Our third item of business is evidence on the Scottish rate of income tax from Professor David Bell. I welcome Professor Bell, our former adviser, to the meeting and invite him to make a short opening statement.
This meeting and the associated papers are important because they place on record the simple point that what you think that you may get if you increase Scotland’s income tax by, say, 5 per cent is not necessarily what you will actually get. Multiplying the average additional tax liability by the number of Scottish taxpayers gives an arithmetic estimate of the extra revenue but the outcome may be quite different, because that simple calculation does not take account of how taxpayers’ behaviour may change if tax rates change. Changes in taxpayers’ behaviour are difficult to predict, and it is very difficult to establish what we call the counterfactual—what would have happened had rates not changed.
One finding on which there is agreement is that increases in taxes lead to reductions in taxable income. That can take a number of forms: taxpayers may reduce their hours, drop out of the labour market, retire or even emigrate. All those responses would lead to lower tax receipts from other taxes as well as from income tax, which might have implications for the second no-detriment principle. Alternatively, taxpayers may reduce their effort at their workplace or negotiate higher wages to offset the higher taxes, which would lead to lower company profits. Finally, they may seek to avoid tax by reclassifying their income as, for example, profits, which would lead to a transfer of tax receipts from Scotland to the rest of the UK.
The group that is most likely to change its behaviour is the very highly paid. Scotland gets about 20 per cent of its total income tax revenues from the top 1 per cent of earners, which means that it faces considerable revenue risk if that group alters its behaviour. To me, that suggests that the Scottish Government must proceed cautiously—ca cannie—when considering significant changes to the structure of income tax in Scotland. It is also worth bearing in mind that the perception of Scotland as a high-tax jurisdiction might affect the behaviour of not only Scottish taxpayers but skilled migrants who might come to Scotland and have the potential to boost the Scottish economy.
Thank you. You used the words “significant” and “perception”. What is the definition of something that is significant? For example, the committee undertook a two-day study visit to the Basque Country last week, and one of the first things that I asked the minister for finance about was the impact of taxes being a lot higher in the Basque Country. He looked at us in bewilderment and told us that there has been no discernible impact. The Basque Country’s GDP is 22 per cent per capita above the Spanish average, its productivity is 28.5 per cent above the Spanish average and its human development index is higher than those of Norway, Sweden and Finland.
Incidentally, it is only 35 minutes by plane to Madrid from Bilbao, which is even closer than London is to us. The view of the people whom we spoke to is that the higher taxation levels had not had any adverse effect. They said that it “was not an issue”. The Basque Country more or less raises all its taxes, and it pays 6.24 per cent of all Spain’s taxes. It has much higher devolution than we have or are likely to have in the foreseeable future.
Labour elasticity is covered substantially in your paper. Where does that become significant? One could argue that, for the super-rich, even a marginal increase could deter someone. Ultimately, if we are looking to increase revenue, where does that occur on the x-axis and y-axis in the UK context?
We have quite a strange tax structure: we pay 20 per cent on income between about £10,000 and £42,000, and then pay 40 per cent on income up to £150,000. The big action is more likely to be down towards the lower end of the income tax schedule. I think that we raise about £700 million out of £11 billion from the additional rate. Let us say that we want to raise a lot of money. If we increase the additional rate a lot but there is no behavioural response, we will not get a huge amount of extra revenue. If we move the additional rate and have it kick in at £120,000, that will make a pretty significant difference.
We have to think about both bands and rates. You made a good point about perception. People see rates as the key issue, but bands are as important. What we see in the Scandinavian case is what used to be the case in the UK, where the bandwidths, if we can describe them as that, were smaller, and people moved up steps in a smoother fashion than they do currently. It is at the lower incomes where we would have a big impact on tax revenues. Of course, potentially that has its political costs.
It is good that in your paper you included national insurance, so that people can look at the real rate of income tax. People often forget that when income tax goes from 20 to 40 per cent, national insurance goes from 12 to 2 per cent, so there is really an increase of 10 per cent rather than 20 per cent.
You have talked about migration, which is also covered in Professor Gavin McEwen’s excellent paper. You said:
“the flow of migrants into Scotland could be reduced if the higher tax rates in addition to the social and psychological costs deter immigrants.”
Professor McEwen said:
“In a 2014 paper, Anouk Bertier of SPICe”—
the Scottish Parliament information centre—
“surveyed the evidence for tax-induced migration in Switzerland.”
Obviously, Switzerland and Sweden have very high migrant populations, so they are not areas where you would expect there to be, for example, any issue in attracting migrants. Professor McEwen continues:
“While there is a federal income tax in Switzerland, it is ... relatively low rates and Cantons and Communities have considerable taxing powers including the right to tax income. She reports that two studies she consulted found evidence of high-income taxpayers taking level of taxation into account in choosing their place of residence while one found little such evidence.”
There seems to be quite a divergence of views on the issue.
Professor McEwen also talked about the diversity of taxation throughout the United States, with a plethora of corporate taxes, sales taxes and so on. He said:
“There is undoubtedly competition between authorities, and influence on migration as a result, but the diversity has been remarkably stable and has even tended to increase somewhat.”
If we look at pure economic factors, we would clearly expect migration to decrease if taxes went up, but what impact does the expenditure of those taxes actually have? For example, if I am thinking about moving the wife and weans to somewhere, I will be looking at not just a marginal tax rate but the schools, the public realm, the levels of crime, the relative cost of housing and so on. How much of an impact would these marginal rates have if, as you seem to be assuming, Scotland has higher tax rates and if folk are actually looking at Scotland as a place where they can have a higher quality of life?
12:00
That is a very fair point, and one that I tried to cover in my paper when I discussed the forces that cause people to migrate or not. Typically, it is the young who migrate, and they do so because they gain the benefits of migration over the rest of their lifetime. The better off and the higher earners also tend to migrate because the costs of moving mean less to them than to lower-income people.
Of course, the costs of moving are not just economic; there are psychological and social costs associated with moving from the place where your children are established in school, where your wife or husband has a settled social pattern with groups of friends and so on. As for giving all of that up just for a lower tax rate, I have to say that it is not obvious that that lower tax rate will make the big difference.
The evidence around migration is very mixed. I refer in my submission to a study involving a case in which tax rates were used—successfully—as a specific policy to attract migrants into a jurisdiction. That happened in Denmark, and the paper is by Kleven et al. Tax rates were lowered for a specific group of high earners, which resulted in an increase in the number of high-income-earning migrants into Denmark. Quite a good counterfactual was established with earners who earned slightly less than the threshold at which the lower tax rates kicked in; their numbers did not change, but there was a big influx of very high earners.
There are a number of reasons to be sceptical about applying the same argument to Scotland. That was a specific policy aimed at attracting migrants, and I do not think that, under the powers of the Scotland Bill as currently constituted, Scotland could offer a lower rate of tax to people migrating into Scotland while keeping the same tax rates for residents. Given my understanding that that would not be a potential policy and that it might not be applicable to Scotland—and given that, of course, the situation in Denmark is somewhat different from that in Scotland—it is not clear whether the lesson necessarily applies.
One final point that I would make is that, in general, the costs of migration have probably fallen in the past two or three decades, mainly because of the ease of international travel and the way in which the labour market has become much more international than simply being national or regional.
Migration is obviously a key issue. On page 3 of your submission, at the end of that long section 2, you state:
“During the period 1950-2000, Scotland experienced net emigration of around 900,000”.
That is obviously the legendary union dividend coming into play again. Down in London, house prices are phenomenally higher, so younger people who might want to go there have to try somehow to get on the housing ladder. My understanding is that the schools are not exactly top notch for most people and that everything is more expensive.
Assuming that there are higher taxes in Scotland, if we can deliver a reasonable quality of life, a higher level of employment and more equality—the things that we all want—would that not deter emigration on tax grounds? The Basque Country is high on the index of equality. It is much better than anywhere else in Spain and above the European Union average. Incidentally, 29 per cent of the Basque Country’s population are immigrants, and most of them are from other parts of Spain rather than from developing countries. Surely, the issue is about what we do with the resources that we have, all else being equal.
At the moment, the cost of housing in London probably deters potential emigrants in, let us say, the financial sector in Scotland from going down south. On the other hand, if someone is established in the London housing market, a higher tax rate in Scotland might be a negative factor in their consideration of moving up here. People could adopt different strategies around that. That would not necessarily overcome the lack of migrants going down, but I guess that, although it is probably true that for most people the costs of emigration exceed the benefits, we have to be careful about losing even a small number of high earners. We can take the view that that is that but, if the top 1 per cent pay 20 per cent of total revenues, we have to be careful.
Let us look at the issue from another point of view. Let us say, for example, that whoever won the election next year decided that they were really worried about the effect of higher taxes and about losing some of the higher earners, so they reduced taxation. What would be the benefit of that to Scotland? Obviously, we would lose tax revenue, which would have an impact on our public services. What would be the economic impact? In other words, is there an equal relationship? Are people as likely to flow into Scotland if we reduce taxes as they are to leave Scotland if we significantly raise taxes? Has any work been done on that?
No, not that I know of. There would be less of a fall in revenue than we would expect if we just arithmetically worked out the cost—
Of course—just as we would never get as much the other way.
Yes.
We would lose less income tax revenue than might be expected and we might gain a bit of VAT revenue. Of course, if the value of national insurance, the other part of VAT and most other taxes was increased, that money would just go to HMRC. It is important to think about where people are going to come from. If they come from outside the UK, that is of benefit to Scotland and to the UK taxpayer. If they come from other parts of the UK, their increased VAT spend in Scotland will be mirrored by the reduction in VAT south of the border, so HMRC will be no better or worse off as a result.
Indeed, but let us say, for example, that we reduced the higher rate by 5p or something like that. The suggestion seems to be that, if we put it up by 5p, that would have a detrimental impact on the number of people who have a domicile in both places classing themselves as Scottish taxpayers. If we were to reduce the rate by 5p, would the amount that we gained offset the amount that we lost? The argument seems to be that we would not gain much by putting the rate up by 5p. Would we gain a lot by putting it down by—taking a random figure—5p? It seems to be considered that that figure would have an impact, whereas it is perceived that a 2 or 3 per cent reduction would not have quite the same impact.
I refer to Alan Manning’s paper and his view on the increase from 45p to 50p that was proposed by the Labour Party at the most recent election. The arithmetic calculation was that UK tax revenues would increase by £3.3 billion. However, Manning’s analysis, which was based on behavioural responses and on effects on other tax revenues, was that, depending on elasticities, the figure would range between a gain of £2.8 billion at the top, which was the most optimistic outcome, and a loss of £700 million, which was the most pessimistic outcome. We expect an increase of £3.3 billion if we just do the arithmetic but, in fact, there is a wide range of possible outcomes because of the uncertainty, and the figure would be between -£700 million and +£2.8 billion.
I have no reason to believe that the effect would not be symmetrical if the rate went in the other direction. You would probably lose about four fifths of the tax revenue that you expected to lose if you cut the Scottish rate from, say, 45p to 40p.
You are saying that, against the rough prediction of an increase of £400 million by 2018-19—the total for the higher rate would obviously be much lower, but I am just looking at this for arithmetical purposes; I realise that the figure would be about £70 million or £80 million for the higher rate—if we cut the SRIT by 5p, we would regain a fifth of the money that we lost. Let us say that we lost £100 million. We would gain £20 million of that back, so the net loss in overall revenues would be £80 million.
It might be, but it is difficult to say. At the UK level, migration has not really been an issue, but what has been an issue is people avoiding tax. Manning shows that there is a huge bunching of the earnings of self-employed directors at just below £150,000. They are taking the rest of the money as profits. There is no reason why there would not be that reaction in Scotland if you bumped rates up, and there might be the reverse reaction if you knocked rates down. That would affect the division of moneys between the rest of the UK, which would take its share of taxes on profits, and Scotland, which would gain only on the income tax.
I have a couple of points to finish off with, before I let colleagues in. It is clearly the case that, the lower someone is down the income spectrum because of a lack of mobility, the easier it is to raise or cut their taxes with a surety of what the financial impact will be. I imagine that very few basic rate taxpayers would leave the country if their tax went up. They might not be happy about it, but they could not do anything about it. The situation would probably be similar for people in the middle part of the income spectrum. That money would, in effect, stay within the Scottish economy. If it was lost, there would be very little beneficial impact in terms of increased revenue, as folk are not going to move to Scotland because the basic rate has fallen or risen—is that correct?
Certainly, migration would not be an issue. It would be a matter of basic work incentives and of how people responded in terms of traditional labour supply responses. Would they keep up their hours? Would some people drop out of the labour market? There are questions around whether people would take retirement at a different time and whether married women would drop out. In the past, they have been shown to be quite sensitive to tax rates.
There is then the question of what rates would be applied to the lowest earners—the people who are currently getting tax credits. You would have to do quite a complicated calculation to find out what rate they currently face and what rate they might face—depending on what happens with tax credits—if Scotland changed its tax rate. I do not think that migration would be an issue for that group, but you would have to be careful about changes in hours and possible withdrawals from the labour market.
12:15There was an article by Martin Wolf in the Financial Times the other day, which was about the problems of the US labour market. It has a much higher drop-out rate than the labour market in the UK. At the moment, the UK has a very high level of participation in the labour market, even though a lot of people are earning relatively low amounts, but the US has not registered the same increased participation as the UK post the recession. Scotland currently has its highest-ever level of employment because participation rates are high. You want to be sure that any changes that you make towards the bottom end of the spectrum still leave people an incentive to participate in the labour market.
In the last paragraph of your submission, you say:
“there is a strong case for moving cautiously when considering changes to the higher rates of income tax in Scotland, even though there may be no evidence that they have a detrimental effect on rates of economic growth in the long run.”
I find that quite interesting, because it is almost like saying that the economy will grow or not grow whether or not those people are here and paying tax. Is there an optimal level of taxation for high-rate taxpayers at which the economy can grow and we can optimise revenue without people fleeing the country?
There has been a huge amount of literature on the optimal rate of tax, which was started by a member of the Council of Economic Advisers, Jim Mirrlees. I think that he won the Nobel prize on the basis of that work. There is a huge amount of work on the issue. I thought that I might attempt to do some work on the issue for my submission, but I did not get round to it.
Yes, there is an optimal level of tax, and I can send the committee some recent estimates that are quite a bit higher than some of the rates that are being levied at the moment in the UK. Looking at the long-run performance of economies, in my last statement—the one to which you refer—I was thinking about Scandinavia. The Scandinavian economies have had very high rates of income tax for a long time but have not fallen significantly behind the rest of Europe in economic growth.
Stephen Boyd told us that tax revenues in Denmark are 12 or 13 per cent higher than here across the board and that that is borne by the population. However, surely, that is borne by the populations in Scandinavia because they have more left after tax—their incomes tend to be higher and they have more disposable income.
Yes, that is true, but they also have a high standard of public services and people buy into the notion that that is part of the deal. It is a somewhat different culture from one in which people do not particularly value public services and are much more concerned about their personal consumption. It is really a cultural issue, and it is difficult to prejudge how that system might work in the UK. Countries such as Denmark do not appear to have suffered significantly in the long run from having higher rates, but we might get a reaction in the short term.
Thank you. A number of members want to ask questions.
I was on the Scotland Bill Committee much earlier in this parliamentary session, when we discussed the implications of the full devolution of income tax. There was quite a discussion about the potential for having a predatory tax policy, about cutting the tax rates in Scotland to attract high-income earners and about whether there would need to be agreement between the Treasury, the UK Government and the Scottish Government that such a policy would not be pursued. Does the fact that there has been no such agreement, similar to the ones that sometimes exist between states in America, not indicate that the received wisdom is that migration is not going to be an issue and that, if it is an issue, it will be marginal? Alternatively, do you think that those matters will be covered by the no-detriment policy that is outlined in the Scotland Bill?
It is not clear to me how the no-detriment policy will cover all the ramifications of the effects on the other taxes that I mentioned: national insurance, VAT, corporation tax, excise duty and fuel duty.
I think that we are all guessing on the migration question. I would be more concerned about the long-term effect of not having people migrate to Scotland than about people emigrating from Scotland. That is my particular concern, and what matters is perception. It might be a general perception rather than one based on the detail of the additional rate being 47p rather than 45p, but if Scotland is perceived to be a high-tax jurisdiction and public benefits are not perceived to offset high taxes, there could be a negative effect. It is really difficult to predict.
I accept that. However, in the European context, if we raised the additional rate to 47p rather than 45p, we would still be behind France, Germany, Denmark, Sweden and others. Surely, therefore, the impact would be marginal when compared with, for example, European migration, which has been important for Scotland over the past few years.
Yes, but it is also true that, if you raised the rate to 47p, you would not raise a heck of a lot of additional revenue. That would be the downside, and that might well be the case. If you changed the amounts of revenue that are coming in through the Scottish rate of income tax, you would probably have to look a bit further down the income spectrum to make big hits.
The answer to most questions about what would happen if we did X on tax seems to be that we do not really know. Do you think that enough modelling is being done on the potential impacts of various tax changes? I suppose that I am asking whether any such modelling is being done. Could more modelling be done to inform Government policy and public and parliamentary debate?
That is a good question. I will say a little about that kind of modelling. The organisation that does most of that modelling is the Institute for Fiscal Studies, and we are doing some work with it just now. However, it talks mostly about short-run changes and focuses on the arithmetic effect of changes in taxes, which is reasonable. In the IFS’s analysis of the budget, whether 3 million people will be worse or better off or whatever, the behavioural effects will not really kick in in the short run. It is not the case that everybody will emigrate before next year; it will take longer for the effects to be seen.
We have done a little work on behavioural responses, but it has been around a general area—for example, what would happen if we applied the Danish rates to Scotland. That is making a big jump and an assumption, so any result that we arrived at certainly could be debated and we would have to qualify it by asking whether it was applicable in the Scottish context.
There is a good argument for doing more work so that such issues can be debated, but there is another issue, which harks back to what was said in the previous evidence session. If you focus on the very top earners, you will find that there is not much information on that group. There are about 11,000 additional rate taxpayers in Scotland, and surveys do not tend to pick them up.
Doing the arithmetic calculations of the impact of taxes is difficult enough; modelling a behavioural response adds a further layer of difficulty. I am not saying that it would not be a good idea to try that, but the results would have to be heavily qualified.
Thank you, Professor Bell.
On page 2 of your submission, you talk about “elasticity of labour supply”. I was struggling a little to get my head round that, especially the bit where you talk about what could happen if there was an increase in income tax. You say that that could cause
“real wages to fall by 1% and ... workers reduce their hours by 0.5%. Then the value of the elasticity of labour supply is 0.5”.
Could you explain that?
It is just a ratio. It is the ratio of the percentage change in the amount of labour supplied to the percentage change in the after-tax income of the individual. It is just a number; it does not have any dimension. Small numbers mean that there is not much of a response; large numbers mean high elasticity and a very marked response to changes in the tax rate.
Right, but are you saying that if you increase income tax, less work will always get done?
I quote a guy called Manski, who is a guru on economic policy in all kinds of areas. He says that there is no consistency on the estimates of elasticity that I have just described, but there is an agreement that if you increase tax rates, you will reduce taxable income.
I presume that that is overall, because some individuals, in order to maintain their income, will do more work.
Absolutely. The theory says that that is a perfectly plausible response to changes in tax rates. If you want to keep your real income up—if your income is what you really value—you will increase your hours of work. Some people do that but, on average, the findings have been that people reduce their hours. Of course, you can do that at what is called the intensive margin, which is hours of work, so you work fewer hours, or at the extensive margin, where you just drop out, so you do not supply any labour at all.
In a slightly different context, you say on page 10 of your submission:
“Cultural acceptance of tax rate differences is probably country specific.”
Are you arguing that there are quite a lot of differences between how people in one country—in Denmark, for example—and another country behave?
12:30
Absolutely. You can see pretty significant differences across the US, Canada, Switzerland and Denmark. This harks back to earlier remarks about what the people in those countries are getting in return for their higher tax rates; perhaps they just take the view that there is a compensating differential with regard to the public services that are being offered. I guess that what I am saying is that we do not know the detail of that, so it is pretty difficult to take lessons from any of those countries and apply them to Scotland. However, it is worth noting that having differences in tax rates would not make Scotland unusual in the set of developed countries in the world, because differences in income tax rates are relatively common.
Is the point that people’s behaviour is not in a sense logical, or that the logic is different in different places?
It is that there are differences in the costs and benefits in different places. What other taxes do people face? How do they respond to differences in indirect taxes? It does not necessarily prove that they are acting illogically, but they might respond to incentives in different ways or they might be facing other considerations such as differences in indirect taxes or the public services that are available to them.
I know that this is a separate area, but I was struck by the 80 per cent drop in the use of bags as a result of the 5p bag tax. After all, people are paying £40 or £50 for their supermarket shopping, so why should they not want to pay 5p for a bag? In the scheme of things, that might seem illogical, but sometimes people react quite strongly to a very small thing, which I suppose might be the risk here.
Yes. There are two modes of economic thought: the first is that people act logically all the time, and the second, which comes through the area of behavioural economics and might not have had the appreciation that it deserves, is that people are influenced by all kinds of things and might not act all the time in what might be their best interests. I should also say that the University of Stirling is particularly strong in behavioural economics.
I am not trying to suggest that people will behave logically or in their own best interests at every juncture; I am just saying that that adds to the complication of interpreting what happens in other countries. People might have a culture of working hard, and that might supersede the effect of changes in tax rates. Indeed, Manning found that, on the intensive margin, high earners do not change their hours of work much in response to changes in tax rates. In other words, if someone earns £500,000 a year and the tax rate goes up by 5 per cent, that person will still put the same amount of effort into their job, but they will also put more effort into finding and pursuing ways to avoid paying the higher tax. That might not seem logical, but the fairly uniform finding is that such people still maintain their work effort.
I am finding it increasingly difficult to predict what might happen, but your suggestion that we proceed cautiously is probably good advice. Can we pin that down? If I were to say, “Let’s increase the rate by 2p next April and see what happens,” would that be proceeding cautiously?
I guess so.
There is also the issue of signalling. For example, you might say, “Let’s increase it by 5 per cent over five years.” I think that that kind of behaviour would precipitate action around the tax avoidance behaviour, because I think that one of the things that happened in response to the announcement that the 50p rate would be taken down to 45p was that the announcement of the tax itself had an effect on behaviour.
So it is about the signal that is sent out. For example, we could say that we will raise the rate by 2p in April and leave it sitting there for five years before we do anything else, but that would be different from saying that we will increase the rate by 1 per cent every year for five years.
Absolutely.
Professor Bell, you mentioned people moving income into profit or dividend income in order to avoid paying income tax. You said that a change in the tax rate might encourage or discourage that behaviour. I presume that that would be true only if the tax that was then paid on that income was lower or the same as the rate that someone would pay if it was dividend income. How far do we need to go in order to encourage that kind of behaviour?
People in a position where that matters are probably high earners who can pay for tax advice. As I said earlier, if we look at information about self-employed directors in a thing called the survey of personal income, we see that their pay tends to be bunched just below the £150,000 additional rate cut-off point. I take it from that that people who earn that kind of money have taken advice and have opted to pay corporation tax on the remainder of their income, whatever it might be. If the two rates were the same, there would be no incentive to do that and we would not see the bunching happening.
How much does the difference have to be? People in that position are offsetting the cost of the tax adviser whom they have to pay to give them advice about what to do. As long as they are ahead on the game, they will do that—for example, if they are earning £500,000, it is quite possibly worth their while to do it. Therefore, the issue is both the difference in rate between corporation tax and the top rate of income tax and how much above the top band of £150,000 their income is. If they are on £151,000, it does not matter.
There could also be tax variation at the other end, because although the personal allowance is not going to be devolved, there is talk of the Scottish Government having the ability to set a zero rate. I am not saying one way or the other that that would happen, but that is another way in which tax behaviour can be manipulated, for want of a better term.
Yes, that is true and it raises an interesting point. Even if we do just the arithmetic calculations—the IFS-reaction-to-the-budget kind of calculations—it is really worth doing those, because if our concern is about poverty or inequality, say, changing the personal allowance might not necessarily have as big an effect as we might expect. That is partly because we usually look at issues such as poverty and inequality not on an individual basis but on a household basis, because it is households that are faced with the difficulty of putting food on the table or whatever. In that case what matters is whether there are two people in the household who would benefit from the increase in the personal allowance, or whether there is one person who is working and one person who is not working or a different combination—all those kinds of things are possible.
A very detailed calculation would need to be done to see whether that would benefit people when compared to, for example, reintroducing tax credits or compensating people for the loss of tax credits. Even without adding in behaviour, that kind of calculation is worth doing, because it gives the picture, as far as the household is concerned, on whether, by changing the personal allowance, we are moving people above the poverty line, which is 60 per cent of median household—as opposed to individual—income. I am not saying that it is one way or the other, or that it is not a good idea. I am sure that there are studies out there on how people will react to a change in the personal allowance, but there is always the problem of finding what the counterfactual is—what would have happened if people were not subject to an increase in the personal allowance.
We have spoken about the potential migratory impacts. In a population of just over 5 million, what qualifies as a significant behavioural impact? If the behaviour of only a small number of people is impacted, either positively or negatively, is that really a significant impact or, in thinking about the significance of the impact, do we have to take into account where those people are in terms of the amount of tax that they pay?
Where people are is clearly important. It is for HMRC to make clear or to be able to accurately determine where people are in relation to their tax liabilities. From the news this morning, it is not clear that HMRC has been all that effective, for example, in pursuing people who have apparently been evading tax.
The short-run impact of migration might not be all that much. If the additional rate contributes about £700 million to Scottish revenues, it might not seem to be necessarily all that negative to lose some top earners, but we have to be careful, because those people might also be entrepreneurs and the fact that they move might have impacts on other parts of the economy. One of my big concerns about the Scottish economy over the past 20 years has been about the loss of decision-making power at the top end within Scotland. That certainly used to be there. The number of FTSE 100 companies that are located in Scotland has fallen over the years. I am fairly concerned about that.
In the longer run, for migration, one of the main differences when it comes to why the Scottish population will grow more slowly than that of the rest of the UK in the next 20 or 30 years is that we are forecast to have a lower level of net in-migration. Over time, that builds up. The fact that around 900,000 people were lost in the 1970s and 1980s has had a massive effect on the Scottish population. However, I am not saying that a change in the tax rates will reverse what has been a trend towards in-migration in Scotland in the past 12 or so years and particularly since 2004.
Migration is something that we have to be concerned about, but I do not think that the taxes that people pay is by any means the major issue as regards what determines the overall flows of migrants. That is perhaps more to do with public services, quality of life and so on.
12:45
Sure. I guess the question that I was driving at was, if we were to take the view that using the tax levels that we levy would attract people to Scotland or, on the flipside, drive people out of Scotland, at what point are the effects of that significant? How many people would we need to take in to be able to say that the policy had had a significant effect, or vice versa? Is there a model that we can look at that would indicate that, if we attract or lose 10,000 people on the basis of the policy, that would have a significant impact?
I do not know that we have a model that does that. It would depend on what kind of people they were—what their average income was and, therefore, what kind of tax contribution they were making.
Clearly, if the Scottish economy is growing, migrants will be attracted into Scotland irrespective of tax rates, just because there will be employment opportunities in Scotland. There are plenty of people from around Europe who have already taken advantage of the employment opportunities that are available in Scotland and have moved in over the past decade or so.
Whether changes in tax rates would affect the overall pattern of migration is not entirely clear. We will not be able to do what the Danes have done and offer a specific rate to people coming in, but you have to be careful about losing, or not gaining, a group of high earners within the overall pattern of migration. There is movement, with significant gross migration flows each year, to Scotland and from Scotland. The net difference is around 10,000 to 15,000 a year. That is roughly where we are with migration currently.
It seems unlikely to me that differences in tax rates would make a huge difference to that overall net flow, but you have to be concerned about whether the tax base or tax revenues, whichever way you want to look at it, are disproportionately affected by migrant flow.
On the question of how long you would need to be able to judge the impact accurately, one of the criticisms that has been levelled in relation to the 50p rate is that, because it was applied for only one year, in effect, it was not possible to judge its long-term behavioural impact. You can only really forestall the first year for which a policy comes into effect, but it might not affect those people who move towards taking profits. How long a period do you need to assess a tax change or a tax policy before you can accurately judge its impact?
You would need at least three to five years, I would think, partly because you would not get the information on the overall revenues until about 18 months after the year end.
If we are talking about the self-employed, they complete their tax returns in January after the end of the tax year, then there is a bit of debate, so the accounts are not finished. Some of the debate around the 50p rate was being done in the dark, because that information was not available.
I think that you are right that it will take what I would call a medium-term perspective before we really know what is happening and what effect a change in taxes has had.
I have a quick question about the evidence from what has already happened. In the early 1980s, we were paying 33p in the pound and, because of the political philosophy and policies of the time, that was changed over a few years until we reached the 22p or 23p in the pound that we pay now as a minimum. What evidence does that give? Can you refer me to anything about that? In a sense, that is hard evidence of what happens when, for whatever reason, income tax is reduced, whether or not we approve of that. I know that taxes went all sorts of other ways and there were changes elsewhere. Since then, every Government has been too nervous to change that approach.
There was a considerable shift in tax policy in the 1980s, which is why looking at income tax on its own gives only a partial view of what is happening. High marginal tax rates were cut, but tax revenues did not fall all that much. There was an increase in indirect taxes—in VAT in particular—so the overall tax burden fell a bit, but not massively for the UK’s economy as a whole.
Since then, Governments have continued to rely more on indirect taxes than on direct taxes. That is partly because collecting indirect taxes is easier than collecting direct taxes is. By indirect taxes, I mean VAT, excise duty, fuel duty and a few others. That has a distributional impact, because most people buy goods that are subject to VAT, pay fuel duty and pay excise duty on cigarettes and so on. In terms of real income, the shift towards indirect taxes tends to favour the rich relative to the poor.
That shift did not just happen in the UK. It was partly driven by the fact that collecting indirect taxes was easier than collecting direct taxes. The difficulty for the Scottish Government is that, in the UK, income tax has become sacrosanct. The increases in direct taxes in the past couple of decades have not really come through income tax; they have come through national insurance, which Scotland does not have the power to vary.
Reviews have been done of what has happened. The IFS has done that kind of work and I am happy to give the reference for that.
Thank you—that is kind.
On the basis that, if tax rates increase, people work less, is there any evidence—again, we can go back to the 1980s and 1990s, when tax rates were reducing—that the converse is true? It is not true, is it?
It is difficult to assess whether things work in reverse. The evidence—for example, on married women’s participation—mostly comes from the 1980s, when there were big changes in income tax rates. The evidence that is out there, which is all subject to qualification, is that full-time males tend not to vary their labour input—their working time—very much in relation to tax rates; the same is true of single females. For married females, the position is a bit more doubtful. Of course, that might have changed. The rules that might have applied in the 1980s, or the observations that we have from then, might no longer apply because the nature of work has changed so much—for example, there is the ability to work at home, and people can put in their hours at home. I guess that I am saying that this is all quite difficult.
I have a couple of brief questions. You gave us the example of Denmark using tax rates to attract migrants—I think that the study was by Kleven et al. Do you have some idea of the magnitude of tax change that was used in that case?
I have forgotten. It was fairly considerable, but the response was also pretty considerable. I know that the elasticity was about 1.52, which is very high, but I have forgotten what the initial change was. I will send that to you.
Thank you. The issue of the 45p to 50p rate has been raised a few times. You quote someone else in your submission, but you are basically saying that it is not possible to measure how much extra tax would be collected. What we know for sure is that the static costing will be wrong—that will not happen in practice. Having seen that happen and having seen the rate changed once in each direction, surely we must have some idea of the behavioural response that we would expect if it were to be raised again.
I think that we should know that very soon. I was quite surprised that Manning’s paper, which was produced just before the election, did not have the numbers for what actually happened and how much additional-rate tax was collected, but that information must be available.
In response to a question from Mark McDonald, you talked about being slightly cautious about playing around with tax rates too much, too quickly. You said that migration flows could remain the same in each direction, so we could end up with net migration being exactly the same with a changed tax rate. However, your point was that the composition of those flows could change.
Let us say that we put tax rates up dramatically. Are you saying that, while the overall numbers of people might remain broadly similar, lots of high-rate taxpayers might leave and lots of people who were not high-rate taxpayers might come in, which would lead to an impact on the public finances?
The average age of migrants out of Scotland tends to be lower than the average age of in-migrants. The young leave Scotland and people who are on average a bit older—though not much—come back. In-migration is made up of quite a lot of young migrants coming to Scotland and also people retiring to Scotland. As you say, the proportion of that group that is responding to changes in tax rates is quite important, because that will affect the demand for public services and the revenues that the Scottish Government collects.
That is helpful.
13:00
I have one final point before we wind up. Paragraph 7 of Gavin McEwen’s paper refers to a study by McGuire and Rueben that found that
“how spending is correspondingly reduced matters. Thus tax cuts leading to reduced spending on services which impact on business, such as education, may have an offsetting effect on business location and economic growth.”
In paragraph 11, Gavin McEwen says that the general conclusion is that
“States wishing to foster entrepreneurship should focus on creating a stronger business climate and economic growth rather than on tax policies.”
That is saying that reducing taxes can have a detrimental effect but that taxes overall—again, I refer to the point that you made in the last paragraph of your submission—do not necessarily affect overall long-term economic growth.
That is probably true. Tax is one consideration among many when someone is thinking about locating their business. I have a small amount of evidence that suggests that companies think pretty carefully about, for example, moving people around. An international organisation will have people working for it in different parts of the world. Some companies are already thinking about how they might compensate employees if higher tax rates are levied in Scotland or, I guess, how they might pay employees less if taxes are lower in Scotland. Some companies are thinking about such tax planning.
When it comes to locating a business, a company will assess how important it is for its particular type of employee to have a lower or higher rate of tax and may make compensating adjustments. However, it will then look at all the other costs and benefits that are associated with the jurisdiction that it is thinking of moving to. In Scotland’s case, that would include free education, free health services and all the other public services.
Do you have any further points before we wind up the session?
No—that is it.
Thank you very much for your responses and your submission. I thank Gavin McEwen, too, for the excellent paper that he produced.
13:02 Meeting continued in private until 13:15.