Finance Committee

Meeting date: Wednesday, April 24, 2013

Official Report 485KB pdf

---

Scotland Act 2012

The second item of business is to take evidence on the implications of the financial powers arising from the Scotland Act 2012. I welcome to the meeting Professor Gerald Holtham, chair of the Independent Commission on Funding and Finance for Wales. I ask Professor Holtham to make a short opening statement.

In my remarks, I will focus on one part of the brief paper that I submitted, on the treatment of deductions from the block grant once tax powers have been devolved, which is a matter of immediate practical concern.

The difficulty is that the obvious things that occur to you to do in that situation are not usually the right things to do. The central authority in London says, “Now you’ve got the tax, you don’t need such a big grant. We’re going to take something off it.” The obvious thing to do is to assess the expected revenue from the tax—assuming that you are using the same rate as the rest of the United Kingdom—and deduct that, then correct the assessment for the outturn in the following year.

That is all right, but what do you do in subsequent years? After all, if that deduction was, say, £50 million, and the economy grows, it would seem a little odd to deduct just £50 million for evermore if the tax base was growing with the economy. How do you manage how the tax base, and therefore the deduction, evolve over time?

Two things might occur to you. One is to keep on doing that—to look at how the tax base has evolved and deduct that every year. That seems obvious but, in my opinion, it is a bad idea because it would prevent you from realising any benefits that you may achieve through good policy. It would also protect you against any weaknesses that you may create through bad policy and it would bias the direction in which you move taxes.

Suppose I cut a tax, and it was the right thing to do and the tax base grows. I will get the same revenue even though the rate is lower. In that case, because my tax base has grown, the deduction from the block grant will grow and I will get no benefit from the success of my policy. I have wiped out the gain from the growth of the base, so why would I ever cut a tax in that situation?

On the other hand, if I increase the tax and shrink the base, the deduction shrinks, the block grant goes up and I am not penalised. To be always raising taxes and never lowering them creates a bias as well as, in effect, protecting me from the effects of a tax change. There are things to be said in favour of that approach, but to me that is a big objection.

There is another thing that it might occur to you to do. Let us say that the deduction in year 1 is worth £50 million, which is X per cent of the block grant. You might say, “We’ll just reduce the block grant by X per cent in perpetuity.” Over time, the deduction would grow at the same rate as the block grant, which will be linked to expenditure in the rest of the UK. That is not terrible—it is what the Calman commission recommended, in general.

The weakness in that approach arises if it is applied to a shared tax base—as with income tax—because it gives you no protection if the UK does things that affect your tax base, which it can of course do. The base is shared, but because the UK controls thresholds and allowances it can unilaterally change the value of your base. Even if, for example, it shrinks the base by increasing personal allowances, your deduction will still go up at the same rate as the block grant. You can renegotiate, of course, but it is not good to have to go back and have an argument every time the UK makes a tax change. In short, the proportional deduction has things to be said for it—for a start, it is simple—but it leaves you exposed if you have a shared tax base.

That is why the commission that I chaired recommended a so-called indexed approach to income tax in which you index the deduction to the growth of the UK tax base. As a result, any messing with the base is automatically reflected in your deduction, but if you manage to change your own tax base through good or bad policy, the deduction will not be affected. In other words, you retain your advantages.

Although the Treasury said that such an approach was too complicated and could not be done, it has since said, “Okay—we’ll do it.” However, I know from the case of Northern Ireland that it has tried to apply the indexed deduction approach across the board because it wants only one method. The trouble is that that is not necessarily sensible either. It is reasonable for income tax, but there are certain taxes where you know that the UK tax base is going to grow at a different rate from your tax base. If you are indexed to the UK tax base, your tax base might not keep up with that indexation. An obvious example is any property-related tax. Although London has a large influence on the UK property market with regard to prices and transactions, it is completely detached; if there is a crisis in Greece, London property prices go up. If you are indexed to that tax base, you might be put at a disadvantage. For each tax, therefore, you must ask whether your tax base is likely to keep up with the UK tax base before you ask for that particular dispensation.

The final method of adjusting the block grant deduction is simply not to do it. Instead, you say, “Let’s do a deal on a single reduction from the block grant and leave things at that. We’ll take £50 million off for ever and that’ll be fair enough.” If you can do that, it will be by far the simplest thing and it is, in fact, a perfectly reasonable approach with a number of the smaller taxes that have been part of the devolution agreements. The aggregates tax, for example, has a static base and therefore does not grow, and there is no reason why you cannot simply look at the expected future returns from the tax, discount them at a reasonable rate to get a lump sum, take off that lump sum and leave things at that. As I have said, that would be a perfectly appropriate procedure for some taxes. I have no doubt that you would have to haggle with the Treasury over the amount that you should deduct, but I suppose that that is life.

I think that I have spoken for long enough, convener. I am happy to take questions.

Thank you very much, Professor Holtham. Those comments and indeed your briefing paper have been very helpful. I will ask a couple of questions to start with and I will then open it up to members of the committee. I note that Malcolm Chisholm has already put his name down to ask his questions early on, so he must be really keen.

Our committee adviser has examined the indexed deduction method and suggested a number of risks that he has termed “Cyclical risk”, “UK policy risk”, “Scottish policy risk” and “Asymmetric growth”. You have already touched on some of those, but can you tell the committee what the parameters of uncertainty or risk might be with regard to the share of money that could be raised? Last week, the committee was struck by the change in the Office for Budget Responsibility’s forecasts of not only UK revenues but Scottish revenues as a part of that. I know that this is like asking, “How long is a piece of string?”, but are you able to give us some boundaries in that respect?

You will have to excuse me, convener, because I do not think that I am completely up to date with the debate. I know that the original suggestion was to base the deduction on estimates, but I do not think that that is acceptable. The initial reduction might have to be based on an estimate, but there is no reason why there cannot be an ex post adjustment and a subsequent smoothing procedure.

Of course, however you adjust the block grant, a reliance on your own income tax base will introduce some volatility relative to that grant. Even if the Government cuts expenditure, there will always be some cyclical smoothing of Government spending. That said, one of the nice things about forecasts is that they are never as strong as what actually happens in the economy. I can guarantee—in fact, I bet you—that the OBR will never forecast a recession of any magnitude. As a result, you will never take the initial hit from a recession, so you will have to make an adjustment in subsequent years. The practice of forecasting might actually smooth the deductions that you get. It will not eliminate them altogether because, if you have a recession, income tax receipts will fall. However, because the deduction will necessarily happen in stages with, first, an estimate and then—you hope—a correction, the process will be slightly smoother than if you were simply taking the receipts straight from the economy, as it were.

How big are the fluctuations? As far as I am aware, the elasticity of income tax is not very far from 1 with respect to incomes. If the economy shrinks by a couple of per cent, you will get a couple of per cent reduction in income tax. That is material in any given year. However, the income tax will account for about 15 per cent of your total receipts. During our worst recession, there was a 6 per cent drop in receipts in 2008, and you do not expect to see that sort of thing more than once in every 50 years. As a result, you will be looking at a fluctuation of, say, 5 per cent of 15 per cent.

Is one of the benefits of the indexed method that it should encourage the Scottish Government to prioritise economic growth?

Yes, I think so. With the indexed method, you retain any benefits from the growth of your own tax base relative to the UK tax base.

The Scottish and UK Governments will have to plan to reduce any potential uncertainties. Is it not the case that most of the risk falls at the Scottish end?

Which risk are you referring to?

Well, the risk from getting the figures in these predictions wrong, certainly in relative terms. Do you accept that?

As I have said, the tendency in forecasting is to flatten reality. People do not forecast recessions very much. Guys in the private sector might want to make a name for themselves by forecasting a recession, but official forecasters such as the OBR would never do that. The same is true of the upside. They tend to underforecast, which means that, if there is a boom, they will not catch all of it. I do not know that there is any particular bias in the risk that will make things worse for Scotland than for the UK.

Could anything more be done to minimise that risk either way? After all, what many people are looking for is stable revenues.

09:45

It is good that there is something such as the OBR, which is at least notionally one step removed from the Treasury. A difficulty historically has been that the Treasury has been the judge, jury, prosecuting counsel and star witness, so you have really had no chance. If a forecast is made not by the Treasury and not with an eye to balancing the budget, you will be in a better position. Therefore, it should certainly be insisted that those functions are carried out outside the Treasury. It is a fact that the OBR tends to use the Treasury model and Treasury officials on secondment, so there have been questions about how independent it really is, but the independence is important.

Yes.

On the taxes that have been devolved here, we have talked specifically about the Scottish rate of income tax. Last week, the OBR made it clear that it does not really have the information or data to effectively predict revenues from the landfill tax and the land and buildings transaction tax, for example. I know that you have said that its predictions tend to flatten things out, but it has, for example, predicted an 85 per cent growth in revenue from the LBTT in Scotland over the next five years.

I understand why you are pulling the facial expressions that you are pulling; some of us thought that that was a wee bit overoptimistic, too. We have a concern that the OBR perhaps does not have the data and figures for Scotland. In the past year, it has readjusted down by 33 per cent its predictions for landfill tax income from March last year to March this year. We have concerns about those predictions and what the impact might be if it gets them wildly wrong.

I understand that. The ability to make forecasts on those small taxes is probably quite limited, so you are well within your rights to ask for some investment in that area so that the OBR can do a better job.

The key thing is that there is a mechanism in place so that there is a process of adjustment when the proceeds are realised and people know where they are. Whether that adjustment should be immediate or phased over a year or two in the interests of smoothing is worth talking about, but taking the forecasts reasonably seriously and having a mechanism in place for ex post correction are both important.

I will now widen out the discussion to involve colleagues round the table.

That was a really helpful introduction by Professor Holtham. I think that there is a degree of consensus on the distinction between the smaller taxes, which we are immediately addressing, and income tax, which is obviously the big one. To summarise on the first, I think that Professor Holtham’s suggestion is helpful. When we saw the Cabinet Secretary for Finance, Employment and Sustainable Growth, he suggested that the adjustment for the land and buildings transaction tax should be retrospective, but I suppose that you are suggesting taking a prediction for the next five years and adjusting it if it is wrong. Is that basically what you are saying?

Yes, that is right. If what has been said is right and the OBR is not too hot at forecasting, it will simply be taking history and assuming that what has happened will continue. It will extrapolate, although the approach is perhaps not hugely different. A reasonable projection should be made and then adjusted.

I was not very reassured when you said that the OBR tends not to forecast recessions—in other words, it looks on the bright side of things. I presume that, the brighter things are, the bigger the deduction from the grant will be. That is the problem if there is no adjustment later on for one-off deductions.

Yes. I suppose that that is right. The process is symmetrical, of course. The OBR tends to underestimate changes on both sides, but if it underestimates a recession, it will assume that you will get more income tax than you will actually get. That will add to the volatility of receipts. The issue ties in with the requirement to have short-term borrowing powers to smooth out cyclical fluctuations.

I think that we are getting them, so I hope that those one-off deductions will not be such a problem. However, income tax is the big one, and I am really interested in what you said about it. Will you go into a bit more detail on that?

I think that it has now been agreed between the two Governments that indexation will be the method, but I am trying to work out how it would work in practice. At present, Scottish income tax is 7.4 per cent of overall UK income tax. Would that percentage be fixed? Is that how the indexation would work? We are not talking about all income tax, so would the figure be 7.4 per cent of all relevant income tax? Indexing seems a good idea in principle, but I wonder how it would work in practice.

In practice, assuming that you did not change the rate, the UK Government would make an estimate of what the revenue was worth—the 7.4 per cent, if you like—and make that deduction. Then it would assume that the deduction will grow in future years at the same rate as the UK income tax base grows, so if the UK economy flatlines and income tax does not grow in the rest of the UK, the deduction will be unchanged. As we go along, you would hope that it will be corrected for actual outturn, so there will be small fluctuations from year to year, but the deduction will grow at the same rate as the UK tax base.

Initially, we will get just the 10p rate. Would the calculation be done on the basis of the growth of just that part of UK income tax or the growth of all UK income tax?

This is one thing that we do not like in Wales, but the 10p rate is applied across the income tax range. It applies to people who are right at the threshold of income tax and to millionaires. There is no banding, so, yes, it would be the whole tax base.

Could there be a problem if income tax grows more strongly in England or Scotland not because of Government interventions, but for reasons that are not really to do with Government policy?

Yes, indeed. That is a risk that you are taking on. To be honest, I have not thought of a clever way to protect you against extraneous risks while leaving you exposed to your own risk, if you like. It is not possible to separate completely a situation that leaves you able to change things, penalises you if you do the wrong thing and rewards you if you do the right thing from a situation in which you run extraneous risks. If you are to have the ability to change things, you will take on extraneous risks. I do not think that there is any way to get round that. If for some reason that nobody can do anything about, the income tax base in Scotland grows more slowly than that in the rest of the UK, you will be disadvantaged, and vice versa.

One point that we made in our report on the Welsh case—nobody noticed it, but it is there—is about how long such a situation should be allowed to continue. Suppose it turns out that, for no reason that you can influence, your tax base grows more slowly, which means that your ability to maintain public services slowly falls behind that of the rest of the UK. How long should that be allowed to continue? My argument is that it must be allowed to continue for long enough to retain the incentive that we talked about, but not so long that it leads to a really ugly cumulative divergence. We said that maybe every 15 years, which is three parliamentary terms, the issue should be revisited and, if that has been happening, the block grant should be recalibrated.

How possible would it be in practice to separate extraneous factors from factors that are related to Government actions?

It would not be possible. After 15 years, even the Government actions would be offset, but you would be starting from that base—you would not be going back and correcting for what had happened in the meantime. You would just start from where you were and recalibrate going forward. You would have on the table 15 years of success or failure, which in political terms is a very long time.

The point that you make is perfectly sound; I do not think that you can unscramble these things. You have to leave a long enough period for the right incentives to operate. At any rate, in my opinion, if you are in a union, you need a backstop. Otherwise, you get a cumulative divergence in which one part of the country becomes increasingly less favoured or more favoured than other parts. You can let that happen to a great extent, but you cannot let it go on forever.

Looking back over the past 15 years, do you think that there are some extraneous factors that would explain the percentage of income tax being greater or smaller in England than in Scotland?

Scotland has done fairly well over the past 15 years—the economy has not done badly at all. Of course, it is presumably very dependent on the oil price. The fortunes of a lot of people who work in the oil industry, and the size of the labour force, may depend on that. There is no doubt that Wales and Northern Ireland would have been worse off, but they were in a situation in which they had no powers to do anything. Had they had the powers, they would perhaps not have done as badly, but if you were just looking at the history, you would say that they would have been worse off with their own income tax and minus a chunk of block grant.

It is therefore a bit of an act of faith to say that, if we get the power, we can do something with it. If we look at the situation at face value, there are elements that mean that, given that our economies have grown more slowly than the rest of the UK, we would have been falling behind.

I have one observation to make. It seems, from reading the paper and listening to the OBR presentation—which was not very reassuring in terms of differentials and uncertainties—that the best thing would be to devolve all the tax-raising powers. Do you have an opinion on that?

I suppose the trade-off is that you get more latitude and more possibilities, but you increase the potential volatility. It is the same with income tax. As long as the block grant system persists, if you took all of the income tax, the deduction from the block grant would be that much greater, so more of your revenue would depend on the income tax and less on the block grant. That would tend to introduce more volatility, but it would of course give you more policy autonomy.

I may be missing your point, but I do not see that it immediately changes the forecasting problem. Whether you get a bit of the income tax or all of it, there still have to be forecasts.

I meant that we would move into a situation in which we were not dependent on the block grant, or in which it would be irrelevant in Scotland.

Ah—I see what you mean. Yes, you could move to a situation in which there was assignment of tax revenues. Obviously, as long as we are in a union—it would be different if Scotland was independent—it is difficult to devolve certain taxes, because under European Union legislation you are not supposed to have different rates within the same country.

However, that does not stop you assigning tax revenues. For example, you could say that, although the Scottish Parliament does not have any power to change the VAT rate, you will estimate VAT receipts in Scotland and then assign all the revenue. If you did that in Scotland, it is true that the block grant would be much smaller.

To take a step back, the other issue is that, if you are assigning oil tax revenues, there will be situations in which Scotland is in balance or in a small surplus, or possibly in deficit, depending on the oil price. You would be able to remove a great deal—if not most—of the block grant in that situation of tax assignment. You would be in a more volatile revenue situation, but you could certainly minimise the block grant.

Would it be a more straightforward and much simpler methodology?

10:00

I am not sure that it would, under the present tax collection systems. Taxes would still be collected by a single agency and put in a central pool. The UK would first estimate how much was to come from you; then when it had all come in, it would need systems to identify what actually came from you, and then make the dispensation. The initial estimate would therefore still be required. Of course, if you acquire your own tax collection agency, that might be different. You would then have an overhead to worry about.

I want to build on some of what Malcolm Chisholm said. Your paper refers to “indexed deduction”, under which

“the initial deduction is indexed to an external variable such as the relevant UK tax base”.

A point that has been made to us before and to which you alluded is that, if we are successful, we grow the tax base and then we get the benefit of that. Of course, we get only some of the benefit of that, and as it is basic rate tax, we get 50 per cent. However, the percentage falls when the rate is higher, so we would get only 25 per cent out of the 40 per cent rate. So, if more people start paying 40 per cent, we do not get a big benefit. Does your phrase “relevant UK tax base” mean that we should look only at part of the income tax—that is, the equivalent 10 per cent in the UK—or should we look at the whole income tax base, including the 40 or 45 per cent, or whatever it happens to be?

The base is the income that people receive, which is the base on which you levy tax. Your 10p is levied on everybody, so the UK would look at the growth of everybody’s income, including that of millionaires. Perhaps they will pay 45 per cent, but they will pay you the 10 per cent as well, just like everybody else. You would therefore have to look across the whole tax base.

I am absolutely in accord with your view, though, because I am not too keen on the 10p proportional tax across the piece. The Silk commission recommended having 10p of each tax band, which would be a better procedure, because it would allow you to alter the marginal rates of different taxpayers differentially, if you want to. It would also give your income tax receipts a little more elasticity. For those reasons, I think that there are serious limitations to the 10p proportional tax measure in the Scotland Act 2012.

That said, in determining the deduction, you would look at the growth of the total income tax base in the UK, but not at the receipts. The fact is that the UK gets more receipts because it has the 45p rate, but that is not the point. You are looking at the growth of the base—that is what you are indexing to. What you are asking is: if England had a 10p tax rate, how much would it bring in per head of population? That is what we are knocking off your block grant.

So is it better to look at the tax base, which is the income of everybody who is subject to income tax, rather than at the actual tax take?

Absolutely. The UK has higher tax rates. You have to look at the base.

And you reckon that the base protects us from other changes in income tax that might be made at Westminster.

Yes, because if the UK increases the threshold on the basic rate, that shrinks the tax base and it should shrink your deduction. Any increase in allowances or any changes in thresholds will alter the base.

You mentioned “relative need” as the issue. I think that you are saying that the UK could consider relative need every 15 years. From reading your submission, I think that you feel that, at the moment, Scotland is doing quite well and Wales is doing quite poorly—I will not argue with that—and that, if we did not have that 15-year readjustment, need would not be a factor any more. The amount would be linked historically to what we get at the moment.

Yes, that is right. I am aware of all the difficulties that an explicit adjustment for relative need raises but, unless there is an explicit adjustment, it is more difficult to motivate the 15-year adjustment. You could say, “Look, our tax base has grown more slowly than yours for 15 years,” and the Treasury would just say, “Well, so what?” Unless there are objective criteria to which to appeal, it is difficult to argue about that.

It is like the Barnett formula. We had a squeeze on relative public expenditure—not absolute expenditure—in Wales, for example. We can moan about it but, so what? If there are no criteria that say whether it is good or bad, we are lost. It would be helpful for the 15-year adjustment if there were an explicit consideration of need in the block grant.

As far as you are aware, are there no restrictions on the UK doing that? I think that, in the Azores, if there was a guarantee that the local Administration could not lose and there was no risk, it was not allowed to vary VAT. Was that just a VAT rule and, therefore, not applicable to income tax? I do not know.

I see where you are coming from. The Azores judgment was to do with the devolved authority not taking responsibility for a tax change and being compensated by the centre. I think that we are okay here because, in this case, there would just be a reassessment of the relative need in different parts of the union and an adjustment of the base block grant. The treatment of the block grant going forward with respect to the different taxes would remain the same, so that could be represented as a natural distribution of the block grant, rather than being related to tax autonomy. Any changes that were made in future on the basis of taxes would stick.

I am sure that you could find a lawyer who would say that there was something to worry about—you always can, I find—but it seems to me that we are probably okay.

Thank you for your evidence, Professor Holtham. A system whereby future alterations—I will not necessarily say reductions—to the initial reduction in the block grant are predicated on changes to the UK tax base presumably provides an opportunity for Scotland if the Scottish tax base grows more quickly than the UK one, but problems if the UK tax base grows more quickly than the Scottish one. Is that, broadly speaking, correct?

Yes, that is right.

I presume that the base of taxation can be affected by a number of factors—not just economic factors such as incomes growing more quickly in one part of the UK than another, but other things, such as wider demographic trends that might result in more people paying tax in one part of the country than when we initially set up the system. Is that correct?

You would adjust for population. That is not a particular difficulty. If the tax base grows differentially purely because the English population is growing faster, we would automatically net that out. However, it is much harder in the case of the tax base per head.

The immediate concern for me was that, on current demographic trends, the UK population is projected to grow quite dramatically over the next few decades, but that growth will be largely south of the border; it is not forecast to be the same in Scotland. Are you saying that that will be factored in?

Yes. The Barnett formula does not do much, but it does take account of relative population changes. Expenditure in England is reduced to expenditure per head and Scotland gets the change in expenditure per head, so an adjustment is made for population growth.

I presume, however, that that applies only to the portion of the block grant that continues to be paid in that manner. How does it affect the deduction and the recalculation that is carried out thereafter?

I must admit that I have not looked at that. I have always just assumed that the same adjustment would be made to the income tax base, but you have me at a disadvantage—I am afraid that I am not up to speed with whether that is actually done.

I do not think that I have you at a disadvantage, because I am not aware of whether that is the case, either.

I was going to put it to you that that could be another potential problem for us.

Yes, but it is an easily resolvable problem, because relative demographics move significantly but slowly. There will not be a mega change in one year, so it is an easy adjustment to make.

It is an easy adjustment to make, but only if the will exists to make it.

Quite.

That is what we must focus on.

Professor Holtham, in your submission, you talk about there being a “revealed preference” once a few basic factors have been agreed among the Governments. Correct me if I am wrong, but I take it that that means that what has been agreed will be announced. You go on to say that, when the results are scrutinised, the Governments may decide that they

“are not what they meant”.

In those circumstances, if a secured agreement is in place, will the Governments be bound by that?

You are talking about a needs adjustment.

Yes.

That is a political matter. When all is said and done, we can all read David Hume, but there is no cast-iron way of deciding that one need is more worthy than another. At the end of the day, it is a political issue.

We tried to establish that it is possible to do a pretty reasonable rough cut at it—which is a big improvement on doing nothing—but if people worry about the last ha’penny or the fine detail, they will be discussing it for ever. All that we did was take the way in which the Scottish, Welsh and English Governments in effect provide revenue support for local authorities and derived from that what weight they implicitly attach to the different factors that everyone agrees are relevant, such as demographics, health and poverty. All that we are saying is that no one has done that. When people look at the numbers, they might say, “Oh, I don’t know if we want to do that.” That is fine; it is just a starting point for the discussion.

At the moment with the Barnett formula, no account at all is taken of such things. Even if, in 1970, the Barnett formula was dead right, we are now 40 years down the track. We have found oil in the North Sea and have wiped out the steel industry. Despite the fact that there have been huge changes, the Barnett formula just goes on. There is no reason at all why it would reflect relative need, and it does not.

One of the problems that we run into when we discuss the issue is that, any time a suggestion is made, people say, “That’s not perfect; it could be wrong in a million ways.” That is true, but what are we comparing it with? We are not comparing it with something that is perfect. We are comparing it with what we have got, which involves making no adjustment at all. A fairly rough-and-ready, 80 per cent right adjustment is easy to do and should be fairly easy to agree on. It will not be perfect. If an attempt is made to get it 100 per cent perfect, agreement will never be reached.

We are talking about agreement at the outset, rather than agreement on the outcome.

Yes. There is a political battle to be had, and it is necessary to have a modicum of good will and a willingness to reach an agreement. I am aware that in Scotland that is a political tall order. As far as the calculations that we show are concerned, it is likely that you would take a bit of a hit—not a mega hit, but you would lose a billion or two. It is hard for politicians to tell their electorate that the right thing to do is to agree to that situation, so I understand the problem.

10:15

In a way, we are talking about the short run versus the long run. If the block grant was made on an agreed basis and everybody sees that it is fair—or heading towards fairness—there would be much less aggro about devolved territories having, for example, a lower tax rate. If Scotland were to lower the income tax rate, I would not want to read The Daily Telegraph. However, if the block grant was made on the basis that everybody could say, “Look, it is like this, and that is why it is what it is,” there would be more latitude to change taxes—assuming that we are still in the union. It would also mean that, after 15 years, it would be easy to make the adjustment if you have been losing out. That is a tough one, but even in Scotland there are reasons to move to a transparent system.

In your submission, you talk about “indexed deduction”, which you say

“does not work well where there are reasons to think the UK tax base will grow at a very different rate from that of the devolved territory.”

What is your definition of “a very different rate”?

I was thinking in particular of the transactions tax on property. I am less familiar with the Scottish situation but, as someone who divides his time between London and Cardiff, that is like night and day. If the Welsh tax receipts were tied to transactions and there is a big transactions boom in the south-east of England, the Welsh tax base could in no way keep up. It would be an inappropriate indexation in that case, and I would imagine the same might be true here.

That is not so obviously the case with the other smaller taxes. However, they are so static—the bases are either stable or shrinking—that there is no need for an indexation, and it should be possible to agree on a lump sum, say what the tax is worth and make a one-off deduction.

A problem is that the Treasury wants to standardise—whatever technique is used, it seems to want to use that for everything. One conclusion that we came to is that there is no single method that is good for all taxes. With stable, static taxes with a shrinking base, such as landfill or aggregates, one should be able to cut a deal and say that it is worth X, knock an amount off and leave it at that. I would not want to apply the index method to property transactions, because the property market is so different across the country.

I have a question about relative need. Your submission states:

“many of the variables are redundant in fixing the size of revenue support because they are highly correlated with each other.”

You then list what you think are the underlying factors—dependency, poverty, sickness and sparsity. Is that list exhaustive or illustrative of what you think some of the most important factors are?

We took soundings from local authorities and asked them what the factors are that drive need. They came up with four, which were poverty, demographics, sparsity—because it is difficult when things are spread out—and population.

When we looked at the data, we found that the demographics had to be split in two—youngsters and older people. It was also useful to introduce a chronic health variable in addition to poverty. Once those were added we were able to explain, as I said, about 95 per cent of the variation across areas. In fact, the formulae are like a Christmas tree, with numbers of pregnant women and people with autistic children and a whole slew of different things. However, they tend to be correlated and most of the variation could be found with the variables that we used.

I am not claiming that that is the last word. Perhaps someone might get a university to carry out profound research, which might find that, with eight variables, 97 per cent of the variation can be explained. As I was saying, you have to decide that there is a reasonable cut-off and stop researching and analysing, because there is no natural limit to the debate and analysis unless you decide that there is.

That is helpful.

I have a question on the back of Gavin Brown’s question on the issue of landfill tax. You are suggesting that there could be a one-off reduction because the sum involved is small potatoes in terms of what has been collected overall.

The amount is not growing—or if it is, that is because rates are being put up. There is no suggestion that the tax base is growing; if anything, it is shrinking.

Indeed—that is precisely the point that I wanted to raise. The tax has been set up with a policy objective: it is not primarily about raising revenue. The policy objective is to reduce the amount of waste going to landfill. You have just suggested that the taxation base is reducing, so the policy objective is working. Is there not a danger therefore that if you set up a one-off reduction, eventually the increase in revenue could be less than that one-off reduction?

Yes. I think that this will be a war of forecasts. You come out and say, “We’ll do this, that and the other; we think the base will do this; and if we discount that flow-back, it’s worth diddly and we should only sacrifice this.” The Treasury might then come back with another forecast. The situation might escalate and there would be a haggle. I do not know how to help you. Just make a good forecast and a good case.

Presumably, the ultimate forecast—although it might be idealistic to suggest as much—is that we will not raise anything because nothing will go to landfill.

Yes—but that will be quite a few years down the track.

What happens when we reach that stage?

If you can make the case that after 15 years you will not be collecting any landfill tax at all, and that the reduction is in perpuity, the reduction should be small—absolutely.

Could the case be made for the reduction to reduce over time?

It could, although the question is whether the monitoring of that, and the hassle, would be worth it, or whether you should just cut a deal and say, “We know this will not be a great revenue raiser; therefore the reduction should be really small.” If you can reach a reasonable agreement about what the reduction should be, I would do that.

On that basis, should the reduction be small?

Yes.

That appears to have exhausted the questions from the committee. I have a couple of questions on which I am keen to have your views.

One of the points made by our budget adviser is that the system as proposed under the SRIT contains an incentive to promote higher-wage jobs in Scotland, perhaps at the expense of people who have been out of the jobs market for a long time. For example, you could have 50 workers earning £20,000 per year. If you take away the £10,000 threshold for each worker, each of them will then be liable to £10,000 of tax at 10p, so that would be £1,000 for 50 workers. Fifty thousand pounds would accrue to the Scottish Government. However, if 10 individuals are earning £100,000 per year, they will also have the £10,000 threshold, but each person will have £90,000 available to be taxed for the Scottish Government.

Under the first scenario, the Scottish Government would raise £50,000 in tax, while £90,000 in tax would be raised under the second. Leaving aside other factors that come into play, such as fairness, surely that distorts the incentive for the Scottish Government to focus on high earners rather than on people who have been on the margins of the employment market for a long time. Last year, the committee undertook an intensive study of that issue over several months. What is your view of the impact of the SRIT?

Your assessment sounds right: there is no question that the Scottish Government is getting a higher proportion of the tax paid at the basic rate than it is of the tax paid at higher rates. That situation influences the Government’s utility, if I can put it that way, with regard to who is getting incremental income. That is why I would prefer a situation in which, say, the Government got 10p of each tax band.

You want to aim for a situation in which you get a certain proportion of the tax receipts in each band, which is what my commission recommended. The Silk commission has come out with something that is close to that but is intermediate between what is in the Scotland Act 2012 and what we proposed. It is certainly an improvement.

You said that there should be a recalibration after maybe 15 years, which I found very interesting. Let us say that, after 15 years, the Scottish economy has grown 10 per cent more than the UK average. The north-east of Scotland is booming but the north of England is in a very difficult economic situation—much more difficult than the one in which we find ourselves. Surely the Scottish Government of the day would not be too chuffed if the recalibration meant a huge reduction in its block grant if it had done well.

In the same way, if the Scottish economy grew 10 per cent less than the UK average for whatever reason, would that not create political difficulties for the UK Government of the day? I imagine that Scotland would still be more prosperous than parts of Wales, Northern Ireland and the north of England. If the Scottish economy had grown 10 per cent less and the UK Government were to say that there was going to be a recalibration, there would suddenly be a huge increase in the block grant. How could that work?

My third point on the issue is that it would surely have a big influence on Government policy in the one, two or three years before recalibration if it was known that there was going to be either a huge reduction in the block grant because the Scottish economy had done particularly well, or a big increase in the block grant. The recalibration idea has tremendous dangers. Surely, if we are going to devolve it, for better or worse, we will have to live with what we do here in Scotland.

Yes. However, there is no such mechanism in place at the moment.

I am aware of that.

We need to be a bit careful, as it depends on what we are talking about. Are we talking about moving to a needs adjustment? With a needs adjustment, we take expenditure in England on the devolved areas and assume the same expenditure per head in the devolved territories, making the population adjustment. There is then a needs factor, which could be based on five or six indicators. It will not be 100 per cent of the England figure—it will be 105 or 108 per cent, or whatever the number is. That defines the block grant. We then make deductions for the various taxes. If that mechanism is in place, those indicators will reflect the slower growth of the economy or the tax base over 15 years and the adjustment will be based on where the indicators are now, not explicitly on what the tax base has done.

If the indicators were not in place, we would have to do something on the back of the fact that taxes had grown more slowly and we would run into exactly the problem that you mention of the Government thinking, “Hang on. If we are doing relatively well, who cares that our tax base has grown more slowly? That has just helped us to converge.” It is less problematic if there is an agreed needs-based mechanism in place. I have not thought deeply about how you would recalibrate without a needs-based mechanism. As you say, you would look at what the tax bases had done and make some ad hoc fix on the back of that, which would be open to question.

Your other point is a very good one, which I imagine is a matter of some delicacy here in Scotland. At the moment, English income per head is higher than that in any devolved territory; therefore, all equalisation flows are going out. Now, if we are in a union, or even a federation, and one part becomes a lot richer than the other, it must expect the flows to go the other way. Why would it not expect that? If we are in a club and the richer are helping the poorer, that is the story.

10:30

That just does not apply—and it is deeply unfortunate that there is absolutely no chance of it applying—in my country. It might apply here; it could do. If Scotland does very well and the oil price goes very high, you might be in a situation in which you would expect that flow to reverse. As you said, the 15-year adjustment would have that effect. It is a theoretical point, but there is no point in ducking it. Currently, Scotland should get higher expenditure per head than England—there is no question of that—but will that be true in 20 years’ time? You should hope not.

We do.

Further to that, how do we separate extraneous factors from other factors? It could be argued that what you are suggesting, Professor Holtham, undermines the whole rationale for devolving taxes. If, in 15 years’ time, we are just going to say, “Well, we’ll have a needs-based formula, to compensate for there being less or more poverty in Scotland,” some of which is to do with the economic policies that have been pursued in the previous 15 years, that means that, from a budget point of view, there was not much point in devolving the tax in the first place. The point was to incentivise economic growth, to put it simply, so to punish someone in 15 years’ time because they have had economic growth is to undermine the whole point of the policy, is it not?

The parallel that I make is with unemployment benefit: of course, at the margin, it is a disincentive to getting a job and doing better, but it certainly is not a complete disincentive.

We should be looking to get to an analogous situation in Scotland. If the country does a lot better, it is natural to expect that there will be some reduction in any net transfer from the rest of the union. I agree that you do not want the net transfer completely to remove the whole benefit of economic growth and prosperity, but I do not think that that would be the situation. There would be some moderate reduction in the transfer to the Government, so if it wanted to maintain expenditure it would have to use some of the increased prosperity locally, by increasing taxes, for example. Maybe the expenditure would no longer be needed, because fewer people would be in poverty, or whatever.

You are right in that, as I have said from the beginning, we cannot entirely disentangle the issue of fairness and compensating for relative disadvantage from the issue of correctly incentivising the Government. We would love to completely unscramble the issues, but we cannot do so. I think that what I have suggested is a compromise—but it is one that could be made to work.

Thank you very much. Is there anything that you want to add, which has not come up in questions?

I do not think so. The questions have been interesting and challenging, and I cannot think of anything that we have not covered.

Thank you. Your answers have certainly given us plenty of food for thought.

10:34 Meeting suspended.

10:43 On resuming—

I shall restart the session, although we are still waiting for one of our guests to appear.

Continuing our evidence taking on the implications of the financial powers arising from the Scotland Act 2012, we will now take evidence in a round-table format. I welcome to the meeting Jeremy Peat of the David Hume Institute; David Bell of the University of Stirling; David Ulph of the University of St Andrews; John McLaren of the University of Glasgow; Ken Gibb, also of the University of Glasgow; and, hopefully soon, Drew Scott of the University of Edinburgh. I thank the participants for their very helpful written submissions, which have been circulated to members.

I intend to allow up to 90 minutes for this session. Any participant who wants to respond to a question, make a point or respond to a point that someone else has made should indicate to me or to the clerk, Jim Johnston, who is next to me. We will begin the session by continuing our discussion on the block grant, on which I will shortly invite David Bell to begin. The second part of the session will focus on the economic implications of the Scotland Act 2012, on which I will invite David Ulph to lead the discussion. Finally, it would be useful to get the views of our witnesses on the impact of the new powers on the Parliament’s scrutiny function.

I welcome Drew Scott, who has now joined us for the session.

I am sorry that I am late. I had the wrong time in my head.

You are here now and we have not quite started, so that is not too bad.

David Bell, over to you.

10:45

I will just make a few points about the block grant adjustment.

First, I note that the adjustment in respect of the Scottish rate of income tax, which will be a regular adjustment, is by far the most important of the adjustments that we will consider. According to the white paper “Strengthening Scotland’s Future”, the adjustments in respect of the smaller taxes are to be allocated on a one-off basis, with some form of—as yet unclear—adjustment being made. As the committee has discussed with Professor Holtham, how one might appropriately index those smaller taxes, if at all, is pretty problematic, given their wide range.

The adjustment in respect of the Scottish rate of income tax will, it seems to me, focus risk around areas that are meant to reflect differential economic performance in the two jurisdictions. That is to say that differences in economic growth will probably be largely what drive the differences in income tax receipts north and south of the border. A quick look at the growth performance of the Scottish economy and of the UK economy over the 14-year period from 1998 to 2012 shows that the difference in the growth rate was around 2 percentage points, which is relatively small. Assuming an income tax elasticity of close to 1 in relation to gross domestic product, we are talking about 2 percentage points over 14 years. If that results in an adjustment of 2 percentage points for the SRIT adjustment, we are talking about around £80 million, which is not a huge amount of money.

Looking forward, I think that there are perhaps issues with, for example, differential population growth and differential changes in economic activity, both of which I would be happy to discuss. For example, Scotland’s economic activity rates are similar to or marginally above those in the rest of the UK and our unemployment rate is now somewhat lower. Those are likely to be key drivers of the SRIT adjustment.

We know very little about any behavioural responses to adjustments in the Scottish rate of income tax. For example, if Scotland suddenly decided that it would set a rate of 12p instead of 10p, how would economic actors in Scotland and England respond to that differential in taxes? It is quite difficult to get plausible evidence to support any particular hypothesis on what is a fairly context-specific and quite unique case.

There was mention in the earlier discussion of what will happen if Scotland’s population, for whatever reason, grows much more slowly and, 15 years down the line, the block grant adjustment is significantly larger than the amount of money that we will raise from a 10p rate. The question then is whether we just recalibrate and start again with a new adjustment. It seems to me that that is a political question.

It would be interesting to see what would happen if Wales and Northern Ireland managed to drive a change in the Barnett formula, which we are assuming will always be our comparator for what we raise in the SRIT plus the reduced block grant, and compare that with what the block grant would have been. However, the key issue then is how the principles of transparency and fairness that underlie the agreement between the UK Government and the Scottish Government would be applied if a large divergence took place. We then come back to Malcolm Chisholm’s earlier question—what would be the point if we have some kind of mechanism for the recalibration anyway? The principles underlying the discussion of the block grant specifically say that any review is to apply principles of transparency and ensure that the system remains fair and fit for purpose, whatever that may mean; we do not know what that may mean.

As far as the SRIT is concerned, history suggests that any relative adjustment will be pretty slow as long as there is not some catastrophic event in the Scottish economy, or indeed in the rest of the UK. There is a mechanism for review. Whether we would want to have it applied and what it would consist of are very open questions at the moment.

No one has indicated that they want to speak next so I will ask a question. Professor Bell, annex 1 of your paper, at 1(b), states:

“Apply the overarching objective of fairness to both the UK and Scottish Governments by ... ensuring that the mechanism is not, when implemented, designed to gain advantage in one set of fiscal circumstances or another”.

How could we tweak the mechanism to ensure that that is the case?

As I understand it, what will happen is that the indexation will be based on the comparator for the amount of money that would be raised for a 10p income tax rate in Scotland. The comparator will be trying to apply the same mechanism to the existing rest-of-UK income tax receipts. So we will look at 10p off the basic rate, 10p out of the 40p on the higher rate and 10p out of the 45p on the additional rate. Those will be modelled for RUK and that aggregate will be the basis on which we then say, “It grew by 2 per cent last year; therefore the block grant adjustment as far as the SRIT is concerned should grow by 2 per cent.” The amount that is taken away from Scotland under the block grant adjustment for the SRIT increases by 2 per cent and, if Scotland has not similarly increased its revenues by 2 per cent across the 10p bands, it will be worse off. That is how I think the nuts and bolts of the mechanism for the SRIT adjustment work.

How can we ensure that we generate more SRIT income than the reduction through the block grant adjustment mechanism? Clearly, we can do that by having more workers in the population and more economic activity. In the earlier discussion, the convener made the point that, at least in terms of revenue, there will not be much point in the Scottish Government taking somebody off benefit and putting them into a job that pays £8,000 a year, because there will be no positives in terms of additional revenue coming in through the income tax channel. What the Scottish Government would want to do is to maximise revenue, which means having more workers earning up to £100,000. Beyond that, it does not matter. The 10p rate becomes a flat tax thereafter because the personal allowance disappears.

This is nerdy stuff—I am sorry about this. The point is that, in effect, higher earners generate more revenue for the Scottish Government. However, it is less the case that the Scottish Government should be interested in high earners than that the UK Government should be, because it will take 35p out of the 45p additional rate. The top 5 per cent of earners generate 25 per cent of the income tax, or thereabouts, so those who earn high salaries are of particular interest to HM Revenue and Customs.

Quite a few people now wish to speak. I call Jamie Hepburn, to be followed by Malcolm Chisholm.

Professor Bell mentioned the issue that I raised with Professor Holtham about the adjustment to the SRIT and what happens when we have huge divergence in population growth with big growth in England but not much here in Scotland. Professor Bell suggested that that could cause some problems—I do not know whether that is the right word, but I use it for want of a better one. Professor Holtham seemed to suggest that it is his supposition that that factor will be taken into account in any adjustment, but Professor Bell seemed to indicate that that is not the case. Are you aware that it will be factored in?

I am not clear how it would be. I would have thought that the growth in revenues would be the driver rather than the population numbers. I have looked at the expected growth in population over the 15-year period that was mentioned in the earlier discussion, and between 2010 and 2025 it is expected that the number of people in the population who are aged between 15 and 60 will fall by 1.5 per cent in Scotland, whereas it will rise by 4.7 per cent in the UK. That will create about a 6 per cent differential in population growth over the next 15 years in the group that includes most of the people who are economically active. It seems to me that the way in which we make up for that has to be faster productivity growth in Scotland. That is how economic growth will push things up, and with faster productivity growth would come higher wages.

We will have to really work to increase wages, and doubly so because the adjustment is going to be affected by the differential rates of population growth.

My feeling is that that is correct.

I am interested in the comments in John McLaren’s submission on the way in which Scottish income tax has fluctuated as a percentage of UK income tax. It fell from 8 per cent to 7 per cent, and 7.4 per cent is the current figure that is quoted. I am still trying to get my head round that. Is that a different situation from what David Bell described in terms of the relevant 10p bands? Would they fluctuate in a similar way or would they be a more constant percentage?

That leads me to a second question, which I also raised during the previous evidence session. How, in terms of indexation, can we strip out extraneous factors that have nothing to do with the policies of either the Scottish Government or the UK Government?

11:00

Those large changes involving the 7 to 8 per cent figures happened over four years. It was a pretty big change. It depends whether there has been a real change in activity, in which case it would automatically feed through in the way that it should, or there has been an error in the way in which the survey figures are being collected. Because they were fluctuating so much during the period concerned, the survey size was improved. The figures still fluctuate quite a bit, however, and no analysis is really done about whether or not they match economic performance.

Let us consider the Scottish labour market performance. Between about 2003 and 2008, Scotland had been about 2 percentage points below the employment rate for the UK. By 2008, it was about 1 or 2 percentage points above it. That is a pretty good performance, although it has not been very well explained by anybody. You would expect that to have an impact on Scotland’s share of UK income tax. As far as I can see, however, it did not. Neither, apparently, did it have an impact on Scotland’s growth.

These are the sort of elements to consider. Is Scotland doing better or not? How much of the information is being picked up by the surveys that the OBR or whoever will use to predict how much money Scotland will be getting? That is predicted on liabilities, not on actual tax payments. There are many issues there. This is not so much about the level; the change is more important. If the difference stays the same the whole time, that is okay, but if we cannot explain ups and downs, things could go either way, but Scotland’s budget is not getting the amount of money that it should.

The survey of personal income is very sensitive to high earners, who can cause very significant fluctuations in regional shares of income tax. One of my research assistants was looking into that recently, and found 13 multimillionaires who, because they earn so much, are not allocated to a region, because that might be disclosive. If a relatively small number of people who are allocated to an area are very high earners, and their contribution to the total income tax receipts is very large, that can make the percentages fluctuate quite a lot. The 10p proportion, however, would be much more stable as a consequence.

It relates only to earnings. It excludes income from interest, dividends and so on, which tends to be more erratic, and which the high earners earn more of. That should not be so much of a problem, but I have not seen any analysis to ascertain how much of a steadying impact it has. I am not sure whether David Bell has done anything in that regard.

No.

I want to explore the idea that Scottish income tax revenue has fluctuated as a percentage of overall UK revenue in recent years. I have some figures before me and, as far as I can see, over the past 10 financial years, the lowest percentage has been 7.2 per cent and the highest has been 7.4 per cent. It has been 7.4 per cent in six of the last 10 years; it was 7.3 per cent in two of the last 10 years; and it was 7.2 per cent in two of the last 10 years. It has changed by no more than 0.1 per cent from one year to the next. Why are we defining that as wildly fluctuating?

There are two points there. First, it was fluctuating wildly in the past, and it still has the potential to do so. Secondly, to return to what I was saying before, if Scotland’s employment rate has moved from 2 percentage points worse than that of the UK to 2 percentage points better than that of the UK over a six-year period, why has its share of income tax as a proportion of UK income tax not improved as well? Why has it stayed flat?

The point is that reference was made to income tax fluctuating wildly, and that response does not suggest to me that it is.

Professor Bell, I think that you said that, because the block grant adjustment is to do with the Scottish rate of income tax, it should be linked to the equivalent UK tax. When I asked Professor Holtham about that, he talked about it being linked to the tax base. I struggle a wee bit to know what the practical difference would be between the two.

My understanding is that the Scottish Government wants to compare like with like. Therefore, it wants to do the comparison with the 10p part of the UK income tax base so that indexation would be based on that 10p component rather than the total component of income tax, which, of course, is much more driven by the high earners. We would consider less than a quarter of those additional-rate tax revenues, for example.

So, as I understand it, there are three options: the total UK income tax, the UK 10p bit of the tax or the UK tax base. Presumably, the UK tax base is just 10 times the 10p bit of the tax. Is that the case?

The tax base is the set of taxable units, which depends on the number of people who are employed or self-employed and above the lower income tax threshold. Most of them will go straight into paying income tax in general and the 10p rate as well if they are in Scotland.

I consider the tax base to be the units. Then there is also the revenue, and the revenue for Scotland is based on the 10p rate; the revenue for the UK is based on all of the rates—the 10p, 40p and 45p rates.

That is helpful. I do not know whether anyone else agrees or disagrees with that.

I will go back to first principles for a moment. As I understand it, one of the key objectives of, or opportunities for, the differential rate of income tax is to incentivise behaviour in the economy that tends to lead to economic welfare in Scotland as opposed to the status quo. I use the term “economic welfare” deliberately because it may not be pure economic growth but some other features that the Scottish Government, as advised by the Scottish Parliament, deems appropriate.

In that context, I struggle a little bit to know whether, in principle, the block grant adjustment should reflect changes in the income tax base in the UK that are related to differential population growth and differential rates of economic activity. If the latter are taken as not entirely independent and exogenous—if they are related to policies and to developments in the economy—and if, for example, economic activity grows in the rest of the UK more than in Scotland, should one adjust the adjustment to reflect that or allow it to flow through because that is the way that the incentives and economies have worked and it reflects behaviour performance?

To an extent, one could make the same point about population. Therefore, it is not 100 per cent clear that, if population growth is 5 or 6 percentage points higher in the rest of the UK than in Scotland, one would wish to compensate for all of that in the scale of the block grant adjustment if it was deemed that that was policy related or performance related in some way.

I raise that as a complexity over and above the niceties of the mathematical calculations that we have been going through. There is a point of principle somewhere deep down there.

I will let Professor Bell answer that and then bring in Michael McMahon.

I think that with symmetric shocks such as the recent recession you would expect income tax revenues in Scotland to fall. However, the block grant adjustment would also fall because the growth in UK equivalent revenues would also be falling and, as a result, the block grant would increase. The intention of this mechanism is to pick up such asymmetric effects, one of which, as you have rightly said, is differential population growth; there could also be differential changes in economic activity rates or productivity.

Of course, those asymmetric effects would also include policy changes. You could have an open discussion about whether there are asymmetric non-policy-related effects—one might argue that that is partly what population growth is—but as I understand it there will not be any particular compensation in that regard.

I regret not asking Professor Holtham about this earlier, but all of our discussion so far has purely been about the forecasting of taxation. However, there is also the question of contingency borrowing, and I wonder whether any of our academics wishes to comment on the short and long-term implications of adjusting for any shortfalls in forecasting or individual circumstances that might arise from the use of non-capital borrowing.

I am sure that Drew Scott will want to respond to that question.

I certainly raise the issue in my submission and in fact raised it before the legislation was enacted.

Going back to the steady state issue, I agree entirely with the view expressed, but it presupposes that the steady state situation will not be interrupted by the mechanism that we are introducing. One of the concerns that we had previously and indeed still have is that the borrowing for shortfalls as a result of forecasting errors could be unstable. We simply do not know whether the £500 million cap will be sufficient in smoothing the reconciliation that has been mentioned when it comes to adjusting forecasting errors. We know that there are going to be forecast errors—indeed, they are inevitable—so the question is how significant they will be, how speedily the reconciliation will have to be engineered and how the reconciliation will impact on elements of policy, which, if we are talking about tax changes, will be undesirable if regular and unpredicted, or longer-term drivers of growth such as public spending on many of the devolved competences.

I take the point that, over long periods of time, GDP tends to track these things fairly accurately, but in some senses the real difficulty for me is whether the steady state situation could be interrupted or indeed fundamentally changed if we have forecasting errors that require speedy and unpredictable changes to the Scottish budget to take account of, say, repayments as a result of incorrect forecasting requiring adjustment to the subsequent year’s block grant. There is an issue about the timing and magnitude of those adjustments and what happens when you run out of adjustment headroom on the borrowing side.

A broader question about borrowing is whether the tax base that we are talking about is sufficient to tolerate the borrowing powers in the Scotland Act 2012. After all, we are talking about £500 million plus £2.2 billion that is going to have to be serviced and repaid off an income tax base. I have severe reservations about the robustness of that income tax base and how quickly it can move to manage the repayment schedule that might unfold. There are quite significant difficulties with the borrowing facility, which, as far as I am aware—and I am quite happy to be corrected on this—have not really been analytically examined. I do not know of any work that states why £500 million—or indeed £2.2 billion—was the appropriate number, but I stand to be corrected on whether that evidence has been made available and whether an analytical case in that respect has been made.

Last week, the OBR made a downward adjustment of £18.1 billion in its forecast of income tax receipts for the UK, although the press here talked about a £650 million black hole in Scotland. Is it those dramatic changes in forecast that set the alarm bells ringing?

With changes in forecast of that magnitude, the entire headroom is written off; one year’s headroom would not be enough. That might be an exceptional adjustment, and improving forecasting or data might ultimately resolve the situation, but it seems to me that such instability over the longer term could affect the drivers of growth.

As I have said, we made this point many years ago, but no one has really adequately analysed whether the borrowing figure is appropriate or whether there is a safety mechanism or fail-safe device to deal with any asymmetric shocks or disturbing changes to tax or spending that would have longer-term implications.

11:15

Do you think that Scotland will take a disproportionate share of the risk, given the size of the tax base and the lack of levers relative to the UK?

It is inevitable that, once we start changing the funding mechanism for Scotland, we change the risks. There were risks with the Barnett formula, but they were well known and well established. We know—or at least we think that we know—that there were bypass mechanisms for Barnett in the event of some absolute problem case.

In this case, we are finding that the risk for Scottish funding is to change; it is a new type of risk that we know very little about and have little experience of. The further along the road one went and the more revenue came to the Scottish Parliament from income tax, the more that new risk would emerge.

We need to be clear about how that risk should be handled and the implications of the policy choices that are made to tackle it, one of which might be to underspend. It might be perfectly reasonable to say, “Okay, we’re going to build a contingency fund to ensure that we can meet these risks as they play out,” but that will have implications for current capital spending. I am not saying that the risks cannot be handled, but there will be short-term and possibly long-term implications of how one chooses to manage them—and it seems to me that they will have to be managed.

It is true that we have not calibrated the risks, and useful work could be done on that topic. Convener, you said that the OBR’s forecast for income tax receipts was adjusted by £18 billion, which is a pretty significant change. Given that the Scottish proportion of total income tax receipts in the UK is about 3.1 per cent, the error in the Scottish income tax forecast is 3 per cent of that £18 billion or about £54 million.

I think that you have misplaced the decimal point, Professor Bell. By my arithmetic, 3 per cent of £18 billion is £540 million.

Yes, you are right. In any case, the £500 million cap would be stretching things quite a lot. A bit more work on that would certainly be useful and, as I have said, the risks have not been calibrated thus far.

Jeremy Peat suggested that the block grant adjustment should not be altered, because whatever happens happens as a result of the policies that are pursued. Does he—or the other witnesses—think that that position would be stronger if the Parliament, and the Scottish Government as the executive in the devolved settlement, had greater leeway on policies that might, say, affect population growth? Surely this is not all down to changes in income tax. The Scottish Government has no authority over migration, for example, which one might presume accounts for a huge part of the demographic shift.

I will respond briefly, first by saying that I am delighted that David Bell and others are doing work on migration policy as part of their Economic and Social Research Council activity. We will hear more about that at the back end of this year.

The last time that I looked at the issue with Robert Wright, there were in his view opportunities for differential migration policies to be considered under the existing statutory arrangements. That might be considered, particularly if we are getting huge differences in population growth in the no-policy-shift scenario. You might wish to consider, first, whether to revisit migration policy and, secondly, whether that is feasible within the existing statutory arrangements—let alone anything that might be changed with or without independence—or, in other words, whether further devolution could take place. It is worth addressing that.

I was pleased that the OBR was pressed hard on the issue of huge errors in forecasting. If there were an error of £500 million that worked against Scotland and Scotland had to take out £500 million of contingency borrowing and then discovered that, actually, the error had been the OBR’s, who would pay the interest on that £500 million?

We would be lumped with the interest.

Well, that would be wholly inappropriate.

Indeed—I totally agree.

I will say a little more on immigration. Obviously, if the Scottish Government had greater immigration powers, that would allow it to compete on a more level playing field. However, one issue is that migrants tend to follow where original migrants have gone so, even if the Government had those powers, because many migrants would continue to go to England as that is where previous migrants went, the migration patterns that were set in the past could still work against Scotland. It is fairly complicated to understand the economics of a policy impact and when that policy should be put in place.

Those comments from Jeremy Peat and John McLaren were helpful, but they were more about whether we should have the levers and what effect that would have. We are talking about the Scotland Act 2012, and Mr Peat’s point was that there should be no recalculation of the block grant adjustment because the changes would emanate from policy decisions. The point that I was trying to make is that there would be a stronger case for that if such policy decisions were exercised from here. Given that they are not, does that not weaken the argument that the block grant adjustment should not be looked at again?

There is an issue about our being at the mercy of OBR forecasting, which is obviously one reason why we had Robert Chote at the committee for two hours last week.

I want to switch tack a wee bit to bring in Professor Ken Gibb. We have talked a lot about the SRIT, but stamp duty is also important, albeit on a lesser scale. Professor Gibb, how should the block grant be adjusted—if it should be—following the devolution of stamp duty and its replacement with the land and buildings transaction tax?

The fundamental issue is that stamp duty has historically been a volatile tax and will remain so. The revenue depends on the tax rate, the level of transactions and the average house price. All those variables can have a big impact on the revenues that are raised.

The convener just referred to the OBR, which makes forecasts in the area. It has a rather positive set of forecasts for stamp duty revenue. Those are fairly speculative, given the difficult and challenging nature of the housing market as a whole in the UK.

I point out for others round the table that the OBR predicts an 85 per cent increase in revenue over the next five years in Scotland.

It is really hard to make a ready claim to that. Clearly, growth is happening in the housing market in south-east England, but the key thing that has happened in the past few years is that the number of transactions has fallen sharply. That has been the main driver of the reduction in revenue. Transactions are being hampered by what is happening in the mortgage market and by a lack of general confidence on the demand side. At present, it is hard to see how an 85 per cent change will be generated. That seems to be an excessively confident prediction of what will occur in the housing market. That clearly suggests that, in relation to the block grant adjustment, we should be cautious about what we think will happen.

David Bell’s paper made the good point that, in the future, we will need to have a sense of the appropriate index of the rise in house prices in Scotland that we should use when we think about the rates that should be applied under the new tax. That is also difficult methodologically, as we do not have an absolutely robust way of measuring the house price index, let alone commercial property or land in the first place. That is quite controversial and is quite a fundamentally difficult thing to do.

In the paper that I produced, I drew on the variability of house prices in Scotland. For that, I used unadjusted data from Registers of Scotland. That is a very crude way of approaching the issue, but it shows huge differences.

It is one thing for the Scottish Parliament to say that we will have a national set of tax rates for the Scottish housing market but, of course, there is no Scottish housing market; there are lots of local housing markets. We can get ourselves into some difficulties with that. I would be cautious about the adjustment because of the inherent volatility of the housing market.

I will pick up on a point that was made in the earlier question-and-answer session in response to a question that Jamie Hepburn asked. Do our witnesses think that the grant adjustment mechanism for LBTT should be vastly different from the one for landfill tax, because of the policy objectives of the two taxes? In the longer term, one hopes that the revenue from LBTT would increase. As the convener hinted, the projections in the short term are optimistic but, in the long term, if the economy grows, the receipts from stamp duty or LBTT ought to increase. However, if the policy objective of landfill tax is achieved, eventually that tax will collect nothing; even if the size of the tax increases, the base will decrease over time—there is no doubt about that. Does anyone have thoughts about how the differences ought to be treated when the block grant adjustment mechanism is applied to each?

We do not have anyone who wants to make a point.

I see that I am wrong. Whenever I say that, someone leaps up.

That is because we are just waiting for you, convener.

The point that Mr Brown raises is interesting. If Scotland introduced differential policies that led to landfill tax falling off more rapidly in Scotland than elsewhere and one were to index on the basis of the levels of landfill tax in the rest of the United Kingdom, the extent of the reduction in the block grant would not reflect what was happening in Scotland. In that case, Scotland would be deliberately using its devolved powers to get a different outcome. Careful thought would be needed about what to do about block grant adjustment if a different policy, with different implications, were adopted.

The same applies with LBTT. I think that, last week, Robert Chote agreed that the movement from a slab structure to a progressive structure is wholly appropriate. It allows the opportunity, to which Jamie Hepburn referred, to have different policies. However, the question then is how one allows for the differential policy when considering the adjustment that takes place. If one is going for a different set of policies for reasons that involve a different view of the objectives and priorities in Scotland, one should not always expect exactly the same level of compensation from the block grant as one would have if the changes were due to wholly asymmetric shocks that are totally outwith Scotland’s control.

The issues are complex. Landfill tax should be dealt with on a different adjustment basis from LBTT, but exactly how to determine that adjustment is complex. Ken Gibb probably knows more about this than I do, but I would say that the work must start from first principles in each case in order to build up the appropriate mechanism.

The OBR made it clear that it did not have enough Scottish data and that it was just extrapolating from historical data or using UK data. That is obviously a concern for the committee, given the inaccuracies in that data.

11:30

That is—rightly—a massive concern. We will talk to Robert Chote later, and I am sure that the committee needs to address exactly how the OBR can be encouraged, with support from the Scottish Government and elsewhere, to have the data and analytical tools that allow it to make appropriate analyses to guide the adjustment process. Currently, it does not appear to be sufficiently resourced to do that and give you satisfaction that that will be done appropriately and wholly equitably.

Indeed.

I urge a bit of caution on thinking of the new LBTT as a possible means of intervening actively in the housing market. One of the arguments in favour of having a stable tax with stable tax rates that are adjusted according to inflation and how the distribution of house prices changes is that, at the individual behavioural level, that can create a counter-cyclical tax, which can be of value. As house prices rise and people move into higher tax bands, they are deterred from transacting, which might be quite helpful for the housing market. Equally, if house prices are falling, people move into lower tax bands and may be encouraged to move. Therefore, there is a natural automatic stabiliser, provided that the tax has a sufficient number of bands and is left alone to adjust over time. It is counter-cyclical at an individual level, but pro-cyclical at a revenue level, which is an interesting tension. That suggests why many economists—such as Mirrlees—are probably not well disposed towards such taxes in the first place.

I think that Robert Chote of the OBR argued last week that the number of transactions would go back to its pre-recessionary level, because people are unlikely to move around less than they did before. I am not really convinced by that argument.

A few people—John Mason, Jean Urquhart and Professor Bell—wish to speak. Once they have done so, I want to move on to the next part of our discussion, with Professor Ulph.

I want to clarify what Jeremy Peat said. I am still trying to get my head around the situation with the landfill tax revenue. If we agreed a target with the United Kingdom Government that we would get it down to nil in 15 years and we managed to do that in 10 years, that would mean that we would still have a cut from our block grant for the remaining five years, but we would have no tax, so we would have lost out by achieving our target. Is that right? If we took 20 years to achieve our target, we would still get tax, despite the fact that we had failed to meet our target.

I think that that is correct, but I need to get a towel over my head to confirm it. That could be a perverse incentive to delay the implementation, because there would then be greater compensation.

Yes—that is well put.

I want to ask all the witnesses about clarity. I think that the need for clarity comes up in everybody’s papers. Has there been clarity? Are you all confident that you know exactly how the block grant is currently arrived at? Is there clarity about the block grant that comes to Scotland?

I cannot allow every witness to respond to that, but people can perhaps touch on it in other responses.

We know how the Barnett formula works. It is efficient in the sense that it requires very little administrative input and no political input, and we know how we get there. As Professor Holtham said, it perhaps reflects a level of need that was appropriate in 1978-79, and that has been unchanged thereafter. We understand how the block grant is calculated and we will be able to continue to calculate what the Barnett formula would have determined the block grant to have been even after the Scotland Act 2012 provisions, the SRIT and so on have been introduced, so we will be able to compare how we would have done had we stayed with the existing structure.

Perhaps I am getting a little confused, because I thought that there was to be a one-off agreement about the adjustment in relation to the smaller taxes and that only the SRIT would have a continuing indexation element applied to it. For the smaller taxes, the white paper states:

“Hence there will be no need for subsequent adjustments to the block grant to compensate for changes to these taxes after their devolution.”

The Scottish Government and the UK Government are discussing what exactly a one-off payment means and whether that would be constant in nominal terms or in some way indexed overall. Clearly, no single index would be appropriate for the landfill tax, the stamp duty land tax replacement and so on. I think that Mr Chisholm has entered into discussion with Mr Swinney on that issue but, as far as I know, there is no information as yet from the Scottish Government on how that will happen.

We will move on to the second part of our session, which Professor Ulph will begin, on the economic implications of the Scotland Act 2012.

A number of the issues that I will raise are covered in the submissions provided by Drew Scott, John McLaren and Ken Gibb, as well as to some extent in the paper from David Bell.

In my opening remarks, I will talk about three sets of issues. The first, fairly trivial point is that one implication of the 2012 act is that, across a range of policies in the future, people will need to start to think about not just their policies’ implications for the conventional objectives of Government policy—growth, fairness, sustainability and so on—but their policies’ effect, whether beneficial or harmful, on the income tax base and on the amount of revenue that is available in Scotland to spend on various projects. Essentially, in contemplating any kind of policy change, people will need to ask the additional question whether the policy is likely to have an impact on the income tax base and, if so, whether that will be positive or negative.

One challenge will be whether we have the expertise and knowledge to try to understand what those effects will be. We have already had some discussion of that in considering Jeremy Peat’s question on whether migration policy should be looked at more actively as a policy that might have an impact on the income tax base. The point might be fairly trivial but, across a range of policies, we will see a systematic shift towards policies that promote the income tax base. Some policies that might otherwise have been accepted will be rejected in favour of policies that favour rather than lower the tax base.

A second set of questions, which were covered mostly in Drew Scott’s paper, concerns the fact that the 2012 act will expose Scotland to a series of tax risks to which it is not currently exposed. Those risks will arise through the forecasting of tax revenue but also through the variability in the tax base in Scotland vis-à-vis the rest of the UK.

I should say that, in my remarks, I use the term “tax base” in a slightly different sense from that which David Bell used. By the tax base, I mean the total amount of taxable revenue that is available in Scotland on which income tax can be levied. The tax base is therefore the product of the number of people who have income above the tax threshold and of the amount of income that they have above that tax threshold—it is not just a head count.

One big question that Scotland will need to face, which I will spend a little time thinking about, is whether to set the Scottish rate of income tax at 10p, which would basically restore the position to the basic rate of income tax; whether to raise it to 12p, as David Bell suggested; or whether to go below 10p by cutting the rate to, say, 8p. In other words, do we make the overall rate of income tax in Scotland higher than 20p in the pound or lower than 20p in the pound? We will need to think about a number of factors in making that decision.

One of the key questions that we might want to think about is how changing the tax rate would affect the amount of tax revenue available in Scotland for spending on various projects. As economists, we think about that question quite a lot. Normally, we think about how changing the tax rate has two different effects on tax revenue. First, for a given size of tax base, if we have a higher rate of tax on that tax base, we generate more tax revenue. The second effect that we think about is that raising the tax rate might affect the size of the tax base. On the whole, we tend to think that, the higher the tax rate, the lower will be the tax base, because work and other types of enterprise will be discouraged. The overall effect on tax revenue—whether it goes up or down—depends on the balance of those two effects.

The essential question that we need to consider is, even if we assume that tax revenue goes up when we raise the tax rate and down when we lower the tax rate, how sharply or moderately does tax revenue go up? Conversely, if we cut the rate of tax, how sharply or moderately does tax revenue fall? Other things being equal, if the impact of the tax rate on tax revenue is more moderate—that is, if an increase in the tax rate would produce only a small impact on tax revenue—we will tend to want to cut tax rates. On the other hand, other things being equal, the more sharply tax revenue goes up when we raise the tax rate, the more we will be inclined to raise the tax rate. The question is how sharp or how moderate the effect of a change in the tax rate would be on tax revenue. Economists have spent a lot of time thinking about and trying to measure that.

In the specific context of the 2012 act, two factors arise that make the answer to that question different from that which we normally face in economics. One factor is that the tax base is shared between Scotland and the rest of the UK. That implies that, if we raise the income tax rate, we will get all the benefits of the higher income tax rate but suffer only some of the consequences of the reduction in the tax base, because some of that will be picked up by England or the rest of the UK.

On the other hand, if we cut the tax rate, we will suffer all the loss of tax revenue but get only some of the gains from the higher tax base, because some of that revenue will accrue to the rest of the UK—to HM Treasury. If that factor was at work, it would suggest quite a sharp response in tax revenue to the tax rate. Other things being equal, that would imply that we would want to raise tax—that would be a factor in arguing for pushing above the 10p rate.

A second effect that is at work is that, if Scotland set a different tax rate from the 10p rate, that would mean that the overall basic rate of income tax in Scotland was different from that in the rest of the UK. The question is: what effect would that have on the tax base in Scotland?

We might think that, if Scotland set a lower basic rate of income tax than that in the rest of the UK, that could have a positive effect on the tax base in Scotland because, other things being equal, economic activity that might otherwise have taken place in the rest of the UK might transfer to Scotland. On the other hand, if Scotland raised the tax rate above 10p, activity that could have taken place in Scotland might be displaced to England. That effect would suggest that tax revenue will move rather shallowly in response to changes in the tax rate, which would tend to make us want to lower the rate of income tax in Scotland.

Those two effects work in different directions. The shared tax base would tend to make us want to raise the rate of income tax above 10p. Other things being equal, if having a different tax rate from the rest of the UK generated more economic activity in Scotland, we might want to cut the tax rate.

As David Bell said in his evidence, we know very little about that second effect—how responsive the tax base would be to the possibility of the tax rate in Scotland being different from that in the rest of the UK. However, that could be a really important issue to understand.

11:45

There are another couple of effects of having a different tax rate in Scotland than in the rest of the UK. It is important to think about the possibility that, if Scotland had a different tax rate from the rest of the UK, that could open up scope for tax avoidance. Tax avoidance tends to operate when economic activities are taxed at different rates in different locations. That gives rise to all sorts of opportunities for people to find artificial means of channelling their income through different countries to exploit the lower tax rate in one country vis-à-vis another. Whether the tax rate was higher or lower in Scotland, those opportunities for tax avoidance could arise and affect the size of the tax base.

I am not suggesting that that is an enormous factor that should make us cautious about setting a different rate in Scotland from the rest of the UK, but tax avoidance is driven not only by the behaviour of taxpayers but by the fact that tax systems set different tax rates on essentially the same activity in different places or on different individuals. Individuals and companies will exploit those opportunities whenever they see them arising.

There is also an issue with mobility. How many taxpayers are mobile? How many of us would move over the border for a 1 per cent or 2 per cent tax change?

What the convener just said covers the area in which I am interested. I am looking particularly at the paragraph headed “Increased Business Costs” on page 3 of your submission, Professor Ulph, and some of the ideas in it, such as that there would be

“pressure from unions/workers to … re-locate workers to the lower tax regime”.

Presumably, a Morrisons worker in Aberdeen will not go to live in Carlisle to save 2p in tax. Have studies been done in other countries about how mobile people can be? At the Scotland Bill Committee, we heard evidence that a few percentage points’ difference does not make much difference to where people live or would be willing to live.

I am also interested in your point that the

“cost of doing business in more than one location could induce companies to consider consolidating in one region.”

That is, perhaps they would pull out of Scotland altogether. Are there examples of that? Are there companies that do not operate in Denmark, for example, because Germany has a lower tax rate—I do not know the exact differences—or companies that operate in one state in Canada but not in another? Is that the experience elsewhere?

I do not know the details or precisely how big those magnitudes are. The problem is that all the effects that I discussed operate on the margin. It is not the case that vast numbers of companies or individuals will relocate. Only people who are at the margin and are thinking about whether they will put activity in one place or another will be affected by the differential tax rates, so it is necessary to identify the individuals or companies who operate on those kinds of margins for such effects to start to operate.

Morrisons will not simply pull out of Scotland altogether, but there might be other companies that are at the margin and are asking whether they want to sustain activity here rather than transferring it all to England. If the potential for different tax rates leads to additional costs for them because unions want to negotiate different wages in the two countries, that might be a tipping factor that causes companies to switch at the margin.

Would there be other tipping factors as well? For example, if Amazon is swithering about whether to come to Fife or go to Leeds, would it be a factor if, although there was 2p more on income tax, there was a better-educated workforce in Fife?

As I have said, companies’ location decisions are affected by many factors including the quality and productivity of the workforce as well as a whole range of legal and other issues. It might well be that, for most companies, changes in tax rates will not be the crucial factor in their decision to move, but for some, all those factors will more or less balance out and a different rate of income tax might cause them to move.

A while ago, I carried out some work on where in the world European companies had made direct investment and I discovered that small changes in the tariff regimes that they faced could cause them to pull out of, say, the far east and relocate production back in Europe. For some companies, tax or tariff regimes have no impact on location decisions, but some companies that we have spoken to have dramatically changed those decisions because the tariff regime changed between one country and another. You could identify the marginal companies that might be affected by differential tariff rates, for whom such effects can be important.

I suppose that this is a fairly obvious point, but we should also bear it in mind that we are talking about only one tax. Most Governments have a number of tax instruments that they can use to offset the various incentive consequences of other taxes. After the new regime is introduced, something will happen—either spending or tax will have to be adjusted. That is unavoidable because we are not going to have the full, steady-state Barnett situation. We do not know whether the amounts of money will be small or large, but something will happen.

The concern is that we are trying to work out the multiple effects of a change in one tax. In that respect, we are a bit like a one-club golfer. Instead of affecting distribution by changing income tax or affecting labour incentives, we could work on corporate mobility using some other tax instrument or work on the effect on aggregate demand by changing VAT. The difficulty that we have does not reside only in the points that David Ulph has highlighted; the situation is even more complex because we have only one tax and we are going to have to use it. That raises a number of adjacent problems that increase the risk of using only one instrument that will have multiple effects, and we cannot offset some of what one might see as undesirable effects by using other economic instruments. The story is even bigger and more complicated than has been suggested. It is not just about looking at the labour market effects, which in themselves are certainly not simple.

Those comments are related to the question that I was going to ask Professor Ulph about the conclusions that he has drawn from his very complex and interesting analysis. Drew Scott’s conclusion is that we need to have more taxes at our disposal. I wonder whether Professor Ulph draws a similar conclusion or whether he still thinks that it is economically beneficial just to have the income tax power.

It is beneficial to have the income tax power for some of the reasons that are alluded to in some of the submissions. The basic argument that economists always make for having tax powers is that you have to make tough spending decisions. As Drew Scott’s submission makes clear, the knowledge that you have to confront your taxpayers with the consequences of your spending decisions forces you to make those decisions efficiently and well. As a result, economists tend to believe that there are powerful arguments for having tax powers. My view is that there is a case for Scotland having wider tax powers and that the issue needs to be looked at.

However, I also want to highlight an issue that is tossed about a little bit in Ken Gibb’s paper. In England, there was the Mirrlees review, which looked at reforming the UK tax system, and it might be interesting to think about whether, if Scotland had more tax powers, it would want to implement some of the tax reforms that were recommended in that review. For example, on stamp duty, should we move from a slab system to a more efficient slice system? If Scotland were a small, closed economy, it could think about doing something quite imaginative and implementing all those tax reforms. However, the danger is that, if Scotland alone reforms its tax system, it might open up the possibility of its tax rates varying from those in the rest of the UK.

For a lot of taxpayers, it is not just the Scottish tax system that is relevant; for companies and individuals who might operate not just in Scotland or the UK but in Europe and other parts of the world, the relevant tax system is the global one. If you reform the Scottish tax system by moving rates better in line with each other, you might run the risk of moving them out of line with those elsewhere in the world and actually end up with a more distorted tax system than that in the UK. Reforming the Scottish tax system alone might look desirable, but it might have unforeseen consequences if the rest of the UK does not do the same. I have always been puzzled as to why the UK does not reform its tax system first and then devolve tax powers instead of doing it the other way round, by devolving tax powers and then thinking about making massive tax reforms.

Interestingly, we always look at things in terms of what is happening in the UK. Belgium, which borders the Netherlands, along with Luxembourg, Germany, France and indeed other countries that are bordered by umpteen different countries seem to be able to set tax rates that are appropriate to their needs without worrying about everyone turning up on their doorstep or moving across the border, even with the free movement of capital, goods and labour. Although there can be changes on the margins, I sometimes think that the impact of individual taxes can be exaggerated—if, of course, the taxes in question are not massively different from those in other countries.

Following on from your Belgium/France analogy and bearing in mind Gérard Depardieu’s recent or suggested move—actually, I believe that he was going to Russia—I note that the most mobile people are clearly the wealthiest and that the wealthiest pay the most income tax. As I state in my submission, the top 1 per cent pay 20 per cent of total income tax. If you can find the point at which rich people from the UK would move to Scotland, it might have a large impact on and raise the amount of income tax that is taken under your tax powers. Alternatively, if you raise income tax by a certain amount, you might find rich people in Scotland moving out of the country. It is a very important tipping point among that elite.

With regard to the Laffer curve, which is cited in David Ulph’s submission, you will have a different curve for each type of income. It will change over time and be different for each type of tax. A few years ago, I attended a lecture by Alex Salmond in which he said that the Laffer curve no longer applied to income tax but did apply to corporation tax. I am not sure how true that is, but it is an example of how these things can move. We know from David Ulph’s earlier comments that the impact on the richest people affects Scotland’s share a lot, and understanding the impact on that group of people will be important.

I also point out that, if you are trying to attract more people into the country, you need a track record to give them certainty about what is going to happen. You might, for example, cut income tax by 5p, but people might not be certain that the measure would be in place for more than a couple of years and they might wait before moving in. I believe that Ireland’s famous slashed corporation taxes were introduced in the 1960s but for a variety of reasons they gained traction only in the 1980s. That further complicates the impact that you might get from such changes. Like the 3p-either-way tax-varying power, the 10p-either-way tax might turn out to be a bit of a damp squib. People might not move it much because they will not be sure about its impact.

I think that you will find that since Gérard Depardieu moved to Russia the queues outside French cinemas for his films have diminished somewhat.

12:00

First, to add to what John McLaren and David Ulph said, it is extremely difficult to gather evidence on the behavioural responses to differential tax rates. The convener mentioned other countries in Europe. I think that, in Denmark, local rates of income tax vary considerably between local authorities, but that does not seem to induce much in the way of migration across borders. That is, I think, linked to the fairly sclerotic housing market in Denmark—there is an interaction there. We would potentially have leverage in both areas.

Secondly, we have had opportunities to reform taxes since the inception of devolution but, mainly, we have avoided that wherever possible. There was an opportunity to reform council tax, as I know to my cost. I was involved in the Burt review, which considered the possibility of introducing a land tax to replace the council tax, but our proposals were dismissed on the morning they were published.

We have talked about the Mirrlees review, which discussed transactions taxes on property, such as stamp duty. It stated:

“transactions taxes are particularly inefficient: by discouraging mutually beneficial transactions, stamp duty ensures that properties are not held by the people who value them most. It creates a disincentive for people to move house, thereby leading to potential inflexibilities in the labour market and encouraging people to live (and businesses to operate) in properties of a size and in a location that they may well not otherwise have chosen.”

There are opportunities for reform, even under the Scotland Act 2012 as it stands. I am sure that we would all be willing to assist the Parliament if such reform activities came on the agenda.

I can see that you are itching to do so.

We have been talking for almost 80 minutes and I am wary of the time, so in a moment I will ask everyone whether they have any final points to make before we wind up. First, though, I will throw out something to see whether anyone has a view on it.

What scrutiny role should the Finance Committee have in relation to the devolved taxes, the SRIT and borrowing? What impact will there be on the budget process and what role should the committee have in the reconciliation of forecast tax receipts with outturn figures and the administration of the devolved taxes? What can we as a committee do to scrutinise the changes?

As I am not being trampled over in a stampede, I will just leave that hanging. We have 10 or 15 minutes left, so I give everyone the opportunity to make any final comments on any issue or aspect of the discussion.

Who wants to go first, before I pick somebody? Professor Bell—a volunteer is worth 10 pressed men.

I will answer your question, convener.

The budget process already takes up most of the autumn for the Finance Committee, but the tax powers will involve much greater commitment or a new commitment. I question the extent to which the committee will have time to engage in other forms of investigation if it is focused on the budget and the tax powers. The committee ought to address fairly seriously whether it needs to bolster its analytical capabilities, particularly in relation to the taxes, over the next three or four years, before their full introduction.

There is also always the question of impact: how will the committee impact on the decisions that the Scottish Government might make in respect of taxes? We should remember that it will be possible to pre-announce the income tax rate but that other taxes might be market-sensitive and may not be pre-announced. Therefore, the committee will have to respond to the Government’s decisions after they have been made.

If we are going to take seriously the reconciliation between the forecasts and the budget, that will be another chunk of work. There will therefore be both a prospective and retrospective aspect of looking at taxation. There is also the budget, so the committee will have its work cut out.

Yes—we certainly have concerns about some of the big inquiries we have undertaken recently, such as those on demography and employability, which both the committee and the Parliament found very valuable. Inquiries of that size may be beyond the time that we have available in the future. Those are some of the issues we will be grappling with, so I appreciate the outgoing budget adviser’s comments.

On your request for suggestions about the scrutiny role, I suggest that we look at four areas. The first is to keep a beady eye on forecasting capability and how that will affect the role. That is utterly critical and an area in which the scrutiny role can be very effective. I know that the committee already has that in mind, but I commend it as a priority.

The second area is to consider how the Scottish Government should deal with the increased risks and uncertainties that will arise as a result of the changes. Those uncertainties include whether it may be necessary to use the capability to borrow up to £500 million on a contingency basis and the costs of dealing with that. Is it necessary to have some form of contingency reserve as part of the budget? How will the Scottish Government deal with the increased uncertainties and risks that will inevitably follow and which will be greatest in the early years while it is learning to understand them?

The third area is one that I assume that the committee is considering separately. We will have a relatively large sum available for capital expenditure via borrowing. The decisions on the deployment of those capital sums will be important, and I am sure that the Scotland’s Futures Forum is working on the issue. Prioritising decision making is also important: what test discount rate is required and how will the Government choose among the different uses? There is a real opportunity to enhance capital expenditure, but there are risks about repayment and how to make the right decisions. We cannot necessarily assume that the sums will come back in financial flows all the time; there may be expenditures of capital for non-pecuniary reasons. The committee should therefore consider how to deal with the capital funds within the budget process and how to prioritise.

Finally, given some of the discussion today, I think that the committee may wish to watch for further changes at the margin that would be desirable as a consequence of the changes that are being introduced, in order to ensure that the Scottish Government and Parliament can respond appropriately to the impact of such changes. That is a little like Drew Scott’s example of the one-club golfer: are there areas where a pitching wedge would be helpful? In the fullness of time, you may wish to consider the extension of powers to allow for that additional club. That is particularly pertinent in the context of the changes that are being introduced and the effects that they can have. I do not think that one should treat the implementation of the act as totally static or a once-and-for-all change. One should be thinking about what other changes might be appropriate in the follow-up in order to make best use of the situation and minimise the risks.

I have a general point, which I probably do not need to make, on the quality of the information. There is nothing more political than income tax: it is the most political tax that we know. Ideological differences can be just as deep as the differences between professional economists over forecasting models.

What concerns me about the work of people such as poor David Bell is the question of how the quality of information will be raised to a level where everyone can accept that there is an objective data set. It is a real challenge to ensure that we have a reasonable basis to say that the information from economic modelling or forecasting and the predictions of economists are credible. At the moment, my concern is that we do not have any credible numbers. I do not mean credible analysis because we do have that, but we are working in the dark. I would find myself in a very difficult position if, in future, the committee did not have a better objective data set than is apparent now.

The interesting discussion that we had on the potential behavioural impacts of the different rates of income tax in Scotland also has a direct analogy to the LBTT. At some point in the future, there will be a statutory instrument that will set the rates. That is still to be debated, but many of the same issues apply to it as well. The level at which those rates are set and how they relate to the level and distribution of house prices will affect the way that people consider some housing market decisions. I hope that those critical decisions will also be discussed in full in the committee.

Not everyone has to make a contribution but if anyone wants to make any final points, now is the opportunity to do so before we wind up the session.

I have a quick point. I agree with what the others have said on scrutiny, but it is also important to consider the bigger picture. One of the main reasons the act is being introduced is to try to improve the economic growth rate. As a paper published earlier this week by the Centre for Public Policy for Regions shows, it is not easy to understand what Scotland’s growth rate is. If we use the standard international practice over the last decade the rate is negative, but in real terms it is not. That does not give a true reflection of what has happened in Scotland, but that is what the official figure would have been.

It is therefore important to home in on the measure that we are trying to compare and then see how that has done relative to other countries, whether that is the UK or other small countries. Again, in the case of other small countries such as Iceland, Norway or Ireland, a lot of the rates have to be adjusted to get a comparable practice. If one of the main objects is to improve growth, it is important to understand whether that has worked or not, and I am not sure who will do that if the committee does not take an interest.

I want to emphasise the point that Drew Scott made. The discussion has brought out the fact that a lot of decisions will have to be made on the basis of very little information and understanding about the full consequences of those decisions. That is particularly important with issues of behavioural response, which, on the whole, we do not know enough about as economists to help you.

There are questions as to whether we in academia should think more about the research that we do; whether we can do research, equivalent to the kind that the Institute for Fiscal Studies does in England, to support the work of the Scottish Parliament; and whether there is a need for an equivalent body in Scotland to analyse data and produce the analysis that will help in making some of the decisions that will need to be made in the future.

That seems to be the end of the session, which has taken exactly 90 minutes. I thank everyone for their contributions; they have been very helpful for the committee’s deliberations.

At the start of the meeting, the committee agreed to take the next item in private. I therefore close the public part of the meeting.

12:14 Meeting continued in private until 12:23.