Our second item is to continue to take evidence as part of our inquiry into Scotland’s fiscal framework. I welcome David Phillips and Professor Alan Trench. Good morning to you both. Both witnesses have provided submissions in advance of giving evidence, so we will go straight to questions from the committee. I will ask some opening questions then we will go round the table.
We will look at David Phillips’s submission first. Alan Trench is free to comment too—and vice versa when I touch on Alan Trench’s submission. The first thing that I want to talk about is the no-detriment principle.
David, your submission states:
“The key point is that it is not possible to design a system that meets both no detriment principles, and is at the same time ‘transparent, effective and mechanical’. But it is possible to design such a system that is broadly ‘fair’ if one is more flexible about the no detriment principles. Such flexibility will be key to implementing a workable system”.
You go on to explain why there are contradictions within what Smith has said on no detriment. Can you expand on that a wee bit for the record?
Yes. There are two no-detriment principles—actually, there are kind of three, but the framework has two.
The first no-detriment principle is that, when a tax is devolved to Scotland, Scotland should not be made better or worse off just from the devolution of that tax. That implies that we take off the block grant the amount of revenues that are devolved to that tax.
The way that I have interpreted that is that the spirit of that principle means that, in the years ahead, Scotland should not be worse or better off if things continue on as they were going to anyway, so if the Scottish economy performs in line with the United Kingdom economy, Scotland should not be made worse or better off under the devolution. That seems to be the spirit of that first no-detriment principle.
The second no-detriment principle is that, if a policy decision is made in Westminster, it should not make Scotland better or worse off if it is for an area of devolved competency. Likewise, if a policy is made in Holyrood, it should not have an adverse or positive impact on the budgetary position of the UK Government.
In my presentation that I summarised in my submission, I gave examples of where those two principles could conflict. One of the reasons behind that is the difference in the relative size of the tax bases in Scotland and in the rest of the UK. For example, income tax revenues in Scotland are about 15 per cent less per person than they are in the UK as a whole so, to satisfy the first no-detriment principle, there would need to be a mechanism whereby if Scottish revenues grew in line, in percentage terms, with revenues in the rest of the UK, Scotland would be no better or worse off. If revenues went up by 10 per cent in Scotland and 10 per cent in the UK, you would want a mechanism that meant that Scotland did not suffer detriment, under the principle that it should not suffer detriment because it was growing in line with the UK.
A mechanism that does that can be designed. An amount would be taken off the block grant in year 1 and in subsequent years that would be indexed to the percentage growth in revenues in the rest of the UK. If revenues have gone up by 10 per cent in the rest of the UK, the initial block grant reduction is increased by 10 per cent. So 10 per cent more would be taken off the block grant than was the case originally. That means that, if revenues here go up by more than 10 per cent, Scotland gains, but if it is less than 10 per cent, Scotland loses and if revenues in Scotland grow in line with the UK, Scotland does not lose. That satisfies the spirit of the first no-detriment principle.
However, suppose that there was a change in tax policy in the rest of the UK and that was the reason why tax revenues rose by 10 per cent. That 10 per cent of additional revenue in the rest of the UK would be put into the Barnett formula, so Scotland would get more under the Barnett formula and then have an additional amount taken off its block grant. In that circumstance, Scotland would gain from a tax increase in the rest of the UK. That is not a detriment to Scotland, but it is a detriment to the rest of the UK, because Scotland would be gaining from additional taxes that are being paid in the rest of the UK and not in Scotland. That would not satisfy the second no-detriment principle.
It is hard to explain this verbally and easier to use a written example. Different methods of adjusting the block grant satisfy different principles. The methods to satisfy the first principle of there being no detriment from the act of devolution do not satisfy the principle of no detriment from policy action and vice versa. In the presentation that I sent to the committee, there are some worked examples that show that problem. Both principles cannot be satisfied, so it is a choice as to which principle is prioritised.
Thank you for that clear explanation.
I largely agree with that. No detriment is a perfectly sensible high-level policy objective, but I am not sure that it is a workable, practical principle to use in any system, particularly one that is designed to work in a mechanical way. By its nature, the principle will be pretty subjective in its application and it is hard to see how it can serve that purpose.
Let us consider the history of the idea of no detriment. The first time that I am aware of no detriment being used was in the context of what is now the Scotland Act 2012. The command paper that was published in November 2010 set out the framework and said that there would be a no-detriment principle to adjust for the impact of UK tax decisions on the Scottish income tax take from the Scottish rate of income tax. My reaction on seeing that was to regard it as a very broad and potentially dangerous principle.
Subsequently, the Treasury went away and looked at the work that had been done in a Welsh context on similar proposals by the Independent Commission on Funding and Finance for Wales, which was chaired by Gerald Holtham, who was supposed to be giving evidence today and who I believe sent a note to the committee. That helped to resolve a lot of the problems. Gerald Holtham formulated the principle of the indexed deduction approach that David Phillips has talked about.
That solved one sequence of problems. The difficulty now is partly that excessive weight is being put on the principle in general and partly that excessive weight is being put on the indexed deduction method. The Holtham commission carried out careful analysis and said that there are a number of ways in which we can calculate the reduction of the block grant and adjust it. Different ones will be appropriate for different taxes, because of the different profiles of risk that are involved. The logic of the indexed deduction method is essentially that it leaves with the Scottish exchequer much of the responsibility for dealing with the consequences of devolved Scottish decisions, but it keeps at UK level—and therefore protects Scotland from—cyclic decisions that the UK Government would be responsible for under the current arrangements and under the arrangements that have been set out.
That situation starts to change the more fiscal devolution we talk about. Whether indexed deduction is the appropriate method to use for other taxes is one big question, and whether and when we need to use the no-detriment principle in this way is the second question—that is the bit that is particularly difficult.
David, do you want to come back in?
Yes. I have a slightly easier way of explaining what I was trying to get at in my first answer. The fundamental issue relates to the different ways in which the Barnett formula and the indexed deduction methods can work. The reason why we cannot satisfy the two principles is that the Barnett formula works on the basis of changes in spending on a pounds-per-person basis. If we want to satisfy the second no-detriment principle—the one where, if there is a policy change in England, it does not have detriment in Scotland—the indexation for the block grant reduction also has to work on a pounds-per-person basis so that it offsets the Barnett formula. However, because a given percentage growth in revenues in Scotland is less in pounds per person, doing things on a pounds-per-person basis would mean that, if revenues in Scotland grew at the same rate as in England, Scotland would lose because it has lower growth in pounds-per-person terms and therefore its pounds-per-person growth in revenues would not be as fast as the pounds-per-person increase in the block grant reduction.
On the other hand, if the indexation of the block grant reduction is done on a percentage basis, that is different from the Barnett formula’s pounds-per-person basis, so it would not satisfy the other no-detriment principle. It is all to do with the fact that Barnett is based on pounds per person. In my view, the most sensible way to do the indexation is likely to be on a percentage basis, given that a given percentage growth in Scottish revenues is different in pounds-per-person terms. It is all to do with Barnett and the block grant reductions being on different bases—that is why we cannot satisfy the two principles at the same time.
Percentage growth would surely be impacted by demographic changes, for example if the population grew at a faster rate in the UK as a whole compared to the rate in Scotland.
That could have an impact. There is a real question about whether the indexation method should account for demographic change. Two points are worth mentioning. First, the Barnett formula on the spending side does not account for differential population growth. It updates the population factors every couple of years but it never updates the baseline spending—it only applies the increment. In effect, it does not account for differential population growth. We could say that, if the spending side does not account for differential population growth, should the tax side as well, or are we just picking the ways that benefit Scotland as opposed to the ways that are consistent?
Secondly, as one of the other people who has given evidence has suggested, if we account for differential population growth in the indexation, that removes any incentive to Scotland to improve its demographic profile. For instance, the Scottish Government has said that it would like faster growth in the working-age population in Scotland. If the indexation method was adjusted for population growth, Scotland would get no benefit from that.
On the one hand, adjusting for population growth would benefit Scotland. It would mean that less would be taken off the block grant, to account for the fact that Scotland’s population had grown less quickly, but it would not be consistent with the Barnett formula and it would remove incentives to grow the population, whereas the Scottish Government has said that it would like better demographics in Scotland.
09:45
Professor Trench, you have helpfully called for the setting up of
“an independent body to help resolve disputes”.
Given what we have just heard, I think that that would be very helpful. You have also called for
“Keeping systems under regular review, rather than believing that a system can be introduced and then simply left to work.”
Yes. Would you like me to elaborate on that?
Yes.
Let us start with the first point. We have a fundamental problem with the institutions that we use to deal with financial matters. The problems have been apparent for some time, and they have now become acute. Essentially, one institution does all those things at the moment: HM Treasury. The Treasury is also the finance ministry, as it were, for England as well as for the UK as a whole.
The whole system rests on nothing more than a Treasury statement of funding policy. It does not have a direct statutory or constitutional underpinning; it operates through the mechanism of annual supply and appropriation acts at Westminster.
Any disputes or disagreements need to be raised with the Treasury first and, if one could really be bothered, they can subsequently be pressed in due course to the joint ministerial committee through its dispute resolution mechanism. To be blunt, a fat lot of good that is likely to do anyone, other than the Treasury.
What we need is, at the very least, two new bodies, both of which, I am afraid, will impinge on the Treasury’s ability to make decisions. First, there needs to be some genuinely independent body that is responsible for if not making then advising on the calculations that underpin the system, whether it is the calculation of the quantum of the block grant or the calculation of the amount of any reductions in the block grant by whatever mechanism is used.
There is also a need to do post facto audit, to review what has gone on and to see what has actually happened. It is remarkable how little anyone does that. I think that I referred in a footnote in my submission to the only example of that that I am aware of, which was undertaken by David Heald and Alasdair McLeod. That was published in 2005, based on data from 2000 to 2003, if memory serves. That was a very early spending review period post devolution. That work was possible only because Professor Heald, as a committee adviser, had access to detailed quantitative information from the Treasury that was not normally published then and is not published now. As far as I can see, there is no particular reason why that should not be published—it just is not. Forensic examination shows that there appears to be something of a black hole in how the numbers have worked, even at a time when they were relatively straightforward. The issue is material.
What about the institutional structure of the potential body? One could draw on the parallel of the Office for Budget Responsibility, but I prefer the example that we discussed in some detail in the recent report from the Bingham Centre for the Rule of Law entitled, “A Constitutional Crossroads”, to which I have also referred. There are many problematic features of the Australian system, but the structure of the Commonwealth Grants Commission has proved to be effective and durable. The CGC has existed since the 1930s. It is composed of independent experts, who are mostly people from public service backgrounds who have worked for state Governments and the federal Government. There are usually one or two academics on the commission. They are expressly not elected politicians—the members of the commission have never held any political office. They are accepted as politically impartial, and they are able to do a job of work of advising the commonwealth treasurer about how the system works. They do so sufficiently effectively that the treasurer has always accepted the CGC’s advice. He or she has never tried to alter the recommendations that have come from the CGC. Institutionally, that is attractive.
The second body that I think we need is some better mechanism to deal with disputes and disagreement between Governments. It is difficult to have a UK Government minister deciding on a dispute between a UK Government department and a devolved Administration. There needs to be something genuinely impartial that can impose some stronger sanction on the UK Government than the practically non-existent consequences that exist at the moment. At the very least, such a body could publish a report to state that the UK Government is refusing to act in a responsible manner, if that were its finding. That is where we would need to get to in institutional terms.
I am afraid that I have forgotten the second part of the question.
Basically, I asked about the independent body to resolve disputes, which we have touched on, and about keeping systems under regular review, rather than leaving them to be introduced at some further point.
Again, it is commonplace in federal systems to review such mechanisms periodically. The Australians, being remarkably well organised, have a five-yearly review of the basic framework that is used by the Commonwealth Grants Commission. Every five years, a fairly systematic root and branch review is carried out into the indicators that are used and what they should be. That is in light of directions from the commonwealth treasurer about what the system should seek to achieve—they are told what the system should achieve, and they go away and work out how to achieve it.
Other systems, for example in Canada, review less frequently. There are three or four significant elements to how the Canadian system works. There is the equalisation fund, two transfers called the Canada health transfer and the Canada social transfer from federal Government to the provinces, and a separate mechanism that funds the three territories. There is also the tax capacity of the provinces. Each of those is reviewed on a different cycle, but they are reviewed. Equalisation, which is the largest of the transfer programmes, is reviewed about every seven to 10 years—it is gearing up for another review now I believe.
Thank you. You said:
“the mechanisms proposed by the UK Government in the command paper Scotland in the United Kingdom: an enduring settlement have significant shortcomings and are likely to rely on negotiations that will themselves be based on data of questionable accuracy. This system is unlikely to breed confidence in its fairness and unlikely to be stable.”
Could you comment on that? Has the Treasury grasped the need for a transformation? It is one thing for us to talk about it, but does the Treasury believe that it must take place? Can improvements be implemented prior to the Smith proposals being rolled out?
I think that the Treasury is starting to learn the nature of the situation that arises. It is a learning experience—I do not think that the Treasury is where it needs to be yet, but that does not mean that it will not get there. Whether it will get there in time for the roll-out of the Smith proposals, as you put it, will depend on the timescale for that. I understand that we are expecting a bill to be announced in today’s Queen’s speech and published shortly thereafter. It is likely to be passed before the end of the current session here and during the Westminster parliamentary session that opens today.
I do not know what that means for applying those further powers in a Scottish context—I do not know whether that will happen a year or two after the Scottish rate of income tax comes into effect next year. The long-ball option would be that it will happen after the next Scottish Parliament elections. I have no idea what the position is regarding that, and the white paper does not tell us anything very useful about it. We may learn more when the bill is published.
Can you remind me of the first part of your question?
Basically, I was asking about transparency and how it can be improved ahead of the implementation of Smith. You said in your submission that the system is
“unlikely to breed confidence in its fairness and unlikely to be stable.”
That appears to me to be the case—I am not sure that I can say very much more than that.
As you know, the data that we have on tax receipts across the UK is quite problematic. We also have a situation in which HM Revenue and Customs is struggling to identify Scottish taxpayers in order to make the Scottish rate of income tax system work. That is not helpful because, of all the taxes that we are talking about, income tax is one of the easiest to devolve. A lot more work needs to go into understanding how we attribute the origin of taxes in a geographical sense. That is necessary on both sides of this equation, not just to work out what Scottish revenues are or might be but also to understand what, as it were, they notionally would be, in order to assess how the first of the no-detriment principles that David Phillips talked about will actually operate. If this is supposed to be revenue neutral, we need to have better data in order to do better projections than we do at the moment.
The first and foremost point to start with is the need for better data and more published data. We have reasonable data for Scotland in the form of the “Government Expenditure and Revenue Scotland” figures. It is not perfect, but it is better than the data that we have for anywhere else. We have rather dubious similarly calculated data for Northern Ireland, we do not have anything for Wales, we have some generalised figures for the UK as a whole and we have some experimental data from HMRC. We have to do a lot of work to make that more useful.
David Phillips, in your submission, you say that, to ensure transparency and effective scrutiny, information on the calculations relating to the block grant and the Barnett formula
“should be published in full at every fiscal event which affects the devolved governments’ block grants.”
Yes. Recently, I tried to do a similar exercise to the one that was done in the 2005 paper that Professor Trench mentioned.
Professor Trench touched on the fact that, although the principles of how the Barnett formula works are published and quite a lot of detail is provided about how the individual departments’ budgets should feed into the Barnett formula, there is no published information on the calculations that are made at the time of each spending review or budget, and there are complicated issues such as how baselines change between periods and so on. The implementation is quite opaque, and information is not published.
In 2009, the House of Lords Barnett Formula Select Committee recommended that that information should be published, but that call was not acted on. I pestered the Treasury and managed to get hold of the spreadsheets that it uses to do the calculations, but we should not have to do that. The spreadsheets should be publicly available so that the work can be critiqued and analysed so that we can understand what the budget allocations are for Scotland, Wales and Northern Ireland.
In the answer to what I think was question 3 in the committee’s call for evidence, I said that I thought that the Treasury or the proposed independent body that might take over the role should publish all that information so that it is open to scrutiny by Westminster and the devolved Governments. The Barnett formula is not just a principle; it is an operational thing as well. Many of the decisions are about what specific things—spending on the Olympics or Kew gardens, for example—are counted towards the Barnett formula. Seeing how all those things are counted and being able to replicate the figures and assess them is important for transparency and for maintaining the integrity of the system.
I was the adviser to the House of Lords Barnett Formula Select Committee. It is not quite correct to say that the changes are not published, because they have been, subsequently. It is one of those things that the UK Government has done without formally announcing that it has done it. The Government puts the figures in some very odd places—I forget where I found the figures for Scotland. It publishes figures for changes that have been made in relation to the baseline or when there has been something that triggers a consequential—a little addendum that sets that out is published each year.
I happen to follow the issue more closely in relation to Wales than to Scotland, and I know that the figures for Wales are, slightly oddly, published in the Wales Office’s annual report, even though it is quite hard to see what they have to do with the Wales Office.
The only issue there is that it does not break down the calculations into their individual sub-components—it just gives a total amount.
Indeed; it is very broadbrush, but it is more than we ever had before.
10:00
I will touch on just one more area, because my colleagues want to come in.
The issue of borrowing is one on which you have different views, which is always interesting for the committee. David Phillips, you say in your submission:
“while for local government borrowing, there are effective means for central government to deal with any problems that may arise if a particular authority borrows what is felt to be an imprudent amount”.
You go on to say that
“the Scottish and UK governments should agree a limit on capital borrowing powers.”
Professor Trench, you say in your submission that
“Scotland in the United Kingdom, in particular, suggests a highly constrained approach, most notably in paragraph 2.2.6”.
You go on to say that
“borrowing choices as a zero-sum game, in which devolved decisions count against UK ones, is an inappropriate way to ensure fiscal devolution works.”
I was asking the question on the borrowing powers with the ideas of the Chartered Institute of Public Finance and Accountancy in mind. CIPFA’s submission contains the idea that Scotland should move towards a system of prudential borrowing for capital borrowing powers. A similar system is used for local authority borrowing, and CIPFA argues that, because that works well for local authorities and we have not seen local authorities getting into trouble, why should we not give the Scottish Government a similar regime of borrowing?
Under a prudential borrowing regime, the Scottish Government would determine its own borrowing limits subject to affordability, and the amount borrowed would be used only for capital expenditure and not current expenditure. It would be an overall prudential regime.
I am not saying that that is definitely the wrong approach. I am saying that there are potentially some differences in the politics between how things work at local authority level and how they would work for the Scottish Government. In particular, such a regime works well at local authority level for two reasons. First, the Government—the UK Government in England and the Scottish Government in Scotland—has the political power to intervene if it thinks that local authorities are borrowing imprudently, even though they have a regime that means that they should not be able to act imprudently in their borrowing. The Government could cap borrowing powers or intervene by sending in commissioners. In effect, it could take back control of those borrowing powers.
In principle, Scotland within the UK could have a regime like that, but the political ramifications would be substantially different from what they would be at local government level. It would cause a constitutional crisis if the UK Government commanded a prudential borrowing power and, at some point in the future, took it back from Scotland.
You are assuming that Scotland would get into some kind of bother.
Indeed. I am not saying that it would, but there is the potential for that to happen. Given the political situation, there might not be the good will on either side for such a system to work. There is at least a risk of that.
The second issue is bailout. If local authorities get into trouble and they fail and need to be bailed out, the amounts are likely to be relatively small. The politics of that would be that people would grumble and they would not like it—people in Glasgow might have to bail out people in Edinburgh, for example. However, the politics of a bailout between the UK Government and the Scottish Government after fiscal autonomy or further devolution would be of a different order of magnitude.
I am not saying that a prudential borrowing regime is a non-starter, but the political issues are quite different from those that apply to prudential borrowing at local authority level. Those would need to be considered carefully by the Scottish and the UK Governments before they decided to go down that route.
If those risks are seen as being too high, a regime with an expanded amount of capital borrowing powers, with a limit, might be a more workable approach that would not have those political problems. Because the limit is known in advance, there is no potential for grievance later on.
It seems that your approach is wholly negative. It is odd that a Government in Edinburgh would have fewer borrowing powers than the local authorities over which it exercises significant powers. Do you not think that that is a pretty illogical place to be?
I agree that you might come to that conclusion if you looked at the government levels on their own and saw that the higher level of government had fewer powers than the lower level of government. However, the issue that I was raising in my submission was the fact that the political risks are different. It is not just a question of the level of government; it is a question of whether, given the political situation and the institutional arrangements, the same mechanism is appropriate. I am not saying that a prudential borrowing regime would not be the right answer; I am saying that there are political issues that mean that it might not be the right answer, even though it works well for a lower level of government.
It is hard to find any objection to the idea that the Scottish Government should have a prudential borrowing power. That issue is not a concern to me. Greater issues arise as a result of the more serious borrowing power that is necessarily implicit in fiscal devolution, which is a power to borrow to address the volatility implications. If one looks at the figures, the amount that you would borrow in those circumstances is much greater than the amount that you are likely to borrow in any five-year period under any prudential power.
That is the point at which things can start to become difficult. We are talking about substantial amounts of money, so fiscal deficits would have to be run and there would be substantial debt service costs. You have to address the question of bailout from the outset—you have to work out who is the lender of last resort.
Again, there are experiences around the world that we can look at. Some countries have very constrained borrowing powers for sub-state Governments, in some cases because sub-state Governments have gone bust in the past and the result has been to constrain considerably their borrowing thereafter. By contrast, there are some Governments that operate with practically no constraint whatever.
Canada is an intriguing example. The provincial Governments go off and borrow very substantial amounts of money on the capital markets by issuing their own bonds. There is absolutely no guarantee of a bailout from the federal Government, and the market appears to believe that. There are cases in which the market does not believe that and thinks that there would be a bailout from the federal Government, which in effect reduces the risk premium that the sub-state Governments would pay on their interest charge. Equally, the Canadian provinces have quite a wide spread of interest rates and risk premiums, although I do not think that any of them has ever defaulted; that has certainly not happened since Newfoundland entered confederation. When Newfoundland was an independent dominion, it overborrowed and went bust as a result. That story resulted in Newfoundland ultimately becoming part of Canada.
In federal countries—this is not a risk that presents itself in the UK context—that situation can affect federal level public finances as well. It has affected Brazil’s public finances. When the richer states of Brazil went off and borrowed a great deal and were unable to pay it back, they had to be bailed out by the federal Government. As a result, the whole borrowing regime in Brazil needed to be reconstructed, because the creditworthiness of Brazil and the exchange rate of its currency were affected.
That happened in the context of some of the largest and most prosperous states of a pretty large country. In relation to Scotland, the risk from a UK point of view has to be different, because Scotland has less than 10 per cent of the population of the UK as a whole. The issue becomes one of UK-wide equity. A system cannot grant Scotland free money simply because Scotland wants it, if that causes disadvantage to the rest of the UK. Finding the way through that is the difficult part.
What is the answer, then?
I hesitate to offer an advanced answer. I am sure that your advisers have been following the work of Angus Armstrong of the National Institute of Economic and Social Research, who has been considering that question in some detail and who is strongly of the view that there needs to be an extensive borrowing power, with the scope to go bust.
I disagree with Dr Armstrong, because I think that there is sufficient risk for the UK Government that it cannot simply say that there will be no bailout of Scotland or that Scotland can borrow freely but at its own risk without there being at least an implicit assumption in the markets that there would be some bailout if things were to go wrong.
My own view, which is not a particularly carefully researched one, is that the best way through this thicket is probably for there to be some sort of ceiling—how that would be calculated, I do not know—up to which the UK Government expressly agrees to indemnify Scottish bonds and above which it expressly says that it will not indemnify any borrowing. Therefore, if Scotland wanted to borrow beyond that amount, it would very clearly do so at its own risk. I think that that is about the best way of balancing the various considerations that are involved.
David, do you want to make a final point before we move on?
As Alan Trench says, there is a real need for Scotland to have further current borrowing powers to smooth the cyclical volatility. To some extent, depending on what block grant adjustment mechanism is used, some of that volatility will be taken up by the block grant mechanism. That is the whole point about the index method of Holtham insulating Scotland from the aggregate cyclical risk.
There will still be idiosyncratic, Scotland-only cyclical risk, which could be quite substantial the more taxes are devolved to Scotland. At the moment, under the Scotland Act 2012, current borrowing can only be used to borrow for forecast errors, but the Scottish Government might forecast a recession and need to be able to borrow for that. There is a need for substantially larger powers on the current side. That is one area in which we need to go much further than is provided for at the moment. There was an indication in the command paper that the UK Government was pushing down that road.
Thank you. I open up the session to members.
On the block grant adjustment, the scenarios from David Phillips were useful. My sense from listening to different people is that the proposal that relates to the percentage change in UK tax revenues is the one that commands the most support and I imagine that it will be the one to happen.
I may be missing something here. You said that Scotland would benefit from tax increases in the rest of the UK, but does that not work in two ways? Is the assumption that Scotland would get more through Barnett consequentials? Surely an increase in tax taken in the rest of the UK would increase the amount taken off the block grant. Is that not right?
10:15
The scenarios that I was talking about were where Scotland would benefit from spending in the rest of the UK. I will give a simple example that shows that.
Let us suppose that, in the rest of the UK, income tax is put up by 2p in the pound. That raises about £10 billion, which is about 8 per cent of income tax revenues in the rest of the UK. Using the percentage indexation method, if income tax revenues in the rest of the UK have gone up by 8 per cent, you would take 8 per cent more off the Scottish block grant to account for that. Eight per cent of Scottish income tax revenues would be about £850 million, so you would take an additional £850 million off the block grant.
Under the Barnett formula, if there was a £10 billion increase in spending in the rest of the UK, that would lead to about £1 billion for Scotland. If you have taken £850 million off the block grant to account for the higher revenues in the rest of the UK, and then Scotland gets £1 billion through the Barnett formula, Scotland gains about £150 million from a tax change that has not affected Scotland but has affected only England, Wales and Northern Ireland. That is because the Barnett formula works on a pounds-per-person basis and the percentage method of indexation is based on percentages, and the pounds and percentages do not necessarily line up, because a given percentage increase in revenues in Scotland is not the same as pounds-per-person increase in the rest of the UK.
It could be the case that Scotland gains a bit from tax increases in the rest of the UK but that it loses a bit from tax decreases in the rest of the UK. The reason why I have come to the conclusion that that would be okay for Scotland is that those things would balance out over time. Income tax revenues will sometimes go up in the rest of the UK and they will sometimes come down, and those things should balance out, but there could be some small knock-on effects from individual tax changes in the rest of the UK, just because of the interaction of the Barnett formula with the block grant.
That is assuming that all the tax increase is spent on devolved areas.
That is right. If it does not go on devolved spending, you get into an interesting situation. My perspective on this is slightly different from one of the other people who have given evidence to the committee. If that money was spent on something that was not devolved, such as defence, no additional money would flow through to the Scottish Government, but £850 million would still be taken off the block grant.
Some people have suggested that that is unfair, because a tax change in the rest of the UK has funded some non-devolved spending and caused a reduction in the Scottish Government’s budget, which means either cutting spending in Scotland or having to increase Scotland’s own income tax to make up the difference. I think that, in the context of a system in which there are still reserved, non-devolved matters, it is not unfair. That is the way that the system has to work for fairness to be in place, because if Scots are benefiting from spending at the UK level—whether it is on defence, state pensions, foreign affairs or the existing national debt—fairness requires that Scots, as well as those in the rest of the UK, contribute to that spending. In the first instance, that can be done by cutting the Scottish Government’s block grant. In the second instance, if the Scottish Government wants to make up for that, it can also increase its income tax rate, just as happens in the rest of the UK. My perspective on that is a bit different from that of some others.
We could make lots of comments on that, but at least you have given us a clear and helpful explanation.
Moving on to the Barnett formula, I believe that you, like most intellectuals, want to change it or get rid of it, and I suppose that a lot of back-bench MPs do as well. My own sense is that it is significant that none of the members on the front benches in the House of Commons does, because they understand the politics of it. However, what I want to pursue is whether we actually need to do that or whether, under the new regime, it will operate as it is supposed to operate in terms of convergence. The main reason why there is no convergence, as far as I understand it, is that the population of the rest of the UK is growing faster than the Scottish population.
I can see Professor Trench shaking his head, so perhaps he will explain his position on that. My understanding is that, under the new regime, if the rest of the UK’s population grows faster than the Scottish population, we will lose out in the block grant adjustment because the tax revenues in the rest of the UK will increase at a higher percentage than ours. Will the new block grant adjustment regime actually make the Barnett formula work more as it is supposed to work in terms of convergence, so that there would no need to get rid of it, as you propose?
To some extent, it depends on what indexation method is used to index the block grant reductions for the tax devolution. The first in the order of methods that will lead to the Barnett squeeze taking place more quickly is indexation, not in percentage terms, but in pounds-per-person terms. That would lead to the fastest squeeze, as Professor David Bell showed in his analysis. A given percentage increase in revenues in the rest of the UK is more, in pounds-per-person terms, than in Scotland. If indexation is in pounds-per-person terms, Scotland’s growth would have to be faster just to keep up with the rest of the UK, so the squeeze would definitely take place under a pounds-per-person indexation.
With percentage indexation, which is the one that I talk about mostly, Malcolm Chisholm is right: if the Scottish population grew less quickly and the growth in revenues per person in percentage terms was the same, that would also lead to the squeeze increasing, relative to the position today.
Finally, if block grant reductions are increased not in percentage terms but in pounds-per-person terms, that would not lead to any squeeze at all, as the population effects would be counteracted.
Therefore, it depends on the precise indexation method that is used to adjust the block grant over time, but Malcolm Chisholm is right: under the method that I discuss, that squeeze would take place.
Differential population change is one of the reasons why the Barnett squeeze has not happened so much. Another is because—this applies to recent years—the squeeze does not take place when spending is being cut; it takes place only when spending is being increased.
There are other factors in the background that have meant that the squeeze has not taken place. One big factor was the devolution of rail spending in 2006-07. Scotland’s percentage of rail spending went up from around 3 per cent to 8 per cent overnight. Spending jumped in those years.
There were some other changes, as well. Formula bypass happens, too. Top-ups are made for the devolved Governments.
Professor Trench was shaking his head.
Yes. Let me try to address the convergence issues.
The arithmetic and the structure of the Barnett formula mean that there should have been dramatic convergence, particularly during the huge public spending boom of the early noughties—roughly from 1999-2000 to 2007-08—but that did not happen. It should have happened, but it did not.
I am not sure about the rail spending point that Mr Phillips mentioned, but I am very dubious about formula bypass, because there is very little evidence that it went on at that time. Indeed, formula bypass has become very hard to achieve and very open when it is achieved, because decisions are made about what is within and outwith the block grant.
What happened to the population of Scotland—not that of the UK as a whole—during that period is material. For most of that period, Scotland’s population was still falling at a quite dramatic rate—0.3 per cent, or thereabouts, a year, I believe. Since 1999, we have rebased the numbers that are used for the Barnett formula every three years through the spending review process. It did not happen regularly before then, and that is part of the reason why convergence did not happen during the 1980s and the 1990s. However, it appears that, when we look at spending on a per capita basis, even during those short spending review periods the rate of population decline in Scotland was sufficient to cancel out the convergence effect. Therefore, the population factors and their accuracy become very important.
It is worth noting that the Office for National Statistics figures for population change for the next 15 to 20 years suggest a much slower rate of population decline in Scotland—the figure is around 0.1 per cent a year. That means that convergence should start to happen.
Do you want me to talk a little bit about the Barnett formula, as well?
That was the point I was raising, so I think that you are agreeing with me that there will be more convergence.
I think that that is very likely. You addressed Mr Phillips about people wanting to get rid of the Barnett formula.
Yes. Was that provocative? I thought that it was a fairly safe point to make.
Indeed. I have to admit to being a member of that club as well. Indeed, we very specifically say that in the recent Bingham centre financial report. That is partly because when we were drafting the report and considering how to put the recommendations, we thought that there was very little point in not saying, “Get rid of the Barnett formula,” because what we were proposing amounted to doing that.
We could have avoided talking about getting rid of the Barnett formula, which would not have got us a headline, but that would have been misleading. Thinking about what the Barnett formula is, we use the term to cover three things. One is the quantum of money that is paid to devolved Governments every year from the Treasury; the second is the formula itself—the means by which that amount is calculated; and the third is the administrative machinery that goes with it and the fact that the Treasury retains so much authority in relation to all the decisions about it.
As I said in earlier answers to the convener, once you have started to take apart and reconstruct the administrative machinery—once you have started to do the fairly systematic review that you need to do of the formula and how you calculate the numbers—you are ultimately going to have an effect on the quantum anyway. The quantum is not the target, but once you have taken apart the two things that lead to the quantum, you might as well say that you are getting rid of it.
I suspect that what you would end up with would have certain characteristics in common with the Barnett formula. I cannot see any way that you could reconstruct the grant element of funding without using England as your reference point, which is what the Barnett formula does, because you get a consequential share of changes that are made in spending for England. It is inevitable that that will have to be your reference point because there is no other suitable reference point.
However, with that exception, you are going to end up with something that is so changed that, even posthumously, it would be a favour to Lord Barnett to take his name off a formula that he had long been seeking to disown.
I could pursue that but I do not really have the time, as I want to ask one more brief question about David Phillips’s update, which is very topical and very interesting.
I want to ask about national insurance contributions, because you make two statements that appear to be contradictory. On the one hand, you say that
“NICs are ... just another income tax on earned income”.
You refer to Mirrlees, and we know that Mirrlees wants to merge the two. However, you then go on to ask
“who would pay for the pensions of people who had worked in England and retired to Scotland ... ?”
I do not really understand why you ask that question, because I do not think that the Scottish National Party is proposing that pensions should be paid any differently from how they are paid at present under the devolution of national insurance contributions, unless I misunderstand its position.
My initial reading of the manifesto suggested that it was looking for the devolution of not only national insurance but, potentially, welfare in its entirety. That is why I raised that point.
I completely agree—and it is a long-standing position of many researchers at the IFS—that national insurance is just another income tax. However, the UK Government maintains the pretence that it is a social security contribution. There are notional links between NI contributions and benefits received.
I asked who would pay those pensions because if there was full devolution of NICs and welfare, such questions would need to be answered. I think that the answer would be that the pensions would be paid by the Government where the pensioners lived and there would be an adjustment to the block grant to account for pensioner population. However, some technical issues would be involved, given the pretence that we have a contributory system in this country.
But does the general view of the Institute of Fiscal Studies support the Mirrlees position that we should just get rid of it? At present, there is still a fund, but it is not clear to me what its purpose is.
The fund is notional. It does not really link in any way to what gets paid out in benefits. It is an accounting exercise.
Is it notional? You still hear that so much comes from the fund and so much comes from general taxation.
You do, but unless something is completely separate it is just what we call dodgy hypothecation. It does not matter where the money comes from because you just pay, wherever it comes from. What you label it as does not really matter.
10:30
The intricacies of the national insurance fund are something that I would not recommend anyone to try to get their head around. They are quite bizarre, because the fund is underwritten by general taxation—the consolidated fund—so any shortfall is made up from general taxation. However, the fund is used to pay not only benefits such as pensions but a variety of other bizarre charges, including an element of funding for the national health service. It is a system that has badly needed reconstruction for a long time, but nobody has done anything to it since, I think, 1975. In essence it is the system that was put in place in 1948.
It is still our view that the systems should be merged. The devolution settlement as it stands—with income tax devolved and national insurance not devolved—prevents that from happening. Devolving NICs as well could give you a situation in which either Scotland on its own or the rest of the UK on its own, or both, could seek to merge the systems, notwithstanding the political difficulties of letting people realise how much tax they are paying.
Okay, that was helpful. Thank you very much.
I would like to touch on some of the areas that we have touched on already, including the Barnett formula. Professor Trench, you talk about
“a more rationally-constructed grant”,
which I assume means less than we are getting at the moment. [Laughter.] Mr Phillips, you say that
“such a system would be ‘fairer’ than the existing system, which involves a block grant that has no relation to current needs”.
Again, I assume that means that we would get less. Is that in fact the case?
I have not done a needs assessment, so I do not want to come to any firm conclusions on that. However, the most recent needs assessment—a good piece of work by Gerald Holtham in Wales—suggested that Scotland was receiving more funding than it would if it was getting the amount equivalent to what it would need relative to England. That is not saying the absolute amount—you could argue what the absolute amount should be. However, Gerald Holtham’s work suggested that the relative amount that Scotland receives is higher than it would need to be if compared to England. Scotland stands out as being relatively well off in terms of income per capita and yet having very high spending.
Surely, while needs form one part of the equation, the other part of the equation would be that a no vote was meant to make sure that we were not any worse off—there was a commitment there and there is the idea that we need compensation because we have given up sovereignty. Are those not factors as well?
On the politics side you are right. Fundamentally there is a question about what form of equalisation and fiscal redistribution should take place in a fiscal union. There are various ways of looking at that, including a needs basis; a revenue equalisation basis, which is equalising not on needs but on the revenues capacity side; or a contributions-based basis, in which the amount that is paid in affects the amount that is paid out. There is a range of options.
On the political issue about whether a commitment has been made to the Scottish people to keep the Barnett formula, it seems to be the position of the Scottish Government and of the UK Government at the moment that the Barnett formula is there to stay. That is a political decision. Whether that decision is fair in the round, I leave to the politicians to decide.
The levels of spending in Scotland compared with other parts of the UK look relatively high, according to the most recent assessment—not my assessment, but Gerald Holtham’s. On that basis, moving towards a needs-based assessment would seem fairer. I agree that there is a question about how that fits into the political promises or pledges that have been made to the Scottish people.
I have already talked about the problems that are implicit in the present Barnett arrangements. I do not want to repeat that, but it is all relevant in this context.
We sketched out a mechanism for a rationally constructed grant in the paper “Funding devo more: Fiscal options for strengthening the union”, which I am sure you have seen before and which we published in 2013. A much more effective way of trying to manage devolution and finance generally would be to combine the elements of grant and fiscal devolution.
In addition, it is worth bearing in mind that, under the Smith proposals, taxes that flow directly to the Scottish exchequer—both fully devolved taxes and the assigned share of VAT—would equate to something between 50 and 60 per cent of the Scottish budget, depending on how we cut the figures for Scottish devolved spending. The Barnett formula accounts for only about half of that or even a little bit less. There is and can be no objection to higher levels of public spending in Scotland if Scottish taxpayers pay for it. The question is how much of any higher level of spending should be covered by taxpayers from outwith Scotland.
Is that not the promise that was made in the vow?
The promise that was made in the so-called vow was to maintain the Barnett formula but it is also subject to the Smith proposals for fiscal devolution.
Was the implication not that Scotland would not lose out?
That is the implication. The problem with that is that it involves a set of assertions or assumptions about what will happen in the future, which, by definition, we cannot guarantee, not least because of the behaviour and effects of devolved taxation. That, in essence, is what we grapple with when we talk about no detriment. The inevitable problem is that, if devolved tax decisions work, they should shape the tax environment and the public revenues. The question is where the responsibility for that lies.
If the Scottish population does not grow less quickly than that in the rest of the United Kingdom, the Barnett formula implies that a squeeze will take place. Therefore, in the longer term, the Barnett formula would not necessarily be beneficial to Scotland. A needs-based formula might deliver more if the squeeze starts to come into effect.
That is certainly true. The Holtham figure is that Scotland’s relative needs come in at about 104 or 105 per cent of England’s. The Barnett element of block grant funding for Scotland is something in the order of 118 per cent of UK average, so there is a substantial benefit to Scotland from it at the moment. That is, I would think, a logical argument for the Scottish National Party to resist any call for fiscal devolution whatever, and I note that there has been a remarkable absence of that demand.
The convergence factor will take a long time to get down to that level but, ultimately, it will do so because convergence will drive Scotland down to 100 per cent of English spending. That is logically where convergence would take you.
We have mentioned the concept of an independent body to arbitrate between the two Governments in a range of areas if there are disputes. The example of Australia has been given, which is fine, but is the reason that it works there not the fact that Australia has a proper federal system and, I think, a written constitution? We do not have those, and to have an independent body would be for the Treasury and Westminster to give up ultimate control, which is surely impossible for them.
I do not see why.
Devolution fundamentally means that they are graciously granting us a little bit of freedom.
I do not believe that that is how devolution works and I am not sure that that way of thinking about it is a particularly useful way of trying to understand the relationships that are emerging in these islands.
You think that the relationship will change.
Yes. For the UK to work as a devolved union, it needs to change.
To set the budget, the whole framework would need to change. The bigger part—that is, the UK—would set its budget first. The smaller part—that is, Scotland, Wales and Northern Ireland—would then tweak round the edges. However, at the moment, the situation is the exact opposite. I would hope that Westminster would be willing to change that. Is there any likelihood that it will do so?
Yes.
Good.
I talked about an independent body to deal with disputes, but I am rather doubtful that the idea of it arbitrating disputes is necessarily where that would need to go to. It is more likely to be a form of mediation than arbitration. I ought to add that that is not part of the Australian architecture. In a sense, it is a measure of the success of how the Commonwealth Grants Commission does its job that there has never been a need for such a dispute resolution mechanism. If there is a political dispute, the matter would go to a body called the Council of Australian Governments, which is the Australian Government’s equivalent of the UK’s joint ministerial committee. If there is a legal dispute, that would go to the courts.
In effect, the commission makes a decision and all sides, so far, have accepted its decision.
Yes, because it is able and in a position to do its job properly.
You think that we could get to that position here.
That has to be the aim.
I agree that that has to be the aim. The politics is rather different here than it is in Australia. We would have to be mindful of the fact that one of the components might not want to be part of the overall union. The politics of that could have an impact on the operation of these systems. I am not sure that there is a way round that, but everyone must be mindful of the political situation here, where the union is not as stable as it is in Australia.
We have talked today and to previous witnesses about the idea of having a framework that is, on the one hand, mechanical—the word “mechanical” has been used and, I think, the word “automaticity” comes into one of the witnesses’ submissions—whereby something is set up, it works its way through and it is left to get on with it. On the other hand, we could have a framework that is subject to review and negotiation. Some people have suggested that we should negotiate every year, but that would take away virtually all the mechanical side. I think that a period of five years for Australia and perhaps seven years for Canada has been mentioned. Is there a right answer? I would, in part, want to set up a system that would work for 10 or 15 years before we had to go back and look at it.
If you could put in place something that worked for 10 or even 15 years that would have many attractions, but I suspect that you would find yourself needing to do quite a lot of maintenance along the way.
You would probably be better off going in for a regular service, as it were—and knowing that in advance—rather than having to take that metaphorical car to the garage rather more often because this bit or that bit has gone wrong, the windscreen wipers have fallen off, or you suddenly need to replace the wheels or something.
The two approaches should not be seen as being in conflict. The idea of a mechanical framework, which works on a year-to-year basis, has attractions. Indeed, the methods of indexing along the Holtham lines, where you deal with the tax changes directly and the interactions of the Barnett formula fall out directly, have their attractions, because they avoid the year-to-year negotiations that would be needed.
Even if the framework works on a year-to-year basis, it would not all be automatic, particularly if the no-detriment principle is invoked for the knock-on effects of policies—that is, the compensating transfers among the different Governments. However, in as far as it is possible to do it year to year, there are benefits to the framework being mechanical. However, I agree that it would have to be reviewed. I am not sure whether there would be an optimum duration.
I am not sure whether it would make political sense to have the reviews taking place in non-election years. If the reviews were to happen then that would perhaps take the politics out of it, but that may just be hoping for too much.
10:45
Let us discuss further the process around the budget. Professor Trench, you were very confident that the Treasury’s behaviours will change and develop. That is slightly out of step with some of the other evidence that we have received, which has indicated that the Treasury is unlikely to alter the way in which UK budget setting takes place. What leads you to assume that that change is likely to take place? The Chancellor has had what I refer to as a rabbit-out-of-the-hat approach for quite some time, and that has become behaviourally entrenched in UK Treasury budgets.
We have to distinguish between the various elements of what the Treasury does and how it does them. The central element, unquestionably, is the annual UK budget. I would agree with people who say that the Treasury will not change how it does that; it is certainly not a likely change.
In other respects, the Treasury has to accept that aspects of what it does must change, because that is the logic of the situation. The politics is such that there is increasing awareness of that and there is a will to address those questions.
You have both talked about whether we need more information about how the block grant operates. A great deal of mystery surrounds how it operates and more transparency is being called for. Is there any reason why that information could not be given in advance of the July budget, as opposed to waiting until the enaction of either SRIT or the Smith powers?
There will be practical difficulties in providing more information about how Barnett works, not least because we need to specify it much more clearly than I suspect we have done. I suspect that David Phillips and I both have ideas about what would be needed. We would need to ensure that the Treasury understood that and published it.
Those are practical difficulties. In relation to this year’s budget, there is no particular reason why historic data could not be provided.
The Treasury could publish some spreadsheets that it sent me. They exist on its hard drive and they are relatively easy to understand. I am not in a position to publish them, because I got them directly from the Treasury, but they can be published. I cannot see a reason why, subject to people agreeing it, they should not be published on the day of the budget, along with the other documentation.
The IFS submission says that devolution of social security
“may result in a system better suited to Scotland’s particular needs and preferences.”
We took evidence last week from Professor Michael Keating, who spoke about the tax and welfare mix in Québec. He said that Québec
“has been able to resist the tendency to greater social and economic inequality in the rest of Canada.”—[Official Report, Finance Committee, 20 May 2015; c 29.]
Is there the opportunity for the same approach to be taken in the Scottish context following the devolution of further powers?
I would raise two points in response to that. If there was substantial further devolution of welfare powers, or is there was devolution of national insurance contributions alongside income tax, that would give much greater scope for policy variation. As I said in the update to my submission, it is not just about increasing or decreasing benefit rates; it is about completely restructuring the system.
In principle, there is scope to do things quite differently. You could have higher taxes and higher tax credits and benefits. You could reshape the system to redistribute more or less over life cycles. There is a range of things that you could do. However, you will be constrained by the extent to which people’s behaviour responds. People in Scotland may change the amount that they work and change the amount that they report to the tax authorities. You will also be constrained by the migration response, or people moving between Scotland and the rest of the UK. If there were substantially higher taxes at the top, you might see people leave Scotland to live in the north of England or even move down to London, if they are in the financial services industry. Although there would be scope for policy variation and difference, the extent to which it could actually achieve drastic change would be affected by the mobility of people and behavioural response.
I am not sure about the specifics in Canada. My impression is that Québec’s economy has done somewhat less well over the past 30 or 40 years than that of the rest of Canada. I think that there has been some movement of top earners and top businesses from Montreal to Toronto, although I am not sure to what extent that has happened because of differences in tax regimes as opposed to the potential for secession. Differences in tax rates, when they are large—and differences in benefits, when they are large—can have behavioural responses.
Later in your submission, you look at the SNP Government preference for increased carers allowance and other social security measures, and say:
“None of these could be delivered by the Scottish Government under plans set out in draft legislation to implement the Smith Commission”.
That does not seem to tally with what you have just said about the use of tax powers and so on.
I do not think that that was in my formal submission. Was it on the website, or was it with something else that I sent over?
I do not have the page marked up; I pulled it out to put in my questions.
Oh yes—sorry, I can see it now.
Under the Smith commission proposals, the Scottish Government will have the power to make changes to the systems in the devolved areas of competency. Actually, it could make the changes in some of the areas noted in my submission and in others it could not. There may be a small mistake in that paper. You cannot change the work allowances on universal credit. On halting the move from disability living allowance to personal independence payments, most of that will have been done by the time that the powers under Smith have been transferred. PIPs will be among the things that are devolved to the Scottish Government under the Smith commission’s powers. The Government will be able to undo those changes if it wants to at that stage, but it will not be able to halt the transition as it actually takes place.
Forgive me—I cannot remember whether carers allowance is one of the areas that is being devolved.
It is being devolved.
It is. Unfortunately I made a mistake in this document; I admit that. The Scottish Government cannot do universal credit, but it can do DLA and PIPs once the powers for them are fully devolved. What it will not be able to do more generally is increase other benefits. As we saw in the draft clauses, the top-up powers, which it was initially thought would allow the Scottish Government to top up any benefit, seem to be restricted to top-ups for discretionary payments. It will also be unable to make any further structural changes to the benefits system.
In the submission, I was saying that further powers for devolution would allow much more in the way of changes to the benefits system—there were mistakes in the submission, unfortunately—but there will be some powers for that under the Smith commission.
We are looking at the fiscal framework that is going to be established, and we have spoken about the no-detriment principle. From your perspectives, how much flexibility is there to allow for meaningful policy divergence? A fiscal framework will be developed, but within it there will be multiple no-detriment elements. How constraining do you see them being of the flexibility that is needed for the real policy divergence required for a different economic approach to be taken?
It depends on how seriously the no-detriment principles are taken in terms of their compensation elements. For example, the Treasury has said that, if the Scottish Government was to decide to have a more generous system of support for unemployment by topping up unemployment benefit and giving higher rates of benefit to unemployed people, people might stay on unemployment benefit for longer because it gives them more income. That would mean that the Scottish Government would be paying more, and the UK Government would be paying more because the standard element of unemployment benefit and universal credit will remain part of the UK Government.
That was one of the reasons why the UK Government was concerned about top-up payments under the initial interpretation of the Smith commission powers for top-ups. The UK Government therefore talked about the need for compensation packages, in which the Scottish Government would have to pay the UK Government to cover the impact. Similarly, there was discussion of whether the Scottish version of the work programme would be less effective than the UK Government’s version and, if it was, whether the Scottish Government have to pay the UK Government for the additional benefit payments.
Surely there is a flipside to that. If it was successful in Scotland and there were compensatory savings to the Department for Work and Pensions, would they be retained at the UK level?
No. The idea is that the compensation payments would work in both directions.
There are two ways of thinking about all this. If compensation payments have to be made every single time, will there be knock-on effects? There are knock-on effects whenever there are policy differences, and calculating them is fraught with difficulty. Different assumptions will be made and different modelling will be used by the different sides. It will therefore be hard to get agreement. If the idea is taken seriously and compensation payments need to be made in either direction every time there is an effect, it will become incredibly difficult very quickly to calculate them and that could discourage policy differentiation.
On the other hand, it could be accepted that, under the no-detriment principles, we could not apply such compensation in every single case. It would have to be done when there was a particularly egregious example of a knock-on effect—how agreement would be reached on such examples is another question—and it would have to be the exception rather than the rule. Compensation in such cases would not have such a big influence so there could be more scope for policy differentiation.
There will always be knock-on effects, but that is part of policy differentiation. Other countries, such as the United States, do not have no-detriment principles in their system and there is great policy differentiation in income tax rates, corporation tax rates, sales tax rates and even benefit rates across states. Although it is responsible to impose some kind of limit on how far differentiation can go, or at least to give the costs for when differentiation becomes too great, the no-detriment principles might or might not have an impact, but that depends on whether they are invoked every time or just used for very serious cases.
I agree with most of what Mr Phillips has said about no detriment and how it should work.
On top-up powers, as one of those who were responsible for first proposing the approach of allowing the Scottish Government to be able to top-up and supplement welfare levels or introduce new benefits, I regret that the command paper and draft clauses do not go as far as they ought in delivering that principle. That is an important element of how a post-Smith world should work and I regret that it is not there. I understand that there are some serious practical difficulties from the UK Government’s point of view, but I do not think that they have been adequately addressed.
I will just say something about Québec. I think that Professor Keating may slightly exaggerate the extent to which Québec has pursued a different social policy to that of other parts of Canada. It may be a matter of nuance and I am not quite sure about what period he is talking about, but an important thing to bear in mind in the Québec context is language, not because it reinforces Québec’s difference but because it segments the markets. That means that mobility of labour between Québec and the rest of Canada is a very different business compared with what it would be between Scotland and the rest of the United Kingdom. It is much harder for people to move from Québec to other parts of Canada, because they need to have good English in order to do so as well as a desire. Of course, that language hurdle does not exist in the Scottish context.
11:00
I have a final question on intergovernmental relations and interactions. The committee has received evidence that there needs to be a much more formalised process of intergovernmental relations than that which exists at present. Do you have any views on how that should operate?
I have plenty of views. I am never quite sure what “formalised” means, and I am very sceptical about things like statutory underpinnings because I am not sure that they do any good or how well they work. They seldom exist and even more rarely serve any useful purpose in other parts of the world.
I think that we need to be more systematic about this issue and need much better information about it. The best people to get information out of the Executives that conduct most of these things are legislators. I would therefore look particularly at members of the Scottish Parliament to be trying to get more information from the Scottish ministers about what is going on in the same way as I would hope members of Parliament at Westminster would try to put greater pressure on UK Government ministers to tell us what is going on. At the moment, all that we get is the annual communiqué or report from the joint ministerial committee, which is pretty terse and not very helpful.
Mr Phillips, do you have a view on that?
I have nothing to add on that point. I agree with most of what Professor Trench said.
Mr Phillips, you have explained in your paper and verbally here today quite a lot about how a block grant adjustment mechanism might work in relation to devolved taxes. You touch on this in your paper, but can you say a little about how a block grant adjustment mechanism might work in relation to any welfare elements that are devolved?
In principle, the two adjustments could work in a comparable way, if not exactly the same way. The simplest way might be to index the block grant addition to what happens to spending on comparable benefits in the rest of the UK. That would be the analogue of what might happen on the tax side but, for a couple of reasons, it could be more complicated.
First, there is the fact that structural changes to the benefit system seem to happen more frequently than they do to the tax system, or at least to the income tax system. In principle, you would want the block grant additions to reflect changes in the UK system. As Alan Trench said, it is fair enough if Scotland wants to pay for higher welfare from its own tax revenues, but should funding for higher welfare come from the block grant?
In principle, you would want the system to reflect any big reforms to the UK system. For instance, if the UK were to delay PIP and cut the amount spent on disability benefits, you would want to give the Scottish Government less money; if it wanted to continue to pay more, it would have to fund that itself. However, going along that line is likely to lead to more political difficulties.
Secondly, given the more rapid ageing of the Scottish population, if you had full devolution of welfare, including pensions, would you want taken into account the known differences in demographics that will affect Scotland? In that case, the index might be the population age over the state pension age and some adjustments would be made.
In principle, you can use the same methods as you do for tax, but because of the more frequent and large structural changes, such as the recent introduction of universal credit and PIPs, and because of the known demographic factors, there may be more scope for debate about whether you should take those additional things into account. There could be more difficulties involved. Does that make sense?
I know what you are driving at. Do you have any additional points to make on that, Professor Trench?
When we did the welfare devolution work, that was one of the problems that was lurking in our minds. It is something that we had some ideas about, but nothing very clear. I am afraid that it is one of those cases where the Scotland and United Kingdom command paper badly flunks the test: it does not answer the problem.
The solution will be something like this: you will identify appropriate proxies for a level of funding in relation to the devolved welfare benefits and there will be an element of funding that will need to be in the annual managed expenditure element of the budget, rather than the departmental expenditure limit in the budget, which would flow into and be equally functional with the block grant, so that it could be spent as the Scottish Government saw fit. Your proxy might be as crude as to say that in 2016 or 2018—or whatever year we are talking about—Scotland accounted for X percentage of the UK recipients of benefit Y, and we will assume that Scotland continues to have X proportion of people who will be entitled to that benefit, subject to changes in the demography of Scotland. An amount would be added to the block grant on that basis.
I was disappointed that there was no serious analysis of those problems in the command paper—it skated over such difficulties—because that is what a white paper should be doing.
Okay. Professor Trench, you talked earlier about an independent body being required for the purposes of trying to predict what the block grant ought to be, to carry out a post-event audit and also for resolving disputes.
Not quite. We need two bodies—a second one to deal with disputes and disagreements.
That was going to be my question. There are three functions. Is it your view that the first two functions would be operated by one body and the dispute resolution would come under another?
Yes, those two things are entirely different. We are talking about a specialist expert advisory body to carry out a technical job and then a mechanism to resolve disagreements and disputes, which I suspect is more in the nature of mediation and arbitration and needs to be able to do an element of both. You certainly could not have your UK finance commission—to give it a notional name—as the body that might well be the subject of the dispute also involved in resolving the dispute. That would offend every principle of justice. You would need a different mechanism, as we need for addressing intergovernmental disputes more generally. That is what we are driving at.
That is helpful. For the first body, you said that we could learn lessons from the Commonwealth Grants Commission and gave us some helpful evidence on that. In terms of the dispute resolution body, is there an equivalent body elsewhere—you said that there is not one in Australia—that we can draw lessons from?
There is not, but the reason for that is interesting and reflects on the difference between other places and the UK and the nature of devolved government in the UK. This comes back to a point that Mr McDonald raised earlier, which is the unwritten nature of the constitution.
Federal systems all have written constitutions, which tend to be rigid formal frameworks that are difficult to amend—if they can be amended at all, because in many cases they cannot be. When there is a sufficiently entrenched disagreement that cannot be resolved by political means in the proverbial smoke-filled rooms—I suppose that they are seldom smoke filled now—the matter will end up in the hands of the court, and the court will play that role.
One example of that can be found in Germany. Periodically, the principles—not the detail—of the German equalisation mechanism get out of kilter. On a roughly 12 to 15-year cycle, the matter is referred to the Federal Constitutional Court by the Länder that pay the most into the system. The court will say that the system has ceased to apply its principles properly and it needs to be reconstructed, and the states go away and come to a political deal, having been told to by the court. That has now happened three times, and I am fairly confident that it will, in due course, happen again.
I would not like us to use the courts in that way. I think that it reflects a failure of political governance and it puts courts in a difficult position, because these are not issues that British courts are well equipped to resolve. They do not have the expertise and they are not accustomed to having that role—they might become more accustomed to having that role, but they certainly would not have the right level of financial and technical expertise.
I think that there should be some alternative framework that puts an impartial body in place. When I say a body, I am thinking about a group of perhaps three individuals.
Professor Trench, you said that your view was that there should be some sort of ceiling below which the UK Government would indemnify borrowing and that, although the Scottish Government could go higher, anything above that ceiling would not be indemnified. If we are talking just about the Smith powers—let us use those for the sake of argument—do you have any sense of what those limits or powers would be in monetary terms?
I said that my answer was not a particularly well researched one, so I will have to duck that question.
Does the IFS have a view on the level of borrowing that might be appropriate under the Smith powers?
We have not looked into the numerical issue. All that I can say is that, on the current site, there would need to be higher borrowing—at the moment, it is £500 million. There would need to be a lot more flexibility about how is that is used; that is for sure.
How much higher should the figure be? That would depend on a quantitative analysis of the kind of risk that the Scottish Government would be bearing under the package of Smith powers, and that would need to be reassessed if any additional powers were devolved.
The aim would be to ensure that the borrowing powers give enough flexibility and room for manoeuvre to the Scottish Government, given the amount of risk that the Scottish Government would be bearing. That is an exercise that involves not just picking a number from thin air but looking at the risks that the Scottish Government would face.
In terms of principle, do you take the view that the limits should be comfortably higher than you would anticipate being needed? Presumably, the last thing that we want would be for a limit to be set and then for it to be found that the limit was off and that, in fairly short order, it needed to be extended. Do you think that we should go higher than we think we are likely to need or should we go for a mid-level figure?
I have not really had a chance to think about that in depth yet.
I would be inclined to think that there is every reason for the limit to be a generous one and, therefore, to be higher than we would expect the Scottish Government to need.
The approach would necessarily have to be flexible, because this is something that will evolve. As we have been discussing in the context of no detriment, the tax situation can change after a few years. We have a high deficit at the moment, essentially because we have maintained a constant or a somewhat increased level of public spending when tax receipts have dropped significantly. One must deal with the fact that, if one is devolving tax powers, any missing tax receipts will be missing from the Scottish exchequer, not the UK Exchequer, and the borrowing powers must be able to cope with that adjustment.
Professor Trench, I want to ask you about the issue of Barnett convergence. We are debating the introduction of a needs-based formula. That means that, eventually, the issue that we are discussing might not be to the detriment of Scotland because, ultimately, Barnett convergence will be complete and there will be parity of spend. You said that that might take quite a long time. Do you have any details on how long that period might be?
My slightly glib answer might be, “Ask me again in July once we have seen the UK budget.”
That seems a short amount of time to get more information on what it would be.
11:15
What drives Barnett convergence is nominal increases in public spending. Those increases were both nominal and real, and they were huge during the noughties. That is why it is so remarkable that convergence did not happen during that time—if ever there were circumstances when convergence should have been happening, it was then.
I suspect that we are unlikely to see a similar increase in public spending for a substantial period, or a Government that is able to engage in such an increase in public spending. I would be fairly cautious about saying, “It’ll happen in five years anyway,” because I am fairly sure that it will not. It will happen over time, assuming that Scotland remains as attractive a place to live as it seems to be, so that therefore the population stops declining or even grows.
I have no predictions for when convergence will take place. It is unusual for me to be in agreement with him, but the work that Jim Cuthbert did in his submission, showing the interactions of expansion and growth, population growth and the lags between how often the population is updated, is a good piece of work. It shows how those three things interact and what that implies for the point at which they converge. Is it 100 per cent, 105 per cent or 110 per cent?
That is a good exercise. It does not give us the answer because it does not state what spending or population growth will be in the next few years, but we could probably run some scenarios using forecasts for population growth in Scotland and England, and different projections—such as that by the OBR on what we will be spending over the next 50 years—to see what sort of convergence we are looking at.
Introducing a needs-based formula, or whatever it is, as part of the process would presumably not be to the financial benefit of Scotland in that timescale.
Looking at the next five or 10 years, it is unlikely that convergence would be down to a level below 105 per cent, which I think is what the Holtham commission said the needs-based level would be. We have not made a needs-based assessment at the IFS, and it is not clear what the needs-based level would be at the moment. I agree that in the short to medium term it is difficult to see a needs-based assessment that would lead to higher funding than Scotland currently receives.
Finally, given the current absence of a group of individuals—presumably appointed by the Scottish and UK Governments—to help mediate over disputes between Governments, the point that you made, Professor Trench, about scrutiny of how the Joint Exchequer Committee does or does not meet and work, is important. This committee, and presumably also select committees in Westminster, must scrutinise that closely. Presumably there should be standing items for committees both here and in Westminster.
Indeed. Part of the issue at Westminster is that responsibility for these matters is fragmented between a number of departmental select committees, in particular the Treasury Committee and the Scottish Affairs Committee. Ensuring that the issue has a high enough profile, particularly for the Treasury Committee, is something of a challenge. It has not shown much interest in devolution issues until very recently—towards the fag end of the last Parliament it started to do some work on that.
Presumably it would also be appropriate for the Scottish Affairs Committee to consider the issue.
One would expect so. Particularly given its likely composition in the new Parliament, one would expect it to be particularly vigilant.
Thank you.
Thank you, all; that concludes questions from the committee. I have just one point: you talked about Scotland’s population declining by 0.3 per cent a year in the 1990s, and by 0.1 per cent now, but Scotland’s population is at record level. It has grown in recent years.
The decline went into reverse for a couple of years, as I recall around 2007. It is a relative decline rather than an absolute decline.
Thank you very much for answering our questions so directly once again.
11:20Meeting suspended.
11:30 On resuming—