Finance and Constitution Committee

Meeting date: Wednesday, December 19, 2018

Official Report 741KB pdf

---

Budget Scrutiny 2019-20

Good morning and welcome to the 32nd—and, I am glad to say, last—meeting in 2018 of the Finance and Constitution Committee. If people with mobile phones could put them into a mode that will not interfere with our business, I would be most grateful. Actually, I had better do the same.

Today’s round-table evidence session will focus on the Scottish Government’s approach to taxation. We will take evidence from Mark Taylor, assistant director, Audit Scotland; Charlotte Barbour, director of taxation, Institute of Chartered Accountants of Scotland; Alan Bermingham, policy and technical manager, Chartered Institute of Public Finance and Accountancy; Dr Angela O’Hagan, chair, equality budget advisory group; David Phillips, associate director, Institute for Fiscal Studies; Dr Alison Hosie, research officer, Scottish Human Rights Commission; and Joanne Walker, technical officer, Chartered Institute of Taxation. I warmly welcome everybody who has come along to help us in our budget deliberations.

This is the first of our sessions on the Scottish Government’s budget. We will be doing a lot more in-depth scrutiny in the new year with not only the cabinet secretary but the Office for Budget Responsibility and the Scottish Fiscal Commission, but I am grateful that the witnesses are here today to help us. I also thank those who made written submissions.

The round-table format is intended to be as free flowing as we can possibly make it. As we go through the morning, committee members will make a number of points to try to tease out some of the issues that are important to them, but we have laid out four distinct areas to cover. Inevitably, with this type of discussion, those areas will merge into each other at some stage, so I will make a judgment as we go on about whether a particular area has been exhausted.

The first of our four themes is tax policy and the differences in that respect between Scotland and the United Kingdom. I invite Murdo Fraser to begin.

I want to ask about the complexity of the budget and tax policy divergence. As Audit Scotland has commented on those issues, I give Mark Taylor a bit of notice that I might come to him first.

That is much appreciated. Thanks. [Laughter.]

One thing that the committee has identified over the past couple of years is the increased volatility and uncertainty arising from the operation of the fiscal framework. Previously, we dealt with a fixed budget; if you took a pound from one place, you had to find another pound somewhere else. However, the budget now has a lot more moving parts; for example, we have the fiscal framework projections for tax receipts, which have to be added in. We note in the budget that income tax is down a bit on the previous year, but that is balanced by the block grant adjustment, which again has to be compared with previous years.

Overlaid on that complexity is an additional level of complexity—the issue of tax policy divergence. There has been commentary in relation to last week’s budget about what that means for non-savings, non-dividend income tax and the way in which the rates in Scotland will differ from the rates in the rest of the UK. Given that we are already dealing with volatility and uncertainty, to what extent does this increased tax policy divergence increase that volatility and uncertainty? Is it just something that we have to live with, because that is the position that tax devolution leads us to? How can we best try to reconcile all these moving parts?

As you have said, there are a number of moving parts in the budget, and one of the key things is to recognise that what ultimately matters is their overall effect. Of course, there is a lot of value in drilling down into and exploring some of the different elements, but it is the overall effect of bringing those things together that ultimately matters.

With regard to the divergence in tax policy, it is the interplay between and divergence in the approach of the two Governments that have an effect. The choices made by the UK Government and the Scottish Government affect the budget’s overall impact, but it is hard to see in the budget how that builds up through time and over the years. In assessing the overall impact, the Scottish Government has put forward the figure of £500 million, but the fact is that the overall impact builds up through time.

With regard to your comment about volatility, the way in which the system works means that additional complexities arise from different tax rates being set, and forecasters have to assess that. Ultimately, however, whether there is divergence or status quo between the two Governments, that volatility will still exist in the system, because it is driven by factors such as economic performance and the forecasting approach. They are still in the system, irrespective of whether there is tax divergence.

There is an additional layer of challenge in the forecasting process, with the additional set of assessments about the relationship between tax decisions made by both Governments and the impact on the economy. Those questions are part of that wider question. However, sticking with the status quo—in other words, having the same tax levels across the two Governments—does not remove that volatility or uncertainty.

Who else would like to chip in?

I would probably echo those comments. The Audit Scotland report that was produced a wee while ago contains much the best explanation that I have seen of how all those moving parts, which are part of the system, fit together.

I am not convinced that the tax policy introduces a huge amount of volatility. Because it is part of the UK system, there are quite a lot of constraints and challenges to what you can do—but we will discuss that later.

On the overall budget risk, it is important to look at how all the moving parts, taken together, affect the budget, but it is also important to break the different moving parts down in order to understand what is driving it. For instance, is any underperformance or overperformance due to a difference in the block grant adjustment—in other words, due to things happening elsewhere in the UK—or is it due to what has happened in Scotland, which is about the actual revenues? If we look at that, the proper Government can be held to account for the impact of the different decisions.

From a budgetary risk perspective, what matters is how it all fits together, but from a budget scrutiny and accountability perspective, what matters is how all the individual parts move. That is a challenge, because there are so many different moving parts, such as the block budget for the current year, reconciliations for previous years and so on. For most people, that is too much information, but it needs to be there in some way for the likes of me and you.

“Too much information” is a nice description.

If I was looking at budget risk as a factor and wanted a better understanding of how all the parts fit together, from a scrutiny point of view, I would be trying to understand some of the sensitivities in the modelling. In the recent UK Government budget, the chancellor spent the extra money from income tax on the health service, but most commentators agree that there was no full understanding of where that extra tax revenue came from. We knew that it came from a range of taxes, but there was no clear indication of what drove it. As it happens, the chancellor has now spent that money on the health service—but without a clear understanding of what drove those increases.

If we apply that logic to the situation in Scotland, I think that understanding some of the sensitivity in the modelling—for example, whether it is sensitive to economic growth factors, the number of taxpayers and so on—and building up that kind of sensitivity analysis through the data would help with understanding and with the scrutiny of the key risk factors. We need a bit more drilling down to get a better understanding of risk and therefore what is driving the changes and where the budget volatility is likely to come from.

How could the Government tackle the sensitivity issue?

If the Scottish Fiscal Commission is modelling the tax revenues, I think that playing around with some of the key determinants in the model will help you understand the impact of, for example, lower taxpayer numbers over the years, lower economic growth or whatever. Modelling that for each element will show the impact on the budget and from that, you can determine whether the budget is more sensitive to one factor or another. If we can see that coming through forecasts—if, for example, the restraints on immigration after Brexit mean that the taxpayer growth numbers are not going to come through—we can then say that that will be a key risk factor in the level of income tax that we are budgeting for.

I want to follow up David Phillips’s point about income tax reconciliations, because they are an important factor. I note that the report produced last week by the Fiscal Commission projects income tax reconciliations from last year’s budget, which will then apply to the 2020-21 budget, as being £145 million down, while for the current year—2018-19—it is forecasting outturn to be down by £472 million. That is a big chunk of the Scottish budget, and it would kick in in 2021-22. I know that these are only forecasts and that we do not know whether the figures will work out until we see the outturn, but is there a danger that, with our current approach to the budget, we might be inadvertently storing up huge budget problems for the next Parliament?

The idea behind the framework is to give the Scottish Government a chance to adapt to changes in the forecast. First, the outturn information is not available for a significant period, and that gap gives the Scottish Government time to prepare for any adjustment required in the budget, whether that be paying back spending, drawing down Scottish reserve or changing tax policy.

One of the issues with the Scottish fiscal framework is that there is less flexibility for the Scottish Government to deal with those sorts of shocks to revenues than there would be if the powers to borrow and save were freer. That is how the fiscal framework has been set up. I am not sure that I would go as far as to say that it is storing up problems for the future, but it certainly means that, given the existing fiscal forecasts, the Scottish Government needs to constantly look ahead and consider what it is going to do two years down the line to adjust to the forecast changes.

We are talking about only one side of the equation at the moment. On the other side, we have the OBR. Does that not introduce even more volatility into the process?

10:15  

It depends on which way the forecasts change. For instance, if the Scottish Government overestimated its revenues, Her Majesty’s Revenue and Customs would have paid too much revenue to Scotland in the year, and there would be a negative adjustment. However, if revenues elsewhere in the UK were overestimated, too, the block grant adjustment would have been too high and there would be a reconciliation payment to reduce it. If the OBR and Scottish Fiscal Commission forecast errors go in the same direction, they tend to offset each other.

It is when the forecast errors go in opposite directions that there is an issue. If the OBR overestimates the block grant adjustment—sorry, does that make sense? Is that the wrong way round? [Laughter.] Is it that, if the OBR has overestimated things, the adjustment is reduced? Let me get this right—it is just so complicated.

If the Scottish Government has overestimated revenues, there will need to be a downward adjustment in subsequent years to take some money off the Scottish Government. If the OBR has underestimated revenues in the rest of the UK, too little will have been taken off, so more will need to be taken off the Scottish Government. Therefore, it is when the Scottish Government overestimates and the OBR underestimates revenues in the rest of the UK that there is a particularly big risk. The issue is when the forecasts go in opposite directions—when there is an overestimate and an underestimate. When they go in the same direction, there is less of a problem.

Sorry about that, convener.

Actually, that was quite helpful. You are describing a situation in which we have the OBR on one side and the SFC on the other—perhaps “side” is the wrong word—and, because of forecast errors rather than any policy changes, the Government might find its budget going significantly up or down.

It depends on how you view the issue. Ultimately, one would want the Government’s budget to be determined by outturns—in other words, what has actually happened to revenues in Scotland and in the rest of the UK. There will, of course, be forecast errors, which can mean additional volatility in the current system. In effect, the volatility is delayed. Things are initially based on forecasts to give the Scottish Government a degree of certainty to plan during the year, and then, when things turn out to be different, there are adjustments at a later date. That is important, given the Scottish Government’s constrained borrowing powers to deal with changes in forecasts in year. Because of the constrained borrowing powers under the fiscal framework, not having that kind of delayed system could pose more rather than less of a problem to the Scottish Government.

Convener, you are absolutely right to say that there are two sides to the equation. On the figures that were quoted, we have forward forecasts from the Fiscal Commission and the OBR, and those figures are the net position—they take the two forecasts into account.

As for the original question, which was about storing up problems for the future, Audit Scotland has suggested that there needs to be an awareness that the volatility is built into the system and that, in the decisions that Government and Parliament make, a longer-term and more strategic approach is needed to take account of that. As well as looking at the detail of the year-to-year effects, there needs to be consideration of the extent of the volatility over time and how to design tax and indeed spending policies that can respond to the volatilities. For example, there is potential to think about where in the spending programmes there is an opportunity to turn the tap on or off a bit without there being general upset or without undermining the way in which public services work and are set up. There is an opportunity for the Government and the Parliament to think ahead more strategically about how all that plays through.

At present, we are in the very early years of this. We have an assessment from the Fiscal Commission that the underlying error level on its side could be about £0.5 billion a year. That is small in the overall scheme of things, but it is still a big number. We would expect a similar error level from the OBR estimates through time. Fundamentally, as has been said, it is whether those error levels amplify or offset each other that will give rise to issues or opportunities.

We often think of the matter in terms of downside risk and what will happen if there is less money. The forecasts that we have come to suggest that there might be less money for the years ahead, but forecasting risk is unbiased and it can go either way. There will be other years—I say this with confidence, but we will see whether it plays through—when the risk goes the other way and the Government and Parliament have the challenge of having more money than they expected.

We cannot depend on such effects because they are not recurring—they are one-off year effects. The question is how spending and taxation policies are determined in order to take account of the volatility that we know will be in the system over the longer term.

Before I bring in Tom Arthur, who wants to contribute, I ask Mark Taylor to pick up on what David Phillips said about constrained borrowing powers to help deal with volatility. Does Audit Scotland have a view on whether the process is constrained?

Our view is that the rules are the rules, and the challenge for the Government and the Parliament is to operate within those rules. As you would expect, we do not have a view on the policy. What I can say, though, is that it is challenging. There is not a lot of room for manoeuvre built into the fiscal framework agreement. In time, as we get more of an understanding of the actual level of volatility, there will be an opportunity to reflect on that through the review.

That was a very diplomatic answer.

You would expect nothing less, convener.

Absolutely.

I will reflect on some of the language that we have heard in the first 20 minutes of this discussion, including “moving parts”, “volatility”, “challenges”, “inflexibility” and “problems around scrutiny”. I am keen to get views from our guests on whether those things are baked into the system and are inevitable consequences of the fiscal framework and the rules that we are operating under or whether, perhaps, five or 10 years down the line, when they have had a chance to bed in, we will be able to observe patterns and there will be more predictability. Will the Scottish Parliament always have to contend with this as a consequence of the way in which the suite of tax powers is arranged between Holyrood and Westminster?

I always expect there to be errors in forecasts, so there will always be a process of reconciliation. Over time, the Fiscal Commission will have more data—more outturns, in effect—to which it can compare its forecasts, and it will be able to look at the most sensitive assumptions around the earnings forecast, the employment forecast and the pensions income forecast. We might expect there to be some improvement over time in the Fiscal Commission’s forecasting capabilities, but forecasts are fundamentally an educated guess, so there will be a degree of error.

I assume that the Scottish Government and the Scottish Parliament will become more au fait with the process, which will make it easier for committees and the Parliament more generally to engage in the process. I also expect there to be a more educated public and media, so people will be able to engage on that side as well. However, I do not see the process suddenly becoming a whole lot easier.

The process exists in order to provide a degree of insurance to the Scottish Government. In effect, the block grant adjustment, the reconciliation process and so on mean that the Scottish Government bears the relative risk of its revenues moving less or more quickly than those of the rest of the UK, but they also insure the Scottish Government against broader macroeconomic shocks. It is a complex situation, because they are trying to do the somewhat difficult task of providing the insurance that Scotland needs because it does not have the borrowing powers.

An alternative would be to give Scotland significantly more borrowing powers and significantly less insurance. Whether the UK Government would want to do that and whether the Scottish Government would want to have that additional risk, given that many of the risks would be outside its control—it does not have levers on monetary policy and things like that—are other questions. However, the current set-up for what the fiscal framework is trying to do means that we inherently have those complexities.

I will comment not on the fiscal framework or the forecasting side but on the Scottish income tax policy side. While income tax remains a shared tax, there will obviously be a lot of constraints on what can be done. That is partly because it is almost a natural or automatic impulse to make comparisons with the policies that are active in the UK. However, income tax is naturally intertwined anyway, so the Scottish Government’s decisions are impacted by, for example, the choice of personal allowance and its level. With better education of the public, it might become easier to diverge and make that more acceptable so that there is a less automatic comparison between UK policy and Scottish policy. However, while it is a shared tax, there will automatically be that constraint.

That is a good point.

I agree with David Phillips’ point that forecasting as a process is never going to be an exact science but will be refined over time as data improves. In Scotland, you are in a particular period of uncertainty at the moment that is perhaps not helping certain factors. Under normal circumstances and with time and improvements to forecasting, you would expect the situation to improve. Obviously, borrowing powers come at a cost, and increasing your borrowing powers increases what you will spend on repaying that borrowing by way of interest out of your normal day-to-day resources. That is a kind of constraint to bear in mind regarding what is affordable.

The issue of managing volatility falls a bit more into the idea of holding reserves. If the impact of forecasting errors and so on can go both ways, there is also an issue about what you do when you have an upside. Do you tuck that away as a reserve? Do you replenish the reserve or whatever and use that to manage the volatility going forward? You need to look at using a combination of tools rather than think that borrowing is your way of managing volatility.

That is quite helpful.

I will pick up on that last point about having a greater suite of tools. Could we increase the limits or caps on what can be drawn down from the reserve alongside having increased borrowing powers? Joanne Walker referred to the interaction with UK-controlled aspects of taxation such as the personal allowance. Another key tax that has a very significant interaction with income tax is national insurance. If, as David Phillips suggested, we moved towards a position of having greater responsibility with less insurance, would that necessitate the Scottish Parliament having an increased range of tax powers along with increased powers over borrowing and more flexibility with the reserve to manage that risk?

Yes. CIPFA would argue that having wider tax powers would benefit you. Obviously, a large amount of your income is tied into one particular tax that you have control over, so volatility in that tax perhaps impacts you more than it would if you had a wider range of tax powers. That position would require increased devolution and so on, so that is a different matter. However, I certainly agree that a wider range of tax powers would help to mitigate some, but not all, of the risk.

National insurance would have many of the same risks as income tax. If you were thinking about diversifying risks, you might want to have taxes with somewhat different bases. You have part of VAT already, but the big one that you do not have part of is corporation tax. However, that has a hell of a lot of complexities around devolution, as they are finding in Northern Ireland.

I have a final question about national insurance and the interaction with income tax, on which I would be keen to hear from Joanne Walker. Clearly, the tax is set UK-wide, but the upper earnings limit has been paired with the UK’s higher rate, which can have an interaction in Scotland that we do not have control over.

In a sense, it is a way of trying—by the back door—to influence decisions that the Scottish Government would take. Would there not be a case either for the UK Government to engage more closely with the Scottish Government to set a differentiated rate in Scotland or for that tax to be devolved? I do not mean to put you on the spot.

10:30  

I do not think I can comment on the policy choices. We will end up with a band for employees of £6,570 that will be taxed at a 53 per cent joint marginal rate and a same-sized band for the self-employed, for whom the joint rate will be 50 per cent. It is a significant kink in the effective marginal tax rate for those earners, and I believe that it will affect about 120,000 taxpayers.

The link is interesting. When the UK Government passed its budget in October, people commented on the fact that it did not mention that, in raising the higher-rate income tax threshold to £50,000, it was also raising the upper earnings limit and increasing the amount of national insurance that the rest of the UK taxpayers would pay. It is a complex issue to decide. If you had more control over that threshold, it would help, but, equally, because of the various links, a lot of things would need to be unpicked.

I apologise. I did not mean to put you on the spot with a policy question; I just wanted to get a sense of how the two taxes interact technically. Normally, any decisions that were taken about one tax would be taken with a view of how it interacted with the other.

We would certainly like to see a lot of collaboration between the UK and Scotland in terms of those interactions between taxes and national insurance, as well as more generally, to make sure that taxpayers are not adversely affected or, if they are affected, at least everyone is aware of those effects.

I will add to the points that have been made about national insurance. You need to be careful about what you are comparing, because all taxpayers right across the UK will pay more national insurance. I think that it has gone up by about £340 because the higher-rate threshold went up south of the border and across the piece. You must be careful about whether you are comparing income tax with income tax or with net take-home pay, where you have one of the levers.

We talked about the complexity of balancing lots of moving parts, and that is an illustration of that. It is interesting that, because the powers around income tax are being only partially devolved to Scotland and the whole system is still intertwined—not just income tax, but how income tax interacts with other taxes such as corporation tax, NI contributions and capital gains tax—all of those taxes meet and one must be careful about how all those bits rub together.

Changes in the Scottish higher-rate threshold shine a light on difficulties with the overall UK income tax system. I do not think that the difficulty sits with the Scottish Government per se—what you do is what you do, and that is income tax. The Scottish Government cannot affect national insurance; there happens to be a knock-on consequence here, but everybody in the UK will pay more national insurance—it has nothing to do with the Scottish Government, although there is a flipside here, because you have changed the higher rate, and higher-rate income tax and NIC are married together in the rest of the UK.

As Joanne Walker mentioned, when the UK budget was announced in October, everybody said that the higher-rate threshold had gone up and that we would pay less income tax. However—it is like David Phillips’s story—although someone might pay less income tax, if the national insurance threshold goes up, too, they pay more national insurance, so the two net off against one another. That is not very visible across the UK.

Throughout my working life in tax, there have been calls to amalgamate NIC and income tax. One of the interesting things that we need to look at—not here today, but much more broadly across the UK—is whether the two of them sit together or whether you should hypothecate them more, so that NIC does whatever and income tax in Scotland does health and education plus.

I will pick up on a point that David Phillips made. We do not talk much about VAT, and this segment of the conversation is about volatility. What do David Phillips and others think about using the partial devolution of VAT, which is still under way, as part of the toolkit for managing the volatility in and around taxation on income?

If I understood him correctly, David Phillips was saying that, if there is concern about the volatility that arises from taxation on income, we do not necessarily want to devolve more tax powers over income. However, we are not devolving powers only to tax income; we are also at least partly devolving powers over taxation on spending. I recall that part of the thinking behind that was that income and spending patterns can be counter-cyclical, which provides a way to manage the volatility.

To what extent do members of the panel think that, when the partial devolution of VAT comes on stream, it will be an effective—although limited—tool in the Scottish Government’s box for managing the volatility that we have talked about, in addition to borrowing powers and other measures that we have discussed?

When VAT assignment—rather than devolution—comes on board, two factors will be going in different directions. VAT involves a different tax base; its risks are not perfectly correlated with the income tax risks and, when two taxes are not fully correlated, they will offset each other to an extent in terms of volatility. However, more revenue will come from volatile streams than from the block grant. Having two different taxes with different risk profiles should help to reduce the risk, but the fact that more of the overall budget will come from risky tax streams than from the block grant will tend to increase the risk in the budget.

Given that VAT will be assigned rather than devolved, it will not provide the same scope as income tax to change rates or policy in order to change the revenue yield or have different distributional impacts. More diversified tax bases involve less risk than a base of one tax on its own, but the fact that the overall budget will depend more on tax revenues will tend to increase risk. I am not sure how those two things will play out, but my gut reaction is that risk will tend to increase overall.

I have a question that relates to divergence issues. The equality and fairer Scotland statement that was published alongside the budget said:

“there will be a separately published distributional analysis of income tax changes across income groups and with respect to age, gender and disabled people.”

Were any of our panellists involved in that analysis? Has it been published yet?

The budget review group’s report recommended that more distributional analysis should be done. A seminar was held in October, a position paper was published and work is in progress to develop that. At the seminar, officials from the Scottish Government and UK Government departments explored data needs, data analysis, approaches to data, gaps in the data and what needs to be remedied to achieve effective analysis in relation to the equality characteristics that have been identified. I understand that that work is going forward.

It is helpful to understand where the work has got to; it is not quite at the full publication stage.

I will build a little on the helpful, though complex, discussion that we have had, in which volatility has been characterised as the moving pieces of decisions that different Governments take, economic performance and the good old science—not—of forecasting. I am interested in hearing our guests’ view on the statement in the Scottish Parliament information centre paper that

“Over the years since 2016-17, the Scottish Government has adopted different income tax policies and the economies of Scotland and”

the rest of the UK

“will have performed differently.”

We have touched on this already, but it seems to me that, even if Scotland adopted the UK tax policy, it might not generate enough in tax revenues to offset the block grant adjustments. To me, that diminishes any argument that we should just make the same tax decisions as the UK.

I am also interested in your views on the Fiscal Commission’s statement that

“changes in Scottish income tax policy and a more positive outlook for the economy since May have increased our income tax forecast.”

That leads me on to the point that our focus should be on the impact of what we do differently, particularly around the three Ps of productivity, population and participation.

I realise that there might be quite a technical or theoretical answer to some of that, but our guests who have views on inclusive growth might also want to input.

Who wants to take that as a starter for 10?

I will give it a go, convener.

In essence, there are three factors that matter: how the forecasts are made and the forecasting risks; how the economy grows and the impact of that over time; and the difference or similarity in tax policies between the two Governments. When we strip back all the complexity and variability, those are the three things that matter to the budget this year, in terms of the amount of funding that is available, and that will matter to the budget in future years because of the knock-on effect. The question then is which of those we can control and influence and on which we are passive recipients. There are a variety of responses to that in different parts but, ultimately, it is those three factors that matter, and that is captured in the SPICe paper and in the Fiscal Commission’s forecasts.

I will give an example to illustrate that. The Government has assessed that the differentials in the tax policies make a difference of about £500 million to this year’s budget. If we look at how that interacts with the block grant adjustment, we see that the net effect is a £182 million addition to the Scottish budget. The difference between those numbers is to do with those other factors about how the economy is performing relatively and how the forecasts play through. The committee might find it helpful to ask for a bit more detail on how that £500 million figure is calculated, as that might give a bit more insight into the make-up of the difference, which would help you to understand what we can and cannot control.

In answering Angela Constance’s question, it is worth going back to the two broad purposes of tax devolution. The first is to provide incentives to the Scottish Government to improve economic performance, grow the economy and boost employment. Those incentives operate even if there is no policy divergence. Even if the Scottish Government had the same policy as the rest of the UK, there would still be the incentive to grow revenues more quickly. Therefore, it would not be pointless to have devolution even if the Scottish Government did not change policy.

However, devolution also gives the Scottish Government the chance to vary policy if, for example, it wants to be more redistributive or progressive, or if it wants to raise more revenues or change how the tax burdens are distributed. That is another tool that the Scottish Government can use to boost the economy, but it can also help it to meet its other objectives. From what the Scottish Government is doing, it is clear that it values the powers to raise more revenues and to make the tax system more progressive.

It is still worth having in mind the system in the rest of the UK, for two reasons. The first is the political reason that, as we have a joint system, people will compare the two. I am sure that the Scottish Government will have that in mind, and the Welsh Government certainly has it in mind when it makes tax policy. It is concerned about being seen as being out of step with the rest of the UK and therefore being labelled a high-tax economy or something like that. The other reason is the issue of behavioural responses. That is not just about migration; it is also about the fact that, because the dividend tax in Scotland is not subject to the Scottish rate of income tax, there could be tax-motivated incorporations if there was a very big extra tax burden in Scotland.

I guess that what I am saying is that there is scope to vary policy, but you need to be careful in thinking about that and must consider the behavioural impacts. You do not always need to change policy to make having devolution worth while, because there are still incentives.

Of the three Ps that Angela Constance talked about, the most important for tax revenues is productivity rather than participation or population. You get more people into work, but a lot of their income will be tax free because of the personal allowance, which is really quite high now. Someone who is in a minimum-wage job is, in effect, paying no tax. So, although it is good from a social perspective to increase participation, that does not really get you much revenue any more, because there is such a high personal allowance.

Productivity, however, means increases in earnings for everyone, including potentially those who are right at the top of the distribution, and that is where you will rake in the extra revenue. Productivity is the most important of the three Ps for the Scottish tax base.

10:45  

As no one else wants to contribute, we will move on to the next area—the number of taxpayers, which fits neatly with what David Phillips has just said. James Kelly will lead on this.

One of the variables that feeds into the tax calculation is the number of taxpayers, particularly at the higher and additional rates, where the bulk of the revenue comes from. The forecast this time last year was for 337,000 higher-rate taxpayers and 18,000 additional-rate taxpayers. However, when HMRC produced its outturn report, it appeared that there were fewer taxpayers in those bands in 2016-17, with 296,000 at the higher rate and 13,300 at the additional rate, while last week’s SFC forecast for 2019-20 shows a slight increase but not to last year’s levels, with 327,600 at the higher rate and 18,800 at the additional rate.

If the number of taxpayers in those bands goes up, the income tax forecasts go up, too, and if they go down, they decrease. It is therefore important to get these forecast figures accurate in future. I am interested in understanding the drivers that feed into the number of taxpayers, particularly in the higher-rate and additional-rate bands. How can we build accuracy into forecasts for these figures?

That is a cracking question.

I will answer the easy part first and then have a go at the harder part.

On building accuracy into the forecasts, the baseline or starting position is now known. As further outturns are published each summer, that position is updated, albeit two years retrospectively—a bit of forecasting remains to be done in that respect—but immediately, we have seen corrections to the OBR and Fiscal Commission forecasts that offset each other almost entirely. As I have said, the baseline position is known.

As for what matters as we move forward, it is the same as with many of these themes: what is the interrelationship between the number of taxpayers across the bands in Scotland and the number of taxpayers across the bands in the rest of the UK, and how does that change through time? Can we grow the number of higher and additional-rate taxpayers faster in Scotland than at the UK level? If we can, that will have a positive effect on the tax take; however, the opposite will have the opposite effect. It is all about the movement through time against the base.

As people around the table will know, we have, on average, relatively fewer taxpayers in those higher bases in Scotland than in the rest of the UK, but we have more than there are in Wales. That is the starting point. What really matters is what happens from here on.

I concur. There was a downward estimate of the number of higher-rate taxpayers in the first year, but that was put into the baseline block on adjustment. What matters is how the numbers evolve, and that will depend on productivity growth among high-earning occupations and higher-paid people in Scotland; responses to Scottish tax policy, whether it be migration, reduction of work effort or tax-motivated incorporation; and any beneficial economic impacts of Scottish Government policy that might boost the numbers.

I will link my answer to this question to the questions that Ms Constance asked.

I certainly have no answer to how we boost the number of higher-rate taxpayers—I do not have a policy magic wand. However, from my reading of the budget, I see significant challenges from having a low-wage or no-wage economy in Scotland and the fact that 2 million people are not paying tax. That presents significant issues with regard to other policy challenges, but it also demonstrates the lived reality of people’s lives and the level of poverty and economic instability that they continue to live with. The current Scottish Government has made a number of policy choices, but there are issues about the clarity with which those policy choices have been presented across the budget in relation to accessing the labour market and so on. There are a number of important interventions, but they are scattered throughout the budget.

Colleagues have commented on the importance of productivity over participation, but I would point out that questions about participation in the labour market include the quality of that participation. What types of jobs are being created? How sustainable are they? What are the earnings levels? Again—and this relates to Angela Constance’s question about inclusive growth—certain spending decisions in the budget mean that proportionately smaller amounts of money are being allocated to the kinds of interventions that will boost earnings capacity through employment and skills. We have to think about who is and is not earning—the distributional analysis will be very important in that respect—and how we characterise productivity.

Although, in policy terms, the investment in the care economy is key to our economic infrastructure, it is not characterised in that way. For example, a Scottish Government publication on the rationale for investing in infrastructure talks about the positive benefits of education, health and other public services, but it says nothing about investment in care. With regard to participation and productivity, how the unpaid economy supports the rest of the so-called productive economy presents a real challenge, and it has to be made much more visible. Talking about participation only in terms of economic activity as it is currently defined is problematic and continues to mask issues around economic contributions that are not measured through earnings and the tax base. We have a way to go to work through the components of our economy and the barriers to inclusive growth. As far as inclusive growth is concerned, we must address the real challenges and unresolved conflicts, as they have been called, around care.

When we look at who is earning—and, indeed, who is earning what—we see that there are 300,000 fewer women taxpayers. That raises a number of challenges within what is already a very small tax base. Figures show that higher-rate taxpayers comprise 91,000 women and 275,000 men, and there continue to be all sorts of structural problems in the labour market and earnings that affect who is paying tax and at what level.

Did you want to comment, Alison?

Angela O’Hagan has just stolen all my points.

But they are worth repeating.

They are. From a human rights perspective, one of the key principles of the obligations of Government is to maximise available resources. However, when we look at our relatively small tax base, we see not a lot of movement—and not much potential for any movement—in terms of varying income tax rates to increase our tax revenues.

A huge and concerning number is the 2 million people who do not pay tax. Angela O’Hagan referred to the differences according to gender; when I read the analytical note that came with the budget, I felt that it presented that gender difference very much as a matter of fact. It was saying, “There is this percentage of men and this percentage of women who are or are not contributing in income tax.” It seemed to accept that as a fact rather than as something that should make us consider our budget policy choices.

If we are to address those 2 million people who are not contributing through income tax, we need to look at our budget policy decisions. Angela O’Hagan referred to the bits and pieces throughout the budget, but what is the aim of, for example, our extension to the childcare policy in Scotland? Do we want an outcome of that policy to be more women in paid employment? There is also the issue of productivity in relation to the types of jobs that we have and how we value them.

The data that we have is limited as far as distributional analysis is concerned, but the question is: do we use the data to ask the right questions? I do not think that we do at the moment.

I want to follow up on a couple of issues. There are connections between spending choices and tax choices. Although we are focusing on the tax side of things, I want to look at the proposed treatment of local government, where gender inequality in employment tends to be less than in the economy as a whole, including at the higher end of the income spectrum. Will the proposed reduction in core funding for local government affect not only the tax receipts to the Scottish Government from that area of employment but gender equality in terms of who is going to be employed? Obviously, the policy question about local government spend is a matter for another day, but the impact on tax receipts and the economic impact of gender inequality are relevant to this discussion.

I agree that productivity is the key driver here. Moreover—and perhaps David Phillips can keep me right on this—some of the devolved Governments of the UK are fairly interested in foreign direct investment and so on. The issue is the nature of the jobs that you are bringing or that you are encouraging others to bring to the marketplace and ways of attracting higher-wage jobs—those in the newer tech industries, for example—instead of there being just a focus on getting people into work. It is about having policies for the education that will support and attract those types of jobs, thereby improving the earnings level.

There is also the demographic factor. People are getting older, and some of those high earners are going to retire and disappear from the tax receipts. If foreign direct investment and bringing in new jobs are factors in the economy, policies in support of that will underpin the future growth of taxpayers in those brackets.

What about Patrick Harvie’s specific issue about the impact on local government?

I will address the income tax implications rather than the gender equality implications, which I think are complicated. The issue is all about the kinds of roles that are being taken out of local government and so on, and it is beyond my expertise.

As for the income tax impacts, I will make two points. First, this policy would still save the Scottish Government money. If the Government has to save money—for example, by making cuts—it will save money overall by cutting the funding to local government. Let us say that the Government saved £100 by cutting funding; it would lose income tax from the employees who lost their jobs if they did not find any alternative work, but that might cost it only £20 or £25, depending on where the people who lost their jobs were in the income distribution. Overall, then, the policy would save the Scottish Government money if it needed to reduce its spending.

Secondly, if the economy was operating at close to its full potential in terms of its output and there were vacancies, many of the people who might lose their jobs if there was a reduction in local government employment would be likely to find jobs elsewhere and would therefore be paying income tax. The question, then, is whether the individuals who might lose their jobs if there was a reduction in local government spending would be able to find other jobs. If they could, there might not be much of a reduction in income tax receipts.

This comes back to my earlier point about how income tax devolution is designed to incentivise the Scottish Government to find ways of helping people get into work, stay in work and improve their productivity. Perhaps now, even more than before, the Scottish Government has strong incentives to help back into work those who have been affected by the Government’s policies or other factors that affect the economy.

11:00  

I have a question that follows on from the comments that were made at the beginning and subsequently about who is paying what. The number of additional-rate taxpayers is set to go up by 25 per cent by the next session of Parliament, and the number of higher-rate taxpayers is set to go up by 10 per cent. Given that we are seeing negative behavioural change, with the current figure of £13 million, which I presume translates into a downward number, can somebody clarify how we will get those increases? As I understand it, that can happen in two ways, either by people moving upwards through the bands or by the bands being moved downwards to capture more people. Given that real earnings are going up by only 0.3 per cent and 0.5 per cent, can somebody explain how valid the 25 per cent figure is?

It seems not. [Laughter.]

Maybe we can package the question slightly differently. HMRC said that, in 2016-17, we had 13,300 additional-rate taxpayers, but the SFC’s forecast says that, by the time we get to 2019-20, that number will have grown to 15,800. What has happened to create that growth? That has been puzzling me, too. You are going to tell me, Mark, are you not?

Unfortunately, I am not. [Laughter.] No doubt that will be at the top of your list of questions for the forecasters. My observation is that those are thresholds, so someone whose earnings are just below the threshold and then move just above it will be re-counted as being above it. The general upward movement of income through time and how that sits alongside the forecasters’ underlying assumptions about how the tax levels might grow will affect that, and both of those things are tied back into their forecasts for the economy.

Going back to Patrick Harvie’s point, I will not go into the local government debate—I am not in a position to do that—but I note that the system has been designed so that spending decisions have a feedback loop into the public finances and how much tax is raised, and increasingly Governments will need to become more sophisticated about how they target those decisions and their understanding of how that works. I say that without making any comment on how that works in the current budget proposals. The system has been designed so that there is a feedback loop between expenditure and the income that is raised through taxes, and the discussion that we have had round the table today has illustrated where there might be some fertile ground for movement in expenditure decisions that might help with that.

Of course, politicians of all parties want the best. The new system gives a hard financial edge to some of that stuff and a hard financial set of motivations, alongside the motivations that we all have for things to be better, to help to play through the system and make those links between expenditure decisions and how much tax is raised.

On the point about the £150,000 threshold, it is important to look at nominal earnings growth rather than real earnings growth, because it is frozen in cash terms rather than in real terms like most of the thresholds. Between 2016-17 and 2019-20, we could see earnings growth of approximately 10 per cent in nominal terms.

Secondly, we could see differential earnings growth across the distribution, so we might see higher earnings growth at the top than at the bottom. I am not sure what forecasts the SFC has made of that or whether it has said what it has assumed about differential earnings growth. There is also the issue of income that is not counted as earnings—the main example being self-employment income for Scottish tax purposes—and whether that is growing at a differential rate.

As Mark Taylor said, the threshold is important. If quite a number of people’s earnings are just below the threshold, even small earnings growth will push quite a few people over it. If I were in the Scottish Fiscal Commission or the Scottish Government, I would look at the detailed income distribution around that point. I would look at surveys such as the survey of personal incomes and the annual survey of hours and earnings to see whether lots of people were in jobs that were just below the threshold, as they would be affected. Perhaps that has been done in producing the forecasts. If that has been done, a bit more information about what will drive the growth could be useful.

It has been well forecast to the commission that the question will come up when it next comes to see us in January.

I hope that the commissioners are watching.

They had better be watching—you are right.

I will revert to Patrick Harvie’s points about the implications for women in local government. Given that women are concentrated in low-paid jobs in local government, all the moving parts—to use today’s language—of the reforms and the restructuring that have taken place, and which continue in the remodelling of social care delivery and so on through integration joint boards, reinforce the importance of scrutiny at the parliamentary level. That is consistent with the budget review group’s recommendation that committees across the Parliament should be alert to those changes—in part, that provides the feedback loop that Mark Taylor talked about.

I do not know that I necessarily share David Phillips’s confidence in people’s likelihood of finding jobs elsewhere. What he said might be true, but we need to look at the character of the jobs. Charlotte Barbour talked about looking not just at the interrelated impacts of the basket of taxation measures but at the interrelated factors around living costs. One issue is the interplay between low wages and the UK welfare system, which has had a punishing effect on women and particularly women of colour, and the living costs that low-income workers experience. Travelling for work presents high cost impacts on low-paid workers. We must look in the round at the impacts of changes in who is employed in the public sector and particularly in local government.

I will pick up on what Angela O’Hagan and Alison Hosie said about women. Of the 56,000 nurses in Scotland, 87 per cent are female. Their earnings are at bands 5 and 6 and intermediate levels, so we can predict what tax we will raise from them.

I am interested in what Alan Bermingham said about immigration post-Brexit. A third of dairy farms employ migrant workers from Poland, Lithuania and Romania. Has modelling been done to forecast the situation post-Brexit if we do not have tax contributions from those folks? They are young, fit and healthy—they do not use the national health service—and they send their kids to local schools. Has modelling been done of their current contribution and of what might be lost?

Is anyone aware of any such modelling that has been done? Do we need to ask the SFC about that when it appears?

I am not aware of Scotland-specific modelling, but colleagues at University College London have done work to look at the fiscal costs and benefits of different groups of migrants, including post-2004 European Union migrants. They looked at the short-run impacts, which tend to pick up a net fiscal benefit—as Emma Harper said, such people are young, fit and healthy and they are in work.

Over time, the benefit has reduced somewhat. Back in 2004, the personal allowance was about £5,000, so most of people’s earnings, even in low-paid jobs, were being taxed. Now that the personal allowance is £12,500, a lot of lowish-paid people, which includes many migrant workers, are not paying as much tax, so the net contribution has probably fallen over time. It will also have fallen as more such people have had children and they have gone to school.

On average, we would expect a migrant with a given set of earnings to make a bigger net fiscal contribution to the tax system for the simple reason that we have not paid for their education. They come here educated and can start work, and we have not paid the up-front costs, if you like. It is not as simple as saying that, because migrant workers do not use services and they make income tax contributions and so on, they make a large net fiscal contribution. We need to look across the life cycle, because there will be costs associated with having children and potentially, if they stay here, costs of retirement. However, because people who migrate here as adults have not been educated here, no cost is incurred for their education, which colleagues in England have calculated costs about £75,000 per child, so that is a net fiscal benefit to the UK or to Scotland. UCL has done some work on that.

I will take that as a nice way of getting into the issue of behavioural responses. Willie Coffey is going to lead on that.

I would like to hear the panel’s views on behavioural change and responses to changes in tax policy. The Fiscal Commission paper suggests that there are two types of response that people might have—one is to reduce the hours that they work, particularly in relation to the higher-rate threshold issue, and the other is to change their residence or location. The commission estimates that the loss in tax revenue from those responses could be about £13 million. If we counterbalance that with the impact of the tax change in Scotland, it is estimated that the receipt would be about £68 million to the Scottish revenue.

Do we need to worry about behavioural change? Is its effect marginal at best? What do you think we should do in the future to try to model it? The Fiscal Commission talks about standard modelling of it, but I wonder exactly what that is. We can look at that in January, but what are your thoughts on behavioural change? Is there any evidence to back up what has been said?

Alison Hosie and Joanne Walker want to respond. Alison can go first.

I know that, in this case, Angela O’Hagan has the right statistics. We talked about the matter on our way here this morning.

On behavioural change in relation to tax, it is important when we are considering negative behavioural change that we also consider social attitude surveys on the issue, which are generally much more positive about people’s impressions of increased tax, paying tax and the value of paying tax. It is dangerous to present only one side of the issue. In the media, we often hear only the negative aspects, and the value of paying tax is underestimated in those discussions.

It is also important to balance the discussion with reference to the cost of living in Scotland, and to introduce that to the discussion. Although we might be paying marginally more tax in Scotland than people pay elsewhere in the UK, it is important to look at the different costs of living in other parts of the UK compared with here.

Another interesting aspect of behaviour is tax avoidance and evasion in general. Mr Coffey mentioned the £13 million figure, but HMRC estimates that tax evasion, tax avoidance and tax debt cost billions of pounds, which dwarfs the small amount that Mr Coffey mentioned in relation to behavioural change. The efforts of the Scottish and UK Governments to promote human rights in general will be meaningless unless there is an adequate effort to collect the necessary funds to provide basic public services. It is important to talk also about that issue of the tax gap when we are talking about behavioural change.

There are probably more potential behavioural changes that can be undertaken by taxpayers. There are fairly simple ones that higher-rate taxpayers might undertake, such as paying additional pension contributions. If a higher-rate taxpayer in Scotland pays more pension contributions under relief at source, they will end up with higher tax relief, so it is beneficial to stay in Scotland and pay more pension contributions.

11:15  

The possibility of evasion is an on-going issue in the whole of the UK. There are issues to do with the number of hours worked and tax-motivated incorporation, although there are some restrictions on that. A person can do it only if they are self-employed. Also, the UK Government has been introducing off-payroll working measures. They are going to come in for the private sector as well from April, and they limit the ability of certain types of self-employed people to incorporate and, in effect, change their income tax from being Scottish to being UK-based.

It is difficult to know how big an issue migration is likely to be because, again, there is a limitation on who can do it. There are people who split their time between somewhere else in the UK and Scotland, and they might have high enough incomes and enough wealth to alter their circumstances and become UK taxpayers, but it is not that simple for many other people to say that they will move house to somewhere in the north of England and commute to Scotland, especially if their job is in Scotland and their children are at school in Scotland—and, of course, it costs money to move house.

Joanne Walker, the committee has done a bit on the incorporation issue, but this is the first time I have heard about the pension matter. What did you mean by that?

It might affect people who have been taken into the higher tax band and been moved from paying 21 per cent to paying 41 per cent. If they currently make a couple of thousand pounds of pension contributions a year, they might be able to raise that enough to make sure that they do not pay any higher-rate tax. That happens because, if you make pension contributions under the relief at source system, it effectively increases your basic-rate band and your intermediate-rate band by the gross amount of the pension contribution. That might mean that you effectively pay less tax. You get a bit of tax back.

I do not quite know how that works in terms of all the block grant adjustments, and I could not say whether that is a massive issue. However, it is part of the tax system.

I know that Charlotte Barbour wants to contribute. Maybe she can tell us.

I am completely with Joanne Walker. I do not know how it flows through into the block grant adjustments and that kind of thing, and I do not even know how easy it is to measure, given some of our more basic points about the number of taxpayers being difficult to measure. To go back to our earlier conversation, decentralising statistics from the centre to get national or regional—call them what you will—statistics will probably involve a steep 10-year learning curve if you are to pull out what you want.

Coming back to the point about behavioural changes, I do not have hard measurable evidence. A lot of what we have is anecdotal but accountants’ bread and butter is about making sure your tax is right less rather than right more—might that be the way to put it?

You must look at two categories of people in the debate. The first category is people who are in Scotland. The issue of whether anyone would migrate to the north of England to sidestep some tax is a moot point but, on the other hand, as Joanne Walker has discussed, they can take some other measures, and we can come back to them. The second category is people who might think of migrating to Scotland—we need doctors and what have you—and who might be receiving messages about Scotland being expensive. I know that some folk are driven to look at the cost of living and the overall package. In pure tax terms, however, if you compare a Scottish taxpayer with somebody who lives south of the border, if they pay the higher rate and upwards, they will be paying more. That must surely be a driver of perceptions and behaviour.

Those who move into the higher-rate band will have two considerations. First, they will be paying more. Secondly, if they look at their pay packet and see that, by the time they combine NIC and income tax, if they earned another £100, they would get only £47 in their hand at the end of the day—that is less than half—they may think about what they can do about that. They could work less, take unpaid leave—if that is part of their package—or not do overtime, depending on where they sit. There may also be benefits that are worth looking at. The pension contribution is certainly a sensible proposition—what is not to like in it? You are saving for the future and that is surely to be encouraged. It would reduce your tax bill now; the theory is that you might pick up tax in years to come. However, looking at in pure, stark, tax terms, there is quite a bit there that looks like it might influence behaviour. Although it is quite difficult to measure across all higher-rate taxpayers, those marginal changes around pensions and holidays could be quite subtle in the way that they flush through.

I have quite a few people who want to contribute—that is great.

I will make two small points. The first point is that we can think about behavioural responses as short term and long term. Some people can respond relatively quickly by changing their status with regard to how they are categorised, whether as an employee, as a self-employed person or as incorporated entity. Similar short-run responses could involve paying more into pensions or cutting back on hours. There are also longer-run responses, such as migration, that could have a bigger impact over time. In the short term, you might not migrate just because of the change in tax policy, because there are big fixed costs associated with moving house and country. However, if opportunities come up elsewhere and you can move to Scotland or move to the north of England, it is the kind of thing that will affect decisions in the long run. It could also affect decisions about promotions and risk-taking in the long run. If one is thinking about taking a promotion but 53 per cent, rather than 21 per cent, would go on tax, that might affect the decision. If more of the earnings from your small business get taxed, it might also affect your decision to take risks in setting up a business. Short-run concerns versus long-run concerns must therefore be borne in mind.

The differential scale of response at different parts of income distribution is also important. When colleagues and I look at that for the UK as a whole in relation to the changes that we saw in 2010, we can see that responses around the higher-rate threshold—apart from the responses on the part of people with dividends income and incorporated businesses—actually look quite small. However, additional-rate taxpayers—those earning around the £150,000 threshold and above—are much more responsive. The cautious approach that the Scottish Government has taken to the top rate, in thinking about the potential impact, is therefore welcome, as that group seems to be the most responsive to changes in tax policy. Of course, it is also the richest group, from which you can get significant revenues.

This is really interesting. Joanne Walker and David Phillips talked about mobility as an issue that we perhaps do not know enough about, so we do not know how realistic it is to consider it as a factor.

When I have looked for information on the issue, I have found some research from other countries but very little from this country. However, the information that I found from other countries seems to imply that, other than a really small number of the hyper-mobile super-rich, the bulk of people at the upper end of the income scale are less mobile when they are later on in their careers, and they have invested economically and socially in the place that they are. They have greater mobility earlier on in their careers when things such as house prices, or clusters of skills that are relevant to the industry that they are working in, are a bigger factor in where they locate themselves.

How much information do we need to gather and how much research do we need to conduct to understand the evidence about the issue in Scotland and the UK? Is anyone doing that work at the moment, either in the Scottish Government or in academia? Are there other aspects of behavioural effects that we need to better understand in order to inhibit or reduce more active or—if I can put it this way—antisocial tax avoidance?

Goodness me; there are lots of things to talk about here. I would like to focus on public attitudes and perceptions about tax. Earlier, Joanne Walker made the point that we need better public education about tax. For us to better understand collectively in Scotland how tax revenue is raised, how it is spent and what the benefits are, there is a need for improved public education and information on tax.

There are a number of questions around public attitudes and perceptions that need to be addressed in relation to taxation. Ali Hosie referred to some of the data. In preparation for this morning, I looked at the Scottish social attitudes survey and, from the last count, it is pretty much a tie when it comes to public attitudes to taxation: 47 per cent of the population think that we should pay more and spend more and 46 per cent think that taxation should stay the same, with only 7 per cent saying that taxation should be cut.

That data on public attitudes in Scotland gives us quite a fertile basis to support a tax system that is specifically appropriate for Scotland and one that talks about the kind of social contract benefits—to use Scottish Government language—that colleagues have been talking about in relation to access to public services such as publicly funded childcare.

However, I come back to the focus on behavioural effects for higher earners. If you were to ask some very low-paid workers, they might say that reducing their working hours is a problem that they would like to have, rather than working three cleaning jobs and still not earning enough to come into a tax band at all. Whose behaviour might change there? Again, we need to consider behaviour in the context of the relationship with the labour market and the relationship between earnings and the quality of work.

Patrick Harvie made the point that, with the exclusion of a mobile super-rich group, there is not much evidence about mobility. Given that the evidence seems to be anecdotal at best, I am unsure what the evidence base for this mobility issue is.

There is something quite reductive and sterile about the assumption that individuals would necessarily leave or defer decisions about seeking promotion based on the tax regime. It is almost predicated on a classical view of human beings as homo economicus. The work of Kahneman and Tversky is about 40 years old but behavioural economics has been quite popularised and we understand that people take decisions based on a broad range of issues. It is a holistic approach.

The question that I would put is, what is the flipside? What is the behavioural driver of public policies? For example, if Scotland sought tax parity with the remainder of the UK, that would blow a £500 million hole in public services. If that meant an end to the pupil equity fund and the reintroduction of prescription charges and university tuition fees, surely that would have an even greater bearing on individuals’ decisions on whether to seek to work in Scotland or to remain in Scotland?

I appreciate that this session is focused on tax but if we are going to look at behavioural changes, it is rather artificial and abstract to simply consider tax as a driver of behaviour in isolation.

There are a few points that relate to both Tom Arthur’s question and Patrick Harvie’s question. We have no evidence about mobility within the UK because tax rates have been the same across the UK until very recently. Colleagues of mine have been thinking about whether there is now enough of a tax gap between Scotland and the rest of the UK to do some analysis. Probably not, so if you want to help some economists do some research, keep going with policy divergence and we will have a nice, natural experiment.

Done!

In terms of the international evidence, I have seen some work from the US that suggests that one thing that matters in determining how mobile people are is whether they are in major agglomerations.

The evidence from the US is that if you are in New York or California, high tax rates do not affect you as much as they would if you were in Arkansas or Mississippi because you have big non-tax reasons for being located in New York and California if you are a higher earner. That is not so much the case if you are in Mississippi, Alabama or Arkansas. The agglomeration effects can be an important countervailing factor that allows attractive locations to set higher tax rates.

The question for Scotland is whether it has those attractive attributes, such as the agglomeration effect in the Edinburgh and Glasgow area, to boost a strong economy, quality of life or other aspects that mean that Scotland can use its amenity value to set higher tax rates while not losing people.

11:30  

On the point that Tom Arthur raised, it is correct to say that, if we were to cut taxes, that would mean lower spending, which could also impact on quality of life and opportunities in Scotland. I will make two points. The standard model, in effect, looks at the tax side in isolation and says that, if there is a higher tax rate, that money goes into a black hole. Therefore, it looks at what the impact of that will be on people’s work incentives, migration incentives and so on, but it is important to think about how the money is being spent and whether it can help people to stay in work and act as a magnet for people to come to Scotland to increase the labour supply.

However, not all behavioural responses require migration. People could still benefit from Scotland’s higher spending on universities, hospitals and so on, while, for example reclassifying their London home—if they have one—as their main home, paying more into a pension or incorporating their business so that it pays the UK tax rate while they still live in Scotland.

The benefits of the extra spending are spread across the whole population, whereas the additional tax burden is focused on a small part of the population, so the benefits from spending might not fully compensate the group of people who pay the higher taxes. It is really important to think about how the spending side affects the equation. We need to acknowledge that people can avoid the tax or reduce their tax liabilities without leaving Scotland, and that the tax comes from a small group whereas the benefits of extra spending are spread among a wider group. Therefore, it is still important to think about the behavioural responses to taxation.

That would perhaps necessitate powers over tax avoidance coming to this Parliament, so that we can reduce, mitigate and eliminate such behavioural responses.

Well—

Perhaps I could interrupt. On tax avoidance and things that should not be done, we need to be careful about what we are talking about. If I were to start up in business, it is perfectly legitimate for me to decide whether I want to be Charlotte Barbour, sole trader, and pay income tax and have unlimited liability and all the things that go with that. Tax is a part of that consideration. It is every bit as legitimate, and within the laws of the country, to say, “Actually, I would sooner set up a company and have my business over there.” The tax regime that would flow from that is not only different in terms of corporation and income tax; there would be a further interplay between Scottish income tax and UK corporation tax or UK income tax on dividends. It is wrong to classify taking that approach as something that should not be done. There are legitimate business decisions to be made, and politicians need to work with that. There needs to be work across the UK.

Going back to anecdotal evidence, I recently went to a training course on the budget. We talked about the fact that businessmen—people who tend to be at the upper end of higher earning—will wonder what is not to like about incorporating their business. Such a decision will be tax driven, but I go back to the point that that is completely legitimate; it is not tax evasion, which is illegal. During the course, we were told that the UK dividend rate will stay constant and that corporation tax will go down. Where would you put your business?

This is a very interesting area—

We have talked about it before.

We have been here before, but I am conscious of the time, and we need to get through all our questions. We began to move into areas of data, with the issues that Angela O’Hagan raised, so I will go to the other Angela to continue that discussion.

So, all we want for Christmas is reliable forecast data and some transparency, please. I am really interested to hear what our guests think are the main issues for the published data in relation to income tax, and how we can make the fiscal framework more transparent and consumable. Because it is Christmas, I am going to fling in Brexit. The Fiscal Commission’s forecasts assume an orderly Brexit, although I am not sure that I am going to lay money on that outcome. Paragraph 25 of the Fiscal Commission report highlights the fact that

“there is no meaningful basis for making predictions of the development of the UK-EU economic relationship over the next few years”,

We discussed earlier the Fiscal Commission’s prediction of lower growth that would be driven primarily by lower productivity, which will be due in part to Brexit.

There are certain elements on which the Scottish Fiscal Commission is really excelling itself in terms of the information that it provides. I am incredibly impressed by the information that it has provided on policy costings, for example. In comparing that information with what the UK Government provides in its policy costing documents—which started off as a couple of pages with a couple of lines saying, “We model this”—you can see that the Scottish Fiscal Commission provides a lot more information for scrutiny and understanding of tax-policy costings.

One thing that is difficult to get from the figures that have been published by the Fiscal Commission and the Scottish Government is a sense of how things will add up over time. I know that the Scottish Government has said that the amount is £500 million, but that number is a bit of a black box, as far as I can tell. Given that the Scottish Government has set a clear strategic priority to increase tax revenues relative to the UK, and to do so in a progressive way, it is important to look at the strategy and the policy as a whole. More transparency and more information about the various components and how the figure has been calculated would be useful.

For the overall fiscal framework, the information is there in the Scottish Fiscal Commission’s report, but I found it a little bit hard to get to grips with it all because it is in different tables and different sections of the report. It could become a monster table, with the block grant adjustment and the different components of the block grant adjustment for each of the axes, as well as the revenue gain from Scottish Government policies, but having it all in one place would—for me, at least, if not for most people—make it easier to see how the Scottish Government’s budget works overall.

Are you looking for somebody to paint you a picture?

Almost, yes, because for a numbers person, having everything line by line in one table helps you to see how all the various components add up to a Scottish Government budget with £137 million more income tax, and how much is because of Scottish Government policy decisions, how much is because of UK Government policy decisions and how much is because of Scottish economic performance. It is important for Scottish voters to know what the Scottish Government is responsible for, how much the policy changes are responsible for making them pay more, how much is being raised from taxes, what UK Government policies are raising or costing and how it is affecting them, how economic performance in Scotland compares with the rest of the UK, and how voters can hold the Scottish Government to account.

I am not sure exactly what that table would look like, but having all the information in one place, whether as an infographic or a table, would be very useful to a group such as this committee, which is holding the Scottish Government and the Scottish Fiscal Commission to account. If that information could be disseminated in some way, there is at least a subset of the population—an engaged part of the population—that could also use it to hold the Scottish Government to account on its policies.

I like that idea—having one table for the whole budget would make things much more simple. It is a cracking idea.

I tend to agree. Transparency is a fundamental principle of the human-rights-based approach. Fiscal policies are generally perceived by the general public to be inaccessible. There is a need for more transparency in order for there to be a good relationship between the Government and the general public, and so that there is understanding of what money is being raised and how it is being spent.

I have a beautiful lot of little tickets on my budget document—we all have them—and I cannot count the number of different ways that I find figures being presented. Some are percentages, some are shown over three years, some do not add up between the table and the paragraph, and some headings in the paragraphs are not the same as the headings in the budget lines. I am not bad with numbers, but that is confusing. Is it deliberately confusing? The impression that it gives is that it is difficult so that people will not interrogate it.

The fiscal information needs to be more straightforward so that the general public can engage with it. I class myself as a member of the general public in terms of fiscal ability. Transparency is critical for any kind of scrutiny, whether by the public or by scrutiny bodies. I would vote for having more transparent and easy-to-understand information—in one table, for example.

Do you think that we need less, rather than more?

We do not necessarily need less, but we need information that matches so that it is the same in one place as it is in another. It is very difficult to follow the budget through, to find the information that is needed and to match it up.

The challenge is that the jigsaw has lots and lots of pieces—it is a 10,000-piece jigsaw, not a 100-piece jigsaw—there is no picture on the box and the pieces are held by different people. The Scottish Government has some, the Fiscal Commission has some, the OBR has some and the Treasury has some. How do we bring all those pieces together?

There are two broad points to make. First, there is a cumulative effect. Often, the focus is on the margins—the year-to-year change or the effect of a particular policy. We need to consider how to pull that together to show the overall effect over time.

Secondly, we want to be able to drill down into some of the detail that we were talking about earlier in order to find out the economic effect, the policy effect, forecasting risks and so on. There is a real need to compare what is in the Fiscal Commission forecasts with what is in the block grant adjustment, and for visibility of how the block grant adjustment is constructed, such as we have from the Fiscal Commission. That is really difficult.

There are some basics that we do not have yet. We do not have information about the calculation of the block grant adjustments, but that needs to be included in order that it can be shown how it links back to OBR forecasts. That will be the basis for a conversation with the OBR about how its forecasts compare with the Fiscal Commission forecasts.

Audit Scotland has a job—although we all share it—to try to explain those things. It is inherently complicated. We will build up that understanding over time, and everyone who has the various pieces can explain how the whole system fits together.

Throughout my long career in tax, I have said that no one has ever understood tax, although we can keep trying.

To be more sensible and to follow on from what Mark Taylor has said, I note that some of the interesting bits that need greater clarity do not sit only in Scotland but—in particular on the tax side—fall across Scotland and the UK. It would be far better to work together. To go back to the points on national insurance, I point out that in the UK there was a lack of clarity in that income tax went down, but national insurance went up. If we put that into the mix in Scotland, where there is Scottish income tax and UK national insurance, that adds to the complications.

If we do comparisons—they seem to be inevitable—it is incumbent on all of us to say whether we are comparing a Scottish taxpayer this year with a Scottish taxpayer next year, or comparing a Scottish taxpayer with a UK taxpayer, because the pictures are completely different. There is quite a lot that could be done to add clarity.

To follow on from Alison Hosie’s point, I agree that transparency is really important for accountability purposes, and that presentation plays a really big part in that, as has been noted in recent comments.

On Charlotte Barbour’s point, we have been comparing Scottish taxpayers with either Scottish taxpayers last year or UK taxpayers this year, but there is a point at which we will need to know how much of the effect is from the UK policy and how much is from the Scottish policy. One big table might be really good, but there is probably still a need for a few other tables, too.

11:45  

I echo the points on presentational clarity and ease of access. Questions about consistency of presentation across the budget documents took up a lot of time in the budget review process. I appreciate the challenges that colleagues have mentioned, but there is still a way to go.

For our revised and, I hope, re-energised equality budget advisory group, our work plan for 2019 includes a focus on revenue and taxation, considering the equalities dimensions of the Scottish Government’s principles and tests, the implications around revenue raising and improvement in the equalities analysis of outcomes. I have learned a great deal about the technical aspects today. Part of the challenge in scrutinising the outcomes is in moving from the technical to the human, and in understanding the difference that policies are making in people’s lives.

I do not see that anyone else wants to comment on transparency. Do you want to finish anything off, Angela?

Nobody took the bait on Brexit.

I got that.

Consistency of presentation is important, but I make the point that one size does not fit all. The idea of a single table for the more figures-minded people would be very good, but for Joe Public who, to be fair, probably does not take much interest in public finance, an infographic summarising the key messages is probably all that is needed on the budget.

I would love to see the infographic that included the block grant adjustment, the fiscal framework, the difference between Scottish taxpayers and taxpayers in the rest of the UK and the difference between what we paid last year and what we paid this year in Scotland. That would be a cracking infographic.

There could be a design competition.

You could submit the first entry to that competition, David.

I thank you all for your comments. This has been a very useful discussion that will be helpful for our budget report, which we will publish towards the end of January. I wish everyone a nice festive period, when it comes.

11:47 Meeting continued in private until 11:51.