somewhat during the past four years, because the additional rate tax base is a smaller share of the overall Scottish revenue. Even if higher incomes in Scotland had grown in line with those in England or Northern Ireland, that would not have had the same effect on total revenue growth. Therefore, it would have been beneficial for the Scottish Government to have the measure in recent years.

Thank you.

That exhausts questions from committee members, but I have one to finish off. Section 2.3 of your paper lists the four taxes that Wales has gained in the past seven years: full devolution of non-domestic, or business, rates, the land transaction tax, the landfill disposals tax and the Welsh rates of income tax. If we look at the entire UK tax structure, a huge amount is still reserved to the United Kingdom, such as savings and dividends, income tax, VAT, national insurance, fuel duty, alcohol and tobacco taxes and inheritance tax. Where does Wales go from here? Is Wales settling in with the taxation that it has and seeing how things progress, or is it looking at whether it can devolve further taxes?

That question intersects with politics, naturally, as it involves questions about the relative balance between reserved and devolved powers, both in the tax area and more generally. The Welsh taxes that you listed would be very familiar to a Scottish audience. Even though in Wales we have had our own commissions, the recommendations that have subsequently been implemented by law have tended to follow the models that were laid down a few years prior in Scotland. In many ways, the range of possibilities for tax devolution in Wales is limited by what has gone before in Scotland. I would say that there is a Scottish slipstream, in the sense that, although Wales will not necessarily gain all the powers that have been devolved to Scotland, it sets the menu of options that might be chosen in the Welsh context.

12:30  

In Wales, politicians of various persuasions have been looking at income tax powers. Scotland obviously has much more extensive income tax powers, although it is constrained by the block grant mechanism that we have been talking about. Consideration is being given, in particular, to changes to the thresholds and the fact that all non-savings and non-dividend income is devolved to Scotland. That is the logical next step in the Welsh tax devolution journey, although the discussions on whether such developments occur are, of course, entirely integrated with politics.

If you were to model an income tax system, you would not pick the bands and thresholds that Wales has at the moment. Given the relatively lower incomes in Wales, the UK thresholds do not match how you would model a progressive tax system. Therefore, one issue that is being considered is thresholds, and the convener mentioned others that have been evaluated in the Welsh case—corporation tax, VAT, the aggregates levy and air passenger duty are often mentioned. However, in the context of the difficult intergovernmental relations between the countries of these islands at the moment, that is very much a political question as well as a fiscal one.

Yes, I do not want to focus too much on political issues. Scotland might get powers over corporation tax and a share of VAT assignment, and I am quite keen on getting powers on alcohol and tobacco duties, which would certainly help, given some of the issues that we have to deal with on those matters.

I thank Ed Poole and Guto Ifan for their evidence. I apologise for delaying you by some 30 minutes because the previous session overran. I appreciate your patience.

The committee will consider the next steps on this work once the UK and Scottish Governments agree the terms of reference, timetable and authorship of the report. In fact, we might invite Ed Poole and Guto Ifan back. We did not spend a lot of time asking questions, mainly because their submission is so detailed, which meant that we already had many of the answers before us, so I thank them for that.

12:32 Meeting continued in private until 12:41.