Revised

Members present

Lord Hollick (Chairman)

Baroness Blackstone

Lord Layard

Lord May of Oxford

Lord Monks

Lord Turnbull

Baroness Wheatcroft

________________

Examination of Witness

David Phillips, Senior Research Economist, Institute for Fiscal Studies

Q56 The Chairman: Mr Phillips, thank you very much for joining us. I had understood that a colleague of yours was coming with you.

David Phillips: She sent her apologies.

The Chairman: Okay. Thank you very much indeed for joining us. We are part-way through—probably a little over halfway through—our inquiry into devolved Scottish financing, a topic I know you have spent a great deal of time thinking about. Thank you very much for your written submission. Could we start off by getting your views on how the block grant should be calculated to reflect the devolved powers of taxation in Scotland and the framework that was laid down by the Smith commission?

David Phillips: Okay. Clearly, with additional revenues and additional spending being devolved to Scotland, you need to make an adjustment to the block grant. At least in principle, in the first year of operation that is quite straightforward. You add an amount to the block grant equal to the forecast revenues you are devolving to Scotland and you subtract an amount equal to the forecast spending that is now being devolved to Scotland. The issue that is more difficult is what you do in subsequent years. The Smith commission said—a little cryptically—that they “should be indexed appropriately”. It did not directly say how that indexation should take place, but the principles it outlines—issues such as the UK bearing the risk of economic shocks that affect the whole of the UK, the “no detriment” principles and the fact that Barnett should remain in place—have some implications for the kind of system that you would want to use to index, because some indexation systems will comply better with those principles than others, although none will comply completely with the principles.

One way of doing things that would have advantages and would be an attractive way of doing the indexation is to take the initial block grant adjustment you make—if, say, in the first year you take £10 billion off the Scottish block grant to account for devolving £10 billion of tax revenues—and to link that to some measure of the change in the tax revenues in the rest of the UK from the equivalent tax. For instance, you could index it to the percentage increase in the equivalent tax in the UK. If revenues went up by 10% in the rest of the UK, the next year you would increase the block grant adjustment by 10%, so up to £11 billion. You can do it in other ways—in per capita terms, rather than in percentage terms—but linking it in some way to what happens to revenues in the rest of the UK is appropriate, I think. For instance, it ensures that the UK bears the risk of shocks affecting the whole of the UK and it comes close, at least, to satisfying the “no detriment” principles, although it cannot do that perfectly.

The Chairman: When you say the detriment principle, you mean the second detriment principle, rather than the first one.

David Phillips: Actually, it satisfies both of them to some extent. You can think of the first “no detriment” principle as operating only in the very first year, but if you devolve a tax in such a way that it will automatically cause detriment in subsequent years, you can think of that as violating the first “no detriment” principle. I think it helps to satisfy both of them.

The Chairman: When we were in Edinburgh last week, we were briefed by the Finance Committee of the Scottish Parliament that it had wrestled long and hard with the second detriment principle but had concluded that in fact it was unworkable and it would just see it as a high-level principle. Do you agree with that approach?

David Phillips: I think that getting that principle to work fully would be completely unworkable. For instance, if there are any knock-on effects on the other Government—for instance, if the Scottish Government were to change income tax rules and that had the knock-on effect that it increased entitlements to, say, universal credit in the UK—one part of the second “no detriment” principle says that some compensation should be paid to compensate the Government for those knock-on effects. In certain circumstances that would be feasible, but in others it would not, so applying it in a rigid way to every single case where there is detriment or there are knock-on effects would be impossible. I agree that it should be used as a high-level principle and that it can be applied where there are particularly large knock-on effects and a common assessment of the effects can be agreed upon.

The reason I said that indexing to the revenues in the rest of the UK is a good method—and it is pretty consistent with the second “no detriment” principle—is this. Imagine there is a change in the taxes in the rest of the UK and the UK Government decide to put up income tax in the rest of the UK to fund higher spending on health there. That higher spending on health would feed into the Barnett formula and increase the block grant to Scotland. Scotland would therefore be gaining from higher taxes in the rest of the UK without paying any higher taxes itself. However, if you index the block grant adjustment to the revenues from income tax in the rest of the UK, because those revenues have gone up, you take off more from the block grant to account for the fact that income tax is devolved. That would more or less offset the extra money from the Barnett formula, leaving Scotland pretty much unaffected by the higher taxes and higher Barnett-able spending. It would not be perfect, because the tax bases and spending levels are different, and the Barnett formula does not align with the ways in which you can do the block grant adjustments, but it would be pretty close.

Q57 The Chairman: Which of the three methods that you set out when you met the Scottish Finance Committee do you favour as being the fairest and most serviceable method?

David Phillips: By the three methods, I think, you mean a fixed deduction—

The Chairman: Yes—a fixed percentage, an index to percentage change and an index to growth.

David Phillips: Yes. I would say that the least good of those is the fixed percentage, because—as I explain in that evidence to the Scottish Finance Committee—it does not satisfy the principle that the UK Government should bear shocks that affect the whole of the UK. That would mean that the Scottish Government would be exposed to recessions that affect the whole of the UK and would have to borrow a lot to smooth the cycle. I do not think that would necessarily be the most appropriate level of government for that kind of cyclical borrowing to take place at.

That leaves you the choice of a percentage-based indexation or a levels-based indexation—a per capita-based indexation. It depends on which of the “no detriment” principles you want to satisfy, I think. If you index it in percentage terms, it means that if Scottish revenues grow at the same percentage rate as those in the rest of the UK—if revenues go up by 5% in England, they also go up by 5% in Scotland—Scotland does not suffer. It does just as well under devolution of income tax as it would have done under the old-style system. If you were to expect the same percentage-rate growth in revenues, that system would deliver you no detriment to Scotland. There is a question about whether you want to do this as percentage growth in aggregate revenues in the rest of the UK or percentage growth in per capita revenues in the rest of the UK, but, on either of those bases, if you interpret equivalent performance of the Scottish economy as being equivalent percentage growth rates, this method would leave Scotland unaffected by devolution. Scotland would still gain if it grew faster than the rest of the UK, and lose if it grew slower. That is my personal favourite of these methods. The other method is where you index it not in percentage growth rates but in pounds-per-person growth rates. The problem with this one is that we start off from the position where revenues per person from income tax are lower in Scotland than they are in England. If revenues in England went up by, say, 10%, in order for Scotland to keep up with that in pounds-per-person terms, revenues in Scotland would need to go up by more than 10%. If they did not—if they grew at only 10%—Scotland would lose out from the devolution of income tax. Scotland’s revenues would need to grow by a faster rate than those in England under this per capita indexation in order to keep up with revenues in the rest of the UK. We did some analysis that said that, over the space of 10 or 20 years, if Scotland’s revenues did not grow faster than the rest of the UK’s in percentage terms, Scotland could lose billions of pounds per year by using the per capita approach to indexing, rather than the percentage approach. Some might say that that is a way you can reduce the high levels of funding that Scotland gets under the Barnett formula; others would say it is a slightly sneaky way of doing that. I think that I am in the latter camp.

Q58 Lord Turnbull: I was rather surprised by this conclusion. I will illustrate it with a simple arithmetic example. Revenues in total in the UK grew by 5%; 3% was the growth in the revenue and 2% was the growth in the population. In Scotland alone, revenues grew by 3% and their population was not growing, so they both have the same growth per capita. Because this is a deduction, if you are indexing on a per capita basis the deduction is smaller. It therefore seemed to me that that would be more favourable to the Scots, whereas you are saying that on a per capita basis that method is more favourable. I think it is because there is another factor that maybe I have not brought in, which is the starting point. There is a baseline—you are indexing on a sum of money. When that sum of money starts in Scotland, per capita the revenues are lower, therefore the effect that you had was dominant. The effect I was thinking of was that, if you take population growth out, you reduce the deduction and hence increase the grant.

David Phillips: Yes. I would distinguish between two things here. Let us take the percentage method first. You can look at the percentage growth in revenues in total. In your example, the percentage growth in revenues in total is 5%. Then you would index the block grant—the block grant adjustment—by 5%. The second option is to take the percentage growth and work out the percentage growth per capita, which is what—

Lord Turnbull: Indexing the block grant or the block grant adjustment?

David Phillips: The block grant adjustment.

Lord Turnbull: Right. One of these has a positive sign in front of it and the other has a negative sign. You actually said you were increasing the block grant—

David Phillips: I mean indexing the block grant adjustment. Sorry—yes.

Lord Turnbull: Right.

David Phillips: The first way is to do it in percentage terms, for the aggregate percentage change. If revenue has gone up by 5% in total, you change the block grant adjustment by 5%. The second way you mention is to say, “What is this 5% growth in total? What does that work out at in percentage growth per capita?” In your example, it was 3%. You are right that doing that method would be beneficial to Scotland, because what you are doing is taking into account the fact that there is differential population change in the two countries.

Lord Turnbull: Professor Cuthbert showed us that, according to his figures, using the per capita method, the grinding down of Scotland’s position was much slower.

David Phillips: Yes, but I mean per capita in a somewhat different sense to that. Your adviser, Professor David Bell, may be able to explain this in one of your private sessions at some point. I do not mean accounting for the fact that there is differential population growth. Holding the population fixed, 5% growth in revenues in England might mean £500 per person in England. Because revenues per person in Scotland are lower, 5% revenue growth in Scotland per person, holding population fixed, might mean only a £400 increase in revenues in Scotland. In order to get the same £500 revenue increase as in England, revenues would need to grow by something like 6.2%.

Lord Turnbull: But both these factors—my kind of per capita adjustment and yours—are at work, so which is going to outweigh which one?

David Phillips: Exactly. It depends. There are four options, effectively, here.

Lord Turnbull: Not three.

David Phillips: Ignore the first of the three in my paper. I think that is not an option, given the need—

Lord Turnbull: What is the fourth option?

David Phillips: The four options are as follows. There is indexing in percentage terms in aggregate, so this 5% figure you mentioned. Then there is indexing in percentage terms accounting for population change—so the 3% in your example. They are the first two examples of ways of doing things. The next is to say—holding population fixed, first of all—what is the per capita revenue increase in England? In my example, it is £500. Then you take that £500, multiply it by the population of Scotland and say, “We will adjust the block grant by this amount”. The fourth one is to say, “We will do that but also account for the fact that there is population change going on at the same time”. So there is percentage or pounds per person as two options. Then there is accounting for population change or not accounting for population change. They are two independent decisions that you can—

Lord Turnbull: Which do you prefer, now we have four?

David Phillips: Which of those four do I prefer? It depends on the extent to which you think convergence in funding between Scotland and England would be fair or not. If you think that the convergence should be as slow as possible, you would want to have the percentage indexation accounting for population growth, because—as Jim Cuthbert would show—that would give you the slowest rate of convergence. If you thought that actually Scotland was getting too much money under the current system and you wanted to remove that via this system, you would do the pounds per capita, not accounting for population growth.

Lord Turnbull: You have two jurisdictions, absolutely head to head. One of them obviously wants the one that is more generous and one of them wants the one that will grind away over the years at what we believe is overprovision. How do you get that resolved?

David Phillips: That is a very difficult question. I think that is a question for politicians, rather than economists. I can give some principles, though, that could guide this. The way the Barnett formula works is that it does not take account of differential population growth, because it does not update the base-level spending to account for the change in population. You might say that, given that the Barnett formula on the spending side does not account for differential population growth, on the revenue side you also should not account for differential population growth, because otherwise you are ignoring differential population growth where that would penalise Scotland and accounting for it where it benefits Scotland. You might say that the two principles—

The Chairman: I think we are in danger of spending too much time on this. Could we move on?

Q59 Lord Layard: It seems to me to be peculiar to be discussing this issue in terms of how you index what is deducted from something, rather than how you index the thing itself. The prima facie way of thinking about this would be, “This grant has been reduced in the first year. How do we index the remainder into the future?” That would lead you naturally to the first approach—that it is just a smaller amount that is being indexed in the same way as the Barnett formula was indexing the original total. Going on from that, what strikes me is that the Barnett formula is irrational and it is pretty extraordinary to be having this elaborate discussion of how to introduce rationality into a deduction from it when the thing itself is not rational. Why not go for the simple thing that the total has been reduced and then just apply the same percentage to the original Barnett formula?

David Phillips: I can see the logic in that approach. The reason that in my discussion and in my recommendations to the Scottish Finance Committee I said that I do not recommend that way of doing things is that, first, in a similar way, it would not meet the principles set out by the Smith commission. In particular, it would not meet the recommendation that the UK Government should bear the risk of shocks that hit the whole of the UK. To see that, consider an example like this. Say there is a recession that hits income tax revenues across the whole of the UK. That causes a fall in the Scottish Government’s tax revenues. Say the UK Government decided just to leave spending unchanged during the recession and to borrow money to fill that gap. Scotland’s block grant would not change but its revenues would have fallen, so a hole would open up in its budget. You could say that Scotland should be able to borrow that money on the markets itself. That could be possible, but the Smith commission said that it thought the UK Government should bear these kinds of risks that affect the whole of the UK, perhaps because it thought that, with a deeper capital market and access to cheaper borrowing costs, it might be easier for the UK Government to smooth these kinds of big shocks that affect the whole of the UK.

Secondly, there could be political concerns about how this would operate with the type of borrowing powers you grant to the Scottish Government. If you expect them to be able to manage these potentially big shocks to aggregate revenues, not just the kind of relative shock to Scottish revenues, you need to give them more borrowing powers. There may be concerns that they would use those borrowing powers in ways the UK Government did not think were consistent with their overall fiscal stance. By taking the approach of indexing the block grant adjustment, rather than changing the underlying block grant, you find a way to avoid the issue of Scotland having to bear big cyclical risks and the subsequent effects of Scotland needing to be given very large borrowing powers to do that. I think that is the reason why the commission—

Lord Layard: I do not quite understand why you are focusing on exposure to UK risk. If a part of the country is becoming more autonomous, it is getting less from the rest of the country; therefore it is inherently less exposed to variation in the rest of the country and more exposed to variation in itself. That seems to me absolutely inevitable. That is what devolution is about.

David Phillips: I agree that the change in the fiscal system might mean that there is some divergence in economic cycles across the UK compared with what there is now, but economic cycles in Scotland and the UK are strongly correlated. If you have a shock that affects the rest of the UK and causes revenues to fall there, that shock will tend also to hit Scotland. The Smith commission took the position that it thought the UK Government should bear the risk of these shocks that affect the whole of the UK. If they affect Scotland worse or less badly, Scotland bears the differential risk, but it thought the UK Government should bear the aggregate risk. I guess it is part of the idea that there should still be some degree of risk sharing across the UK. If you are happy for there not to be that risk sharing across the UK and that Scotland borrows to smooth the aggregate cycle, as well as its own idiosyncratic cycle or idiosyncratic secular trends, I agree that the system you describe would be much simpler, but it would expose Scotland to additional fiscal risks.

Q60 Baroness Blackstone: Most people in Scotland, for obvious reasons, want to see the continuation of the Barnett formula. Many people outside Scotland who think about these issues do not think it should continue. Would you like to see a replacement of the Barnett formula with a needs-based assessment? If so, what factors should it cover?

David Phillips: Again, that is a slightly tricky question to answer simply as an economist. One thing that it is worth bearing in mind is that the Barnett formula as it is implies some degree of convergence. The evidence that you will have seen from Jim Cuthbert would say that it does not guarantee complete convergence. It depends on relative population growth and the speed at which spending is rising. It could also depend on relative revenue growth and relative levels of revenue under a system of partial revenue. To some extent, if the concern is that Scotland is getting too much money under the current settlement, the Barnett formula would be expected to squeeze that down somewhat over time, although it will not do that when spending is falling and there are some flaws in the way the Barnett formula deals with business rates and certain other things, which mean that that has not been happening over the last 15 years or so. Anyone would say that there is no clear economic rationale for the Barnett formula. It simply takes historical levels of spending and adds an increment to them, based on relative populations and changes in comparable spending in England. There is no assessment of the level of spending in there.

There could be two approaches to a more rational system. Other countries tend to do one of two things. The first is to go for what they call revenue equalisation and account for the fact that there are different capacities to tax in different parts of the country. The system for allocating grants across the country takes into account that there are different levels of tax-raising capacity. That is normally more straightforward to do than a needs assessment because it is easy to agree on what factors affect tax capacity. You can model what a standard tax system would do across the entire country and use that. Needs assessment is more difficult, because of disagreements about what should enter a needs assessment. Here I would defer to the Holtham commission, which showed that actually you can get a pretty good approximation of the needs-based assessment that seems to be going on in England, with a whole plethora of different formulas in education and health. You can approximate that using just five or six factors. My own view is that keeping it simple would have the benefit of avoiding people trying to game the system by adding things and having things taken off that are considered to be beneficial or detrimental within the formula. Whether or not it is politically feasible to have a very simple system if there is pressure to include these other things, I do not know.

Baroness Blackstone: What you are saying is that, in principle, it is possible to move to a needs-based assessment, with a simple formula, without too many complex factors all being introduced. Assuming that could be done, you would not really agree with Alistair Darling, who perhaps was making a political point rather than an economist’s point, when he said that it would not be possible because nobody could ever agree on it.

David Phillips: I agree that it would be difficult, but at some point, if Barnett does lead to convergence, there will be pressure to reform the system. Moving towards a needs-based system or a revenue equalisation-based system would be difficult but would not be an insurmountable challenge.

Q61 Lord Monks: The logic of the Barnett formula was more political, probably, than economic, wasn’t it? It was compensation for the oil revenues in the North Sea coming into Whitehall and Westminster. That is what is bugging me, really, about formulas generally in this devolved world we are embarking on—or greater devolution of tax matters, in particular. I wonder whether you could help me to understand how any of the variations on the formula might work in relation to the revenue decisions of the Scottish Government, because the Scottish Government will have extra tax-raising or tax-cutting powers—probably more tax cutting than tax raising, I would guess, given the political atmosphere in Scotland. Supposing that the Scottish Government cut, say, air passenger duty, which is likely to devolved, and so on; how do these mechanisms work in relation to the Scottish Government saying, “Well, look, our needs now to sustain our spending levels are for a greater amount, through the block grant system, from London”? That is likely to cause tremendous resentment, if there is a feeling that they are giving away things and are claiming more off central UK-wide Government. How does the revenue side act in relation to this?

David Phillips: Under the current system as it is, with the Barnett formula in place, the kind of systems we are talking about if you index this block grant adjustment to what happens to revenues in England mean that Scotland has to bear the cost or the gain from any changes to its own taxes. If they were to cut air passenger duty and that led to a loss of revenue, the Scottish Government would just have lower revenues from their own taxes. There would be no change to what they get under the Barnett formula, so they would bear the cost themselves. That is in accordance with fairness and with what the Smith commission says about Scotland bearing the economic consequences of its own policy decisions. What happens under the current system is quite straightforward.

If you move towards some kind of needs-based assessment, it would work similarly to how it has worked with local authorities, I imagine, where you assess the taxable capacity of Scotland—not based on the actual tax rates it imposes but based on a kind of default tax system. As you do with the block grant adjustments, you would probably take the default tax system as being the tax system operating in the rest of the UK. Again, if their needs had not changed but they cut their air passenger duty—if they decided to cut their own taxes—the block grant would not adjust to compensate for that, because they had decided to deviate from this assumed default taxation system and they have to bear the cost of that themselves.

Where it gets more complicated under a needs-based system is if these tax policy changes have dynamic effects on the economy—for instance, if they cut air passenger duty and that leads to more people flying in and out of Scotland. If you based your assessment of taxable capacity on the observed number of flights in Scotland, you would assume that their taxable capacity was higher than it actually is, because the reason there are lots of flights in Scotland is that they have a very low tax rate. If they imposed the standard UK-wide tax, there would be fewer flights in Scotland and they would have a lower taxable capacity. There is an issue that, when you start to move towards a system of assessing revenue capacity and spending need, the dynamic effects you want Scotland to bear the cost of or to get the gains from can start to be cancelled out by these needs assessments and revenue capacity assessments. There is an issue that arises there, but it is to do with the dynamic effects, not the arithmetic effects of changing the tax rates, if that makes sense.

The Chairman: Lord Turnbull, could you bring us on to question 5, about the institutional structure?

Q62 Lord Turnbull: When we were in Scotland, it was made pretty clear at a political level that the Barnett formula was not negotiable. It had been re-avowed by the vow and they were not at all interested in a needs adjustment, because they suspected that it would probably show what everyone believes—that the thing is overdone. However, it seems to me that, on this second detriment principle, people have recognised that doing it bit by bit is not going to work. What you will have is that, periodically, someone will have to sit down and say, “What has happened since we last set this thing?” It will probably be on a five-year cycle. Is there then a need for some kind of body that is, first, the guardian of the figures—because we can even argue about the figures—and the guardian of the methodology? It does not necessarily go all the way towards the Australian thing of actually taking the decisions, but you would at least get to the point where there is an agreement on how this system works, because no one outside the Treasury and the Finance Ministry knows it. You will need something of that kind if, instead of doing these things continuously year by year, you let them all pile up for five years and then make a correction. You will need to have someone who is able to provide the basis for that kind of review.

David Phillips: I agree with you on the point that there will need to be changes to the institutional arrangements for managing the devolved Government finance. At the moment the Treasury is judge, juror and executioner on this issue. There are some forms of negotiation between the two Governments, but ultimately the Treasury executes and determines the plans. The question is, what form of arrangements is needed here? First, there is a need for much more transparency about the way in which the calculations are done—not just every five years, but year to year—on the operation of the Barnett formula, the block grant adjustments, the revenues and all of these things. Whether this is just the Treasury publishing its workings or whether it is having—preferably—an independent body publishing its workings, so that people can see how the figures are calculated, that is very important. It is something that has not been done under the Barnett formula to date. I think that the House of Lords Select Committee on the Barnett Formula recommended that these kinds of things be published when it looked at it back in 2009.

Lord Turnbull: The Treasury publishing its spreadsheets will not get us very far, because they will be denounced in Edinburgh.

David Phillips: Yes. As I said, a first step would be their publishing it, but it would be better if a body could be set up that could act as an impartial assessor of these things.

Lord Turnbull: Like the IFS.

David Phillips: I am not sure that the IFS would want to enter quite such a minefield. You mentioned the Australian model. As I understand it, the Australian Commonwealth Grants Commission does not actually make the determination. It makes a recommendation to the Australian federal Government, following a technical assessment and consultation with the provincial and the federal Governments, and then the federal Government makes an actual decision. That kind of set-up, where you have an independent body that makes a recommendation about how these things are to work but the ultimate decision rests in the hands of the elected Governments, would be the one most likely to work in this regard.

Lord Turnbull: Is it not inevitable that someone starting work in that area will start to work out a needs-based figure, because that will be one of the ways in which they determine their recommendation? Even if the Scottish do not want this needs-based thing at the start, will it not eventually have to go that way?

David Phillips: It depends on the remit that is given to the organisation. If the organisation is given a remit of saying how this could operate in the context of the Barnett formula and the system of block grant adjustments indexed by percentage growth, what it will then do is forecast what these growth rates will be, forecast the impacts of the policies—the knock-on effects—and decide which ones are big enough to require these compensations. If it is given a wider remit, maybe it will innovate and make suggestions about how the overall system can be changed, but that is a question of the remit it is given by the two Governments. The more difficult question is if it is a body that makes a recommendation as opposed to the decision. Which is the actual body that makes the final decision? If it reverts to the UK Government still making the decision, there could still be difficulties at that level, although you would hope that, if those decisions are based on the independent advice and they follow that independent advice, it would make it somewhat less difficult.

Q63 Lord Monks: Can I move on to a different topic? It is one you touched on earlier—borrowing powers and the form that they should take. Starting this year, Scotland has been able to borrow within certain limits, under the 2012 Act. I am interested to know whether or not you think that borrowing powers should be extended and expanded beyond those that were set out in the 2012 Act. Maybe you could tell us a little about how it is proposed that they use any powers. What plans are extant in Edinburgh to deploy the powers that came in in April 2015, for example?

David Phillips: First, I would say that the extent of the further borrowing powers they are given should depend on the extent of fiscal risk they are going to bear. Thinking about the amount of borrowing powers they need depends very much on the wider set of changes—how you are indexing the block grants and so on. If they go down the route of doing it under these methods where there is some indexing to UK revenues, the risk that Scotland bears is the risk that its own revenues move out of line with those in the rest of the UK. It does not bear the whole cyclical risk of recessions. Having said that, that could still be a pretty substantial amount of risk. There can be shocks that affect Scotland differently from shocks that affect the rest of the UK. For instance, the recent declines in oil prices have likely benefited the UK economy as a whole but are likely to have hurt the Scottish economy, particularly the area around Aberdeenshire, so they still need to have borrowing powers to smooth out the effects of these shocks that hit Scotland on its own.

Under the Scotland Act, the power to borrow for smoothing for current purposes—current borrowing—only allows you to borrow for forecast errors. If you forecast revenues to be £10 billion but they come in at £9.9 billion, you can borrow that £100 million to make up the gap. But if you forecast a recession next year and your revenues to fall, they are not able to borrow to fund that forecast shortfall in revenues—only the unforecast shortfalls in revenues. Now, I do not think that is an appropriate amount of borrowing powers to have when you could be facing these idiosyncratic risks. If you saw the oil price fall coming—of course, people did not see it coming, but if you could—you should be able to borrow to smooth those risks, not just the risks that you do not see coming. They definitely need further borrowing powers on the current side to allow them to smooth forecast as well as unforecast shocks.

How many more of these borrowing powers there should be is an open question at the moment. One of the things that I and colleagues at Stirling University will be looking at is just how big a fiscal risk Scotland will bear—how much you can expect revenues to vary from year to year. That might give you an idea about how much these borrowing powers should be.

Then you have the capital side of borrowing powers. At the moment they have £2.2 billion as the total amount they can borrow for capital spending, with about £300 million per year. I am not sure whether or not that should be increased. It depends on the extent to which you want to allow Scotland the freedom, in effect, to shift resources from current spending to capital spending, because that is in effect what they will be doing with these capital borrowing powers.

One recommendation that has been made, with which I think there are risks involved, is giving the Scottish Government a prudential borrowing regime. CIPFA has suggested that, like local authorities, the Scottish Government should have a prudential borrowing regime, which means that it is up to the Scottish Government to determine what it thinks is an affordable and prudential amount of capital borrowing that it can undertake. The way that works is that they assess an amount of their day-to-day budget—their current budget—that they think they can allocate to capital budget repayment.

That works very well for local authorities, and it might seem strange that you give local authorities more borrowing powers than you give the Scottish Government. However, the political context could be somewhat different with Scotland. In the case of local authorities, if Westminster believed that a certain authority was borrowing too much, it could intervene. Politically, it would have the capital to intervene in that authority’s borrowing and say, “We are going to cap your borrowing power”, or, if it went completely wrong, it could bail out.

In the case of Scotland, the politics are quite different for a prudential borrowing regime. If the UK Government said to Scotland, “You can have this regime”, and then a few years down the line said, “Actually, no. We think you are being imprudent. We are going to stop you borrowing any more”, the political ramifications of that would be pretty big. It would be the same if they were to bail out the Scottish Government, if they were to borrow what turned out to be an imprudent amount. While there could be scope for increased borrowing powers, and maybe a move to a prudential regime giving the Scottish Government the same kinds of powers as local government could work, there would be political risks in doing that. They should be thought about, given the quite different politics of local versus Westminster and Scotland versus Westminster politics.

Q64 Lord May of Oxford: I would like to remark at the beginning, first, that I find much of this difficult to understand, and, secondly, that my brother spent his career in the Australian central bank and I can assure you—at least, from the last conversation I had with him and many earlier ones—that things in Australia are not all that harmonious. They have just the same problems we are talking about here.

My question is, do you really think that the “no detriment” principle, which clearly offers all sorts of opportunities for difficulties, will be put into practice in a way that works harmoniously? It seems to me that tax competition, deliberate or not, would be an inevitable consequence of diverging tax rates and so on. I do not see how just chanting mantras about no detriment will cure these sorts of problems.

David Phillips: I agree with you that there is a strong chance of tax competition, particularly on the taxes where there are more mobile bases—income tax for very high earners and corporation tax in the case of Northern Ireland. If you put complete faith in these “no detriment” principles, you might say, “Okay, you have cut your top rates of income tax. People from our jurisdiction have moved to yours. You have caused us detriment. You have to compensate us”. I think that would quickly become unworkable. It would lead to arguments between different Governments and would potentially remove the incentives that Governments have to engage in policies that are of benefit to their own country. In this regard, what George Osborne said—I am not sure whether it was to the Scottish Parliament’s inquiry or to the Treasury Committee’s inquiry on this issue—was that he did not envisage the “no detriment” principle applying for these kinds of tax competition issues. If Scotland were to cut air passenger duty or the top rates of income tax, or Northern Ireland were to cut corporation tax, and that led to losses in the rest of the UK, George Osborne said that that is just the nature of the competition game and you would not seek compensation for it. Whether or not that can be believed is another question—and whether the devolved Governments would also feel the same way is another question. I agree that if it was implemented in these cases it would be very difficult to do.

Q65 Baroness Wheatcroft: First, I have to echo what Lord May said. I almost feel we are in Alice in Wonderland territory, trying to play croquet with flamingos instead of starting from scratch. Mr Phillips, if you put aside the fact that the Smith commission came up with its proposals and you are trying to work within them, I wonder whether you would not think it more sensible to start from scratch. Even Joel Barnett wanted to do away with the Barnett formula.

David Phillips: As an economist, I would want to start from scratch. One of the issues with the Smith commission is that the “no detriment” principle is by definition very status quo- oriented. Any change from the status quo counts, potentially, as a detriment to one side or the other. Not only do the principles they have used here reflect a system that has flaws and failings in it; they reinforce it, because the “no detriment” principle becomes something that binds you to keeping the system going.

I would say that there is a strong rationale for taking a more first-principles approach and looking at what should be the degree of risk sharing across the UK, what should be the degree of equalisation, what risks should the Scottish Government bear and what risks and incentives should the UK Government bear. I agree that that approach would be economically more fruitful. Politically, however, the UK comes from a system where there was a huge degree of centralisation. The status quo is therefore in people’s minds. I am not sure whether politically a different approach would have worked. Hoping that they could have avoided something like the “no detriment” principles may have been hoping too much. What we can hope for is that these principles do not become fetishised and that there is some flexibility around them.

Baroness Wheatcroft: I understand your point about the status quo, but the whole issue about these moves to further devolution is to change the status quo. With that in mind, I wonder if you could say how you view the rationale between the different taxes that it is proposed to devolve to Scotland, Wales and Northern Ireland.

David Phillips: I will take one of the taxes that has already been devolved under the Scotland Act—stamp duty. It is a property tax. Property taxes are probably one of the more natural things to devolve. It is a less mobile base. It also reflects the fact that there are already devolved taxes—not transaction taxes, but devolved recurrent taxes—on property: council tax and business rates. What it will mean is that Scotland and Wales have full control over both the annual property taxes and the transactions-based taxes. That gives them the opportunity to design a better system by abolishing stamp duty. Neither has done it so far, but in principle it would give them the power to do that. That is a sensible tax to devolve.

The rationale for income tax seems to be that it is a large and salient tax and therefore that it gives political accountability to the devolved Governments. That is true—it is a large and salient tax. However, there are some difficulties with devolving income tax in its entirety. First, it makes the UK almost unique—if not completely unique—in devolving income tax in its entirety to a devolved level of government. The UK Government are now left unable to raise income tax in Scotland.

Secondly, one of the issues that will arise under this new system is what happens when the UK Government change income tax rates in the rest of the UK. The logic of the system of block grant adjustments that we set up is that if they increase revenues in the rest of the UK and decide to spend that on pensions, say, or defence, the revenues in the rest of the UK have gone up, the revenues in Scotland have not gone up and you are taking more from the Scottish Government’s block grant now. Actually, that is fair, because the Scottish Government are benefiting from the higher pensions and higher defence spending but not paying any extra taxes themselves, so they should have less money for devolved services.

However, there are political issues around this. You can just imagine people seeing it as unfair—“You are taking money off us to fund a tax cut in England”—although that would not be the case. A system where you devolve income tax fully is slightly strange and creates some problems. It may have been more sensible to have a system where income tax was partially devolved, not fully devolved. Taking the Calman commission’s recommendations and increasing them, giving Scotland a bigger chunk of income tax but not completely devolving income tax to Scotland, might have had some benefits to it.

Baroness Wheatcroft: And allowing the Welsh to vary income tax by 10p?

David Phillips: That is fairly similar to the Calman commission. As I said, the idea behind that one is that it is a salient tax. What they have done with Wales is actually different from what was done under the Calman commission and the Scotland Act. Scotland has a lockstep—it can only move the tax rates up and down in line with each other. The House of Lords passed an amendment that said that Wales did not have a lockstep, so Wales would be able to vary the rates independently. That is kind of interesting. Suppose that Wales wanted to put up the top rate of income tax, from 45p back up to 50p. How would this work? Wales would get to keep all the additional revenue from the mechanical revenue effect—the extra 5p of income tax. If people cut how much they work or move to England and that reduces revenues, the Welsh Government bear the cost of that only on the 15p of income tax that they now control. The UK Government bear the costs of the behavioural effects on the amount that you would get from the other 35p of income tax and on national insurance. It gives a somewhat skewed incentive for Wales to increase the top rate of tax, because it gets all of the mechanical revenue effects but bears only a small part of the behavioural effects. Wales could have an incentive to put up, rather than cut, top rates of tax, because if it cut top rates it would gain only a small part of the behavioural effects. There are some interesting issues. Maybe there will not be tax competition with Wales but Wales putting up tax rates.

The other one to mention is Northern Ireland and corporation tax. Corporation tax is probably one of the taxes where the rationale for devolving is least strong. It is a very mobile tax base. I have never quite bought the idea that Northern Ireland is a special case because of its border with Ireland, given that business decisions are internationally mobile, not just mobile across land borders.

Baroness Wheatcroft: As you say, there are some interesting issues. Thank you.

Q66 The Chairman: I wonder whether, as a final question, we can climb out of the labyrinth we have been in for the last 50 minutes or so.

David Phillips: I am sorry about that.

The Chairman: Tell us about the benefits that flow from this. Are there any examples of devolved parts of countries having significantly better economic growth than the former host country? What is the evidence to suggest to those who argue for this that there will be an improvement in the overall economic performance of Scotland?

David Phillips: I must say that that is an area I have not really looked at in my research to any significant extent. The only comment I would like to make is that I suppose that would depend very much on just what fiscal arrangements were set up. If you have a system where the devolved area of the country is able to keep any revenue surplus it has—for instance, where you do not have fiscal sharing between the devolved area and the rest of the country—and it is doing well economically, it will get to keep more of the rewards itself. It will therefore have more to invest, be able to afford to cut taxes and maybe do even better again. On the other hand, if you have a system where you still have significant revenue equalisation across areas performing at different levels, these dynamic effects will be smaller. I guess that what I am saying is that there is no simple answer. It will depend not only on what powers they have but on how that interacts with the fiscal system.

The Chairman: So to the extent that the Scottish economy did better than the rest of the UK, these arrangements would mean that they would be able to keep some of that benefit but by no means all of it.

David Phillips: Yes. Actually, under the Barnett basis of it, they would get to keep all of the additional improvement in their economy. If the revenues go up faster than those in the UK, they keep all of that additional increment under the Barnett-based system. If you moved to a system of needs or revenue equalisation across different countries, because you were equalising that away they would not get to keep the additional benefits. Just what benefits accrue from tax devolution depends on the arrangements you have for fiscal equalisation.

The Chairman: Mr Phillips, thank you very much indeed for a very interesting journey through what is a highly complex system. Perhaps one should be thinking about a “no detriment” principle for the rest of the UK, on the basis of your last answer.

David Phillips: Thank you. I hope I illuminated rather than obfuscated.

Examination of Witness

Professor Iain McLean, Professor of Politics, University of Oxford

Q67 The Chairman: Professor McLean, thank you very much indeed for joining us for the second session this afternoon. We visited Edinburgh last week and were struck by the political realities of the situation, but one of the observations that we made was that the detailed workings of these arrangements were not very widely understood. Do you feel that there is sufficient awareness among the population of what is at stake here and that there is sufficient transparency around this process? Generally, on transparency, has sufficient information been put into the public domain to allow those like yourself who study this in great detail to be able really to hone your opinions and views as finely as you would wish?

Professor McLean: Thank you, my Lord Chairman. Taking those in reverse order, I believe that there is not sufficient information in the public domain. From that it follows that there is not sufficient transparency. From that, it follows that the public on both sides of the Anglo-Scottish border are unaware of what is at stake. I think that even many otherwise well- informed people are insufficiently aware.

As members will be aware, yours is one of several parliamentary Committees that is looking into this and closely related matters. I suspect you will be getting the same cycle of witnesses going round. I did not have time to do a written memorandum for you but, in the hope that it might be helpful, I sent in the one I wrote for the Scottish Parliament Scotland Bill Committee—the Devolution (Further Powers) Committee—to which I gave evidence last week. As you will be aware, this House’s Constitution Committee is also holding an inquiry. I expect that either personally or corporately, on behalf of the British Academy, I will be giving it evidence, so my apologies to this Committee if there is any repetition in what I say to, possibly, three parliamentary Committees. However, yours was an easy question, Lord Chairman, because my answer is no to all three parts of it.

The Chairman: Perhaps Lord Layard can ask a slightly more difficult question.

Q68 Lord Layard: We had some difficulty in our previous session in grappling with the question of how one thinks about the total grant to Scotland by focusing entirely on how to replace the bit that is removed, rather than thinking about the bit that remains. If I may reveal some ignorance, I would like to understand a little more how the Barnett formula itself is calculated and why you say that “contrary to widespread misconception, Barnett does not protect the relatively higher public spending per head in Scotland”, as some people assume, and that “If allowed to run on, it erodes it”. Could you give us a little lesson on that?

Professor McLean: I can do my best. Members will probably be aware that the world’s greatest expert on the operation of Barnett is indeed one of your special advisers, who can no doubt mark my homework and correct any deficiencies in what I am about to say. Barnett operates not on the baseline of public expenditure but on changes in public expenditure year on year. In principle, it gives the three devolved territories a population share of changes in public expenditure on the relevant functions of government in England—or it might be England and Wales, but we will leave that complication aside—in the reference territory. Because Barnett now operates on nominal changes in public expenditure, for all three territories, starting on the baseline of higher per head public spending on functions that are devolved to them, in the long run you should get convergence to equal public spending per head in the four territories of the United Kingdom. That is how the differential equation system, as your adviser has elegantly labelled it, works in the long run.

This has not been happening over the very many years that Barnett has operated. There are a number of reasons for this, but the main reason is relative population change. The relative populations of Scotland and Wales have been declining vis-à-vis England and therefore they are to that degree protected from the convergence effects of Barnett. Nevertheless, if the long run is sufficiently long—and we all know what an eminent economist said about that—it is correct that Barnett does not protect the relatively higher spending in Scotland. I hope that helps.

Lord Layard: Right. Given that we are starting from Barnett, we were discussing in the previous session different ways of thinking about how to calculate the value of the remaining part of Barnett. The two most obvious ways are first to shrink it by the percentage by which it shrinks in the first year and then carry it on thereafter, and secondly—I think this is the Holtham formula—to try to recreate, as it were, what Barnett would have produced had there not been the extra autonomy. What is your position as between those two?

Professor McLean: Neither of them is fully specified, so I find it difficult to state a position. My problem begins further back. How is it that, once a new tax is devolved, whoever calculates the Barnett baseline does the subtraction sum? Is the tax that was formerly reserved and is now devolved to be taken as a proportion of all tax revenue in the UK in the year before devolution? That cannot be right, because some of the tax revenue in the UK continues to be used for reserved public spending. Here we might need some algebra and a whiteboard, but I guess it is the ratio of that to all of the tax revenue in year 0 that was spent on devolved services in England—services in England that are devolved in, say, Scotland. That gives you your subtraction factor, I think.

Of the two ways that you suggest of operating it in years 2, 3 and 4, my first inclination is to agree with you that the second way would be more wholesome, but I would rather see the algebra worked out before I could be confident in that opinion.

Q69 Baroness Blackstone: Do you think the Barnett formula could be replaced by a needs-based assessment? Were that to happen, how should these needs be assessed? What kind of principles would you introduce to carry through such an assessment?

Professor McLean: Yes, I do. I have written about this in references that I can supply to your clerks. I have argued for quite a number of years now that the model jurisdiction to look at is Australia, where they have what I regard as a robust needs assessment across the eight states and territories—six states, Capital Territory and Northern Territory. The most important point about the Commonwealth Grants Commission, which operates this, is that it is not owned by the commonwealth Government. It is, in principle, owned by the Parliaments of the commonwealth and the eight territories jointly. Because an element of the distribution is zero sum, it follows that it will never be the case that all nine Parliaments agree on a particular direction to give the Commonwealth Grants Commission and therefore the commission has reasonable liberty to operate.

Its assessment of needs has become very sophisticated over the 80-plus years that it has operated because it has always been well aware that all players in this game will play games and that they will, in particular, want to say that what they do not have much of is a need and what they have a lot of is a policy choice. The Australian commission—and, I would recommend, any UK commission—tries as far as possible to derive its needs from objective factors about the age structure and morbidity structure of the relevant populations, being factors that Governments cannot alter in the short term, although they can alter them in the long term. The second thing the grants commission does is apply some time lags. It smooths its grants to avoid a situation whereby, if a Government made a particular tax effort or had a particularly successful policy in year 1, it did not lose all the results of that good practice by having it equalised away from it in year 2.

To summarise, the essential principles of a good needs assessment are, first, to have the needs calculated as near as you can on an objective basis, immune from manipulation from any of the Governments in question, and, secondly, to have a smoothing system such that any gains made by a jurisdiction are not immediately equalised away.

Baroness Blackstone: If we were to introduce a system of this sort here, how do you see it being implemented? What kind of institutional structure would we have to put in place to do it properly and then make it work?

Professor McLean: A hypothetical UK grants commission clearly would have to report to the four Parliaments, let us assume, in the system—the UK Parliament and that of each of the three devolved authorities. I imagine that each Parliament—and I think it should be owned by the Parliaments, not the Governments, of the countries—might be invited to name commissioners, but the commissioners should not be delegates of their Parliament. There are constitutional forms from the various bodies that have been invented in the hope of keeping politicians honest; I am thinking of the Statistics Authority, for instance, and the Office for Budget Responsibility. This is not my area of expertise, but one could look at the constitutional structures of bodies such as these and apply the most appropriate one that one thought one had found to the UK grants commission, if we call it that, but with the addition that it should be owned—I suppose a lawyer would say jointly and severally—by the four Parliaments.

Baroness Blackstone: So you would see each of the four Parliaments setting up some kind of Committee, whether it be a special Committee or something slightly different from a special Committee, in order to oversee the work of the grants commission, as you describe it.

Professor McLean: That would be up to each Parliament. I doubt if it would gain very much from doing that. You want to avoid a situation in which each Parliament has a Committee that simply complains that the allocation of the grants commission does not favour that Parliament. You want to try to make the system as consensual as you can. The Australians have succeeded, in my view. Maybe they are the only nation in the world that can, but you would want to try to devise a technocratic and consensual system.

Baroness Blackstone: Would it not be better to try to get nominations from each of the Parliaments to take part in the oversight of the commission, so they would meet together rather than separately? Is that the sort of thing you have in mind?

Professor McLean: My rough model is that the commission is a permanent body—staffed, of course, by public servants—that reports to a Joint Committee of the four Parliaments.

Baroness Blackstone: That is what I am getting at. It is a Joint Committee of the four Parliaments.

Professor McLean: But that Joint Committee would have to have Standing Orders such that it did not vote by majority rule.

Q70 The Chairman: When we were in Edinburgh last week, it became very clear that there was a deep attachment to the Barnett formula. We did not inquire of the level of understanding of the workings of the Barnett formula; of course, it would appear that that is quite obscure anyway. Have there been worked examples? For instance, as far as you know, did the Smith commission see worked examples of the existing Barnett formula and a needs-based assessment, and then factor in some of the changes that are proposed to be made year in, year out to see which offers the most workable and least disadvantaged of those combinations?

Professor McLean: To the best of my knowledge—but I can be corrected by those closer to the process—Smith did not do this because it did not have time. It was under a mandate to report in an immense hurry. The previous Calman commission—or, more precisely, the independent expert group for Calman, of which I was a member—did do this exercise to a certain extent. It considered what might be called Australian-style, as I have just outlined, and Canadian-style equalisation rules. A Canadian-style rule would be characterised as one that attempts to compensate for the differential tax bases of the provinces but does not attempt to make a needs assessment. The Calman independent expert group, which heard from both Australian and Canadian witnesses, plumped for something that was closer to the Canadian than the Australian model. That is the last time I am aware of that it has been considered by an expert committee.

Lord Turnbull: I wonder about this point about the Barnett formula supposedly having mechanisms within it that would produce convergence and actually doing the opposite. Am I right in thinking that the population share—the 10 85ths, or whatever it is now—is supposed to be revised census by census? As a matter of fact, do you know how often it has been revised? It seems to me that this number has been around for a very long time.

Professor McLean: I understand that the population relativities are revised at every reissue of the Treasury document—I am blanking on its name—that comes out at each Budget. In principle, I think the population relativities are revised annually, but I do not know whether the relativities that they use are themselves updated annually.

Lord Turnbull: They are revised each time there is a spending review, which used to be annually, but now they are sometimes two years apart or even three.

Professor McLean: Per spending review, then; I apologise. The statement of funding principles is now revised at each spending review—

Lord Turnbull: What I think happens is that the Scottish population grows more slowly than the UK population, or the English population, and therefore this number is always lagging behind the reality. For a number of years, Scotland is being paid for people who are not actually there, based on what the population relativity was five years ago or something like that. It is quite unlike what happens with the taxman. If you put in a tax return and he says, “You didn’t pay enough tax last year”, he will do two things. One is that he will change the coding—that is, the 10 85ths—so that for the next year you are paying the right amount of tax for the income that is expected for that year. Then he will say, “I am going to add a bit more to your coding—actually, to take a bit more off your coding—to correct what you didn’t pay last year”. What happens with Barnett is that that second adjustment never gets made. Each time the base is a bit higher than was really justified, it stays in the base in perpetuity and gets added to. Over 40 years, that turns out to be quite a lot. That is one mechanism.

The other mechanism is that in ’78 the per capita GDP of Wales and Scotland was probably quite similar. Over those 40 years, Scottish GDP has risen—indeed, the Scots boast a lot about how prosperous a nation they are—but that has not affected the spending that they are given, so you end up with not only a higher per capita spending than Wales but even less justification for it. If that is how the thing works, the last thing you want to do is to build those defects into a new system and, in effect, to perpetuate them, but that seems to me what we may be doing.

Professor McLean: Right. These are deep waters. On Lord Turnbull’s second point, as the computer scientists might say, is this a bug or a feature? The fact that Barnett is needs blind means precisely what you have just said, Lord Turnbull—that it has not taken any account of the relatively faster increase in prosperity of Scotland compared with Wales and Northern Ireland, so the outcomes now look more anomalous than they did when Lord Barnett was Chief Secretary. That is just a feature of the fact that it is not and does not claim to be a needs-based system.

As to the first point, yes, I agree that the lag is not caught in the way that HMRC attempts to catch the lag in income for taxation. I do not think that that increases divergence, nor do I think that it changes the baseline, because the baseline is the amount spent in England on the relevant service. Its effect is just that as long as the changes in population relativities—we are into second derivatives here—are smaller in Scotland than in England, the convergence is dampened and the long run remains a bit longer.

The only justification that I can see for keeping Barnett is that it was part of the vow made by the unionist parties and therefore it may fairly be said that these were the terms on which the majority of Scottish voters voted no. That, I think, ties the legislature in good faith to continuing to operate Barnett on the reduced proportion of tax that it already covers, after the 2012 Act and Calman, which will be reduced by even more after the current Scotland Bill is enacted.

Lord Turnbull: If that is where we are—and you may well be right on that—that makes it even more important that as we move forward a proper track of this is kept, through the institutional mechanism that you have set out, so that people can understand whether the system is still diverging and not producing convergence. There needs to be some mechanism for regular renegotiation of it. At the moment we do not have a renegotiation—we simply have a recalculation. That seems to me to be an inadequate feature of the system.

Professor McLean: Yes. We can all agree easily that that is inadequate. I entirely agree that having simply an announcement in the periodic statement of funding policy as to how it is going to be operated, with then a little bit of argument between the Governments at the margin as to which expenditures are to be deemed to carry a Barnett consequential and which are not, is an unsatisfactory system. Even if one did not go the full hog of having an equivalent of the Commonwealth Grants Commission, I would like this to be overseen. It could be, I suppose, by one of the existing bodies—just possibly the Statistics Authority or the OBR, which have been created to be at arm’s length from Government—to do the sorts of calculations that you recommend.

Lord Turnbull: The model that seems nearest to me is the Electoral Commission, which is a UK body that operates in Scotland. The Government do not have to do what it says—Parliament can change it—but it is a very influential body. That tells us that we can create these kinds of federal bodies.

Professor McLean: It so happens that, of my last two research trips to Australia, one was to study the Commonwealth Grants Commission and one was to study the Australian Electoral Commission. Over there they have set them up on very similar lines. The UK Electoral Commission is quite similar to the Australian Electoral Commission in the way that it was set up, so we possibly have quite a lot of appropriate lessons from Australia to pick up here.

Q71 Lord Monks: Can I switch the line of questioning a little to the issue of taxation and the taxes that were chosen? In your interesting article in Politics in Spires, you are rather complimentary about the Smith recommendations, particularly given the time pressure that they were under. Are you still complimentary? Would you still say that those are probably the right choice of taxes?

Professor McLean: To be honest, I cannot now remember what I wrote in Politics in Spires, but I think that at that time I was commenting on the Scottish Government’s choice of what to do with the taxes that it already had devolved to it pre-Smith, in particular land and property taxation. If I said that the Smith bundle was right in that article, I had probably better revise that and have another look at the article, because what I would now stand by—if it is different, I apologise—is what I told the Scottish Parliament last week: that the tax powers and, even more so, the spending powers that are devolved under Smith are a bit of a rag-bag. Because this applies particularly to the spending powers devolved under Smith, the tax powers devolved under it look less coherent than they otherwise would. In particular, because Smith devolves some spending responsibilities for social protection but does not devolve national insurance—either employers or employees—I would say that it creates a new anomaly.

I would probably not have devolved the whole of income tax. I heard the last five minutes of David Phillips’s evidence and agree with the reasons he gives for saying that it would have been better if income tax had been a shared tax rather than a fully devolved one. As I say in my evidence to the Scottish Parliament, if more or different taxes are to be devolved, principles that go back to Adam Smith and David Ricardo suggest that taxes on stuff that does not move are the most appropriate to devolve—those not already devolved are oil and gas offshore, but also mineral exploitation onshore—and taxes on things that move instantaneously, such as corporations, are the least suitable for devolution.

Q72 Lord May of Oxford: I was very taken with your description of the system in Australia. As I was mentioning to the others earlier, my brother spent his life in the middle of that, beginning with “Nugget” Coombs. My question to you is, given that, quite understandably, the Australian way of doing it is best, what do you think is the chance of the countries, particularly Scotland, going along with that? How hard do you think it would be to implement that system here?

Professor McLean: The obvious difficulty is that turkeys don’t vote for Christmas. People in Scotland who are aware of the present system are aware that they get a relatively favourable deal under the Barnett arrangements, even though they may not be aware of why that is.

Lord May of Oxford: That was the assumption behind the question.

Professor McLean: Yes. I think you will find—you probably already have found—that the Welsh Government would be very keen on a needs assessment. We have spoken to successive First Ministers and their staffs about this and they are on board with that idea. Northern Ireland politicians do not seem to know what they think about this matter because they are always thinking about something else. So it might be difficult to get a consensus on a move to a needs assessment.

If, however, a UK Government were to say that the only fair thing to do is to do a needs assessment, they have the power to do that. It would be rather difficult for the devolved authorities to say that they were against a needs assessment. They might then get into the politics of saying, “What is expensive to operate in this country is a need, and what is expensive to operate somewhere else is a policy choice”. I have accompanied Treasury civil servants when they have gone round in these Barnett consequential negotiations and have heard them use somewhat salty language about, for instance, the maintenance of two school systems in both Scotland and Northern Ireland, namely a Catholic and a non-denominational one. I think that the Treasury view—I am speaking for what I understand to be the views of Treasury officials, not of Treasury Ministers—is that having two parallel education systems is a policy choice and should not be regarded as a need and that, therefore, were we to move to a needs-based funding system, Scotland and Northern Ireland should take the consequences of having made that policy choice.

Q73 Baroness Wheatcroft: Clearly, the driver behind the current moves towards devolution is politics rather than economics. Professor McLean, I wonder whether you have found any evidence that devolution of powers to regions or countries produces per se any economic benefits.

Professor McLean: There is no strong evidence about this. There are contested claims. There was—as some members and, certainly, some advisers may know—a rather spectacular bust-up, if I may call it that, in the previous Scottish Parliament when expert witnesses who claimed that devolution did of itself bring economic benefits were systematically taken apart by the then Chair of the Scotland Bill Committee. I think the case has not been made, but it is perfectly possible that there are such benefits.

Where I would first look for the benefits would be in the area of land and property taxation. I repeat that, since I came in at the tail-end of David Phillips’s evidence, I know that you had a discussion of that matter and I would go along with what I heard him say. Scotland was given the power to replace what all experts in the field regard as the most indefensible of all taxes in the quiver, namely stamp duty land tax. They have taken that opportunity, and kudos to them—they have introduced a land and buildings transaction tax. However, one could look at the whole suite of taxes on land and property, which the Scottish Government are currently doing, as members may be aware. They have appointed a property tax commission, to which I have given evidence. There are ways of making the suite of taxes less regressive and more incentive compatible with tax activities that are, in the view of some experts, currently undertaxed. The Scottish Government are at least looking at those options. If I may be slightly cynical, the angry reaction of, in particular, sporting landowners in Scotland suggests to me as a rather cynical political scientist that the Scottish Government may be on to something here.

Q74 The Chairman: Can we look at the arrangements that are being proposed for fiscal autonomy in Scotland, in comparison with what has happened in other countries? We have a useful piece of information here that shows on a graph—the source is the OECD—exactly where it sits in comparison. From a qualitative point of view, it would be helpful to know how you see it in comparison with what has happened elsewhere and whether devolution of fiscal measures—a degree of fiscal autonomy—has worked successfully in other countries, and why.

Professor McLean: I have looked at the OECD numbers in the past, with my co-authors Guy Lodge and Jim Gallagher. I do not have any up-to-date OECD tables in front of me, so I cannot comment on those. What does it mean to say that a devolved tax system works? I suppose a working devolved tax system is one that is able to adapt to local conditions in all the sub-national territories while creating as few perverse incentives to players in the game as possible. If one looks at mature systems such as Germany, Canada, Switzerland and Australia, at an amateur level I can only say that they all seem to bumble along more or less okay. The only two of which I can say something a little more informed than that are Canada and Australia. We went through this in the Calman independent expert group, and I can say again that both of these are systems that work. They do not eliminate the perverse incentives, but those who have designed the system have thought about those incentives and tried to design institutions that suppress them as far as they can.

The Chairman: Have you detected in those systems in other countries some clues as to what needs to be included in the arrangements to give them the best possible chance of success?

Professor McLean: I would say two principles: non-gameability and incentive compatibility, which I illustrated in the Australian case a moment ago. A system should be non-gameable in that it does not give a player, which would be a Government in the system, an incentive to claim poverty by claiming something they have particularly much of as a need. It should also be incentive compatible in the sense that gains from tax effort are not equalised away immediately. Were I designing a system from scratch, I would adopt those two principles.

The Chairman: Is there not a danger in the fact that there is a second principle of no detriment—the ongoing no detriment—which is widely considered to be unworkable and has been put on a shelf as a general principle? Nevertheless, the expectations that flow from that are that, indeed, there will be no detriment over time, yet almost by definition there will be detriment over time—or benefit over time—because autonomy seems to me to imply that. That is one of the arguments for a degree of autonomy. Is there not a problem that, right at the heart of this, there is a principle that could undermine the workings of it and, therefore, the reaction of Scotland and the Scottish people to what subsequently emerges?

Professor McLean: I agree that the second Smith “no detriment” principle is very hard even to parse or interpret. It is also very hard to understand why it should be justified. It states—I have it in front of me—“No detriment as a result of UK Government or Scottish Government policy decisions post-devolution”. A strong reading might say that a tax change is a policy decision and therefore one territory must be compensated for any tax change made by the other. That is patently absurd, so if “policy decisions” does not include taxing decisions, what does it include? Maybe it includes decisions to spend more or less or to change the basket of taxes. There are various interpretations of what this could possibly mean. The only way that I have seen in which the second Smith “no detriment” principle could be made workable would be to restrict it very narrowly to decisions about the rates and base of a tax in one territory that would have knock-on effects in the other. You could do that.

If you apply it to spending decisions, you are constraining the very autonomy of the devolved Governments that the whole successive Calman and now Smith processes are supposed to be about. Members may notice that in my evidence to the Scottish Parliament I gave what was intended as a slightly tongue-in-cheek example, but one to alert MSPs to the issues. As I read it, if the Scottish Government use their to-be-acquired powers over the Crown Estate in a way that is suggested by the explanatory notes to this Parliament’s Bill, and that is less commercial, the second Smith “no detriment” principle could lead to a claim from the UK Government for compensation for the Scottish Government making less tax effort in the form of what it gets from the Crown Estate. The example may be fanciful, but it was designed to illustrate to the Scottish Parliament that the second Smith “no detriment” principle is expressed symmetrically.

Since the population of the rest of the UK is 10 times greater than the population of Scotland, were a UK Government to play hard ball over the second Smith “no detriment” principle, it would severely constrain the freedom of the Scottish Parliament to take autonomous decisions. That is clearly not what the parties who negotiated Smith can have wanted, so one has to try to do some rational reinterpretation of what they may have meant by the second Smith “no detriment” principle, and we face the problem that they may have intended them to mean different things. I can only agree that it is very hard to interpret.

Q75 Baroness Blackstone: Could you tell the Committee how feasible you think fiscal autonomy is for Scotland?

Professor McLean: Did you say “political” or “fiscal”?

Baroness Blackstone: Yes, in a political sense—full fiscal autonomy.

Professor McLean: I differ from a number of people who write about this in thinking that full fiscal autonomy is feasible. It carries heavy consequences, of which an obvious one is, if Scotland has full fiscal autonomy, is there any case for it to be represented in this Parliament? But that is not a fiscal issue. There is a principle that I believe the current Scottish Government would really like. You would have to ask them, and you probably would not get a straight answer if you did, but I think the present Scottish Government would like a regime in which Scotland controlled all the tax rates and bases applicable to Scotland, used whatever proportion it took on devolved services and sent the UK Government an annual cheque to pay for defence, foreign affairs and debt interest. In principle, that was the system devised for Ireland and then Northern Ireland in the 1880s, and again in 1921.

The obvious warning is that it has never worked in Northern Ireland. Northern Ireland always has been a poor territory. It could possibly be made to work in relation to Scotland, given that Scotland is an average-GDP territory, and it would have the merit that it would force the Scottish Parliament as nothing else would, apart from complete independence, to face fiscal reality, which the design of tax and spend in Scotland—in the other two as well, but notably in Scotland—has prevented it from having to do since the Scottish Parliament was recreated.

Baroness Blackstone: Do they want it just as a next step towards independence? Is that really what is behind the drive for full fiscal autonomy?

Professor McLean: It is not for me to guess the motives of the Scottish Government, but I believe that quite a lot of SNP politicians would be quite content with full fiscal autonomy, if only because that is a popular position among the people of Scotland. If you ask the people of Scotland which Government should do what, they say, “We are happy for the UK Government to do defence and foreign affairs, and the Scottish Government should do everything else”. If that is the allocation of spending responsibilities, I think it is right that the allocation of tax responsibilities should follow that.

There is a problem. No less a person than the nationalist Convener of the Committee to which I gave evidence on Thursday said, “Isn’t this transfer of tax powers a poisoned chalice?” He quoted your own adviser, Professor Bell. I said, “Are these Professor Bell’s words or yours?”. The Convener said they were his words. I said, “It is a consequence of any further devolution of tax powers that the Parliament to which they are devolved has to be more serious about matching its taxing and spending”, or words to that effect.

Q76 Lord Turnbull: When we were in Scotland last week, one of the phrases that got a lot of pushback—I cannot remember the exact words, but it is in the UK Government’s document about the way forward—was something like, “We would expect Scotland to make a commensurate move in deficit reduction”. They did not like that. Of course, a lot of the political rhetoric is, “We want an end to austerity”. The UK Government may want to reduce their deficit to zero and even go on to make a surplus, which is controversial even here, but they say, “We don’t want to go that far”. They might want to adopt the old Labour Party policy of having a deficit only of the current budget, and they might have a view on the time over which they wanted to reduce the deficit. How do you reconcile the fact that the UK Government have an overall fiscal policy and powers relating to monetary policy, yet one part of the UK is saying, “We want to be able to determine our bit of that ourselves”?

Professor McLean: I would have to say that, for better or for worse, that is a consequence of the way the Scottish people voted last September. The majority of those who voted, voted to remain in the UK. There is no country in the world—nor, I think, could there be—in which determining the overall fiscal balance is the responsibility of anything other than the upper tier of government. It then devolves into the pure politics of it. Given that the vote was no, it is a necessary consequence that the UK Government are entitled to make the sort of statement that you have just quoted. It is also in the nature of politics that a Scottish Government controlled by a nationalist party is always going to object to that.

Lord Turnbull: It follows from that that all of the discussion about borrowing powers is not about this basic trajectory down to a balanced budget that the UK Government want, but about quite minor variations around that trend. It cannot differ fundamentally; it certainly cannot go off in a different direction. In the end, it always has to come back, even if it takes slightly longer over it. Do you think that thinking people in Scotland have recognised that?

Professor McLean: Here is another point where I may differ somewhat from other academics who look at this. I would maintain that a good way to deal with the contradiction we have just agreed on is for the Scottish Government to have relatively wide borrowing powers, with the effect that market discipline would then be applied to Scottish Government debt. Of course, there would be a question of whether Scottish Government debt was impliedly backed by the UK Government. The UK Government would presumably insist that it was not and would have an incentive to put their money where their mouth is. These politics could then get very painful, but it is safe to presume that Scottish Government debt would carry a risk premium over UK Government debt if the Scottish Government had more serious borrowing powers. For my money, that is all to the good from the point of view of fiscal discipline.

Lord Turnbull: Is this not going to replicate the essential fault of the euro? Ireland, Portugal and Spain all had debts, with risk premiums virtually nothing over German debt. However, when it came to it, it turned out that the markets had guessed correctly that there would be a sort of bail-out. The idea is that the UK Government, in a monetary union, would not bail out one of their component parts, but Scotland is too big to fail and we have to build around that.

Professor McLean: Certainly, that analogy was a powerful driver of two famous speeches of which you will all be aware—by Governor Carney, followed by the Chancellor—both in Edinburgh last spring, saying that a currency union with an independent Scotland would not be acceptable to the UK, for the very reason that you have just said. Can you credibly say that the Scottish Parliament bears the risk of Scottish debt within a union? Here might be a very brutal view. Scotland is not too big to fail. Scotland is 8% of the UK population. If they were to fail to make repayments on Scottish debt issued by and on the authority of the Scottish Parliament, that would be between the Scottish Government and their creditors. The UK Government would not intervene. Why should they, especially as their stake in Scotland would be much less than now? Here one is freely hypothesising about events that would be high risk. I cannot be 100% sure that the UK would not bail out Scotland in those circumstances, but if it did it would be on terms.

Lord Turnbull: Who would those creditors be? Would they be UK institutions? You then get into the complication that our No.2 and No.3 banks are headquartered up there.

Professor McLean: It would be whoever bought Scottish Government debt. “Headquartered up there” is a bit tricky, but the Royal Bank, let us say, would have to take a commercial decision as to whether—or on what terms or what yield—it was willing to buy Scottish Government debt.

Lord Turnbull: The UK Government ought not to bank on the fact that they could just let them fail. They should always recognise that they may find, when the time comes—if only because of the risk of contagion—that they have to, and therefore they should be aware of that from the start. That is why I think the proposition would be that we should not be offering a kind of “no bail-out” threat, because we would not know at the time whether we would ever be able to deliver on it—or want to deliver on it.

Professor McLean: That view may be right, but a concomitant of that is that you are not then going to feel free as the UK Government to give the Scottish Government any extensive borrowing powers. Then we are getting back to the political irresponsibility of the regime. That is not meant as an attack on the present Scottish Government. It is just a fact that a Government who are not fully fiscally responsible have every incentive not to behave in a fully fiscally responsible way.

Lord Turnbull: I think you are saying that accountability will come from one source rather than two. It will come through taxpayers, who might pay an extra bill, but it will not come through pressure of creditors, because we will not be allowing them to play a big part in this.

Professor McLean: Yes. That may be the necessary situation, in which case it is a long way short of the aspirations of the Scottish Government.

The Chairman: Professor McLean, thank you very much indeed for your clear and helpful answers this afternoon.

Professor McLean: Thank you for putting me through a considerable challenge.

The Chairman: Good luck with your forthcoming appearances before other Committees.

Professor McLean: Thank you.