Y Pwyllgor Cyllid - Y Bumed Senedd

Finance Committee - Fifth Senedd


Aelodau'r Pwyllgor a oedd yn bresennol

Committee Members in Attendance

Llyr Gruffydd AM Cadeirydd y Pwyllgor
Committee Chair
Mark Reckless AM
Mike Hedges AM
Nick Ramsay AM
Rhianon Passmore AM

Y rhai eraill a oedd yn bresennol

Others in Attendance

Bert Brys Pennaeth Uned Polisi Trethi Gwledydd ac Uned Trethi Personol ac Eiddo, y Sefydliad ar gyfer Cydweithrediad a Datblygiad Economaidd
Head of the Country Tax Policy and Personal and Property Taxes Units, the Organisation for Economic Co-operation and Development
David Bradbury Pennaeth yr Is-adran Polisi Trethi ac Ystadegau, y Sefydliad ar gyfer Cydweithrediad a Datblygiad Economaidd
Head of the Tax Policy and Statistics Division, Organisation for Economic Co-operation and Development
David Phillips Cyfarwyddwr Cyswllt, y Sefydliad Astudiaethau Cyllid
Associate Director, Institute for Fiscal Studies
Dr Ed Poole Canolfan Llywodraethiant Cymru
Wales Governance Centre
Guto Ifan Canolfan Llywodraethiant Cymru
Wales Governance Centre
Sean Dougherty Uwch Gynghorydd, y Sefydliad ar gyfer Cydweithrediad a Datblygiad Economaidd
Senior Adviser, Organisation for Economic Co-operation and Development

Swyddogion y Senedd a oedd yn bresennol

Senedd Officials in Attendance

Christian Tipples Ymchwilydd
Gemma Gifford Dirprwy Glerc
Deputy Clerk
Georgina Owen Ail Glerc
Second Clerk

Cofnodir y trafodion yn yr iaith y llefarwyd hwy ynddi yn y pwyllgor. Yn ogystal, cynhwysir trawsgrifiad o’r cyfieithu ar y pryd. Lle mae cyfranwyr wedi darparu cywiriadau i’w tystiolaeth, nodir y rheini yn y trawsgrifiad.

The proceedings are reported in the language in which they were spoken in the committee. In addition, a transcription of the simultaneous interpretation is included. Where contributors have supplied corrections to their evidence, these are noted in the transcript.

Dechreuodd y cyfarfod am 09:01.

The meeting began at 09:01.

1. Cyflwyniad, ymddiheuriadau, dirprwyon a datgan buddiant
1. Introductions, apologies, substitutions and declarations of interest

Bore da, bawb. Croeso i chi i gyfarfod y Pwyllgor Cyllid y bore yma. Gaf i nodi fel arfer, wrth gwrs, fod clustffonau ar gael ar gyfer y cyfieithu neu i addasu lefel y sain? A gaf i atgoffa Aelodau i ddiffodd y sain ar unrhyw declynnau neu ddyfeisiau electronig sydd gennych chi? Rydym ni wedi derbyn ymddiheuriad oddi wrth Siân Gwenllian. Gaf i ofyn os oes gan unrhyw Aelod unrhyw fuddiannau i'w datgan? Na. Iawn. Diolch yn fawr.

Good morning all, and welcome to the meeting of the Finance Committee this morning. May I note, as usual, that headsets are available for translation or for amplification? And may I remind Members to ensure that any electronic devices are placed on silent? We have received apologies from Siân Gwenllian. May I ask if any Members have any interests to declare? No. Thank you very much.

2. Papurau i'w nodi
2. Papers to note

Mae yna ychydig o bapurau i'w nodi. Mae yna gofnodion o'r cyfarfod a gynhaliwyd ar 6 Chwefror eleni ynghyd â chofnodion 12 Chwefror, ac mae yna bapur arall i'w nodi, sy'n lythyr gan y Dirprwy Weinidog Iechyd a Gwasanaethau Cymdeithasol ar y Bil Plant (Diddymu Amddiffyniad Cosb Resymol) (Cymru) o'r 11 Chwefror eleni. Ydych chi'n hapus i dderbyn ac i nodi'r papurau yna? Diolch yn fawr iawn.

There are a few papers to note. There are the minutes of the meeting held on 6 February this year as well as the minutes of the meeting held on 12 February, and also another letter to note from the Deputy Minister for Health and Social Services on the Children (Abolition of Defence of Reasonable Punishment) (Wales) Bill, of 11 February 2020. Are you happy to note and accept those papers? Thank you very much.

3. Effaith amrywiadau yn y dreth incwm genedlaethol ac is-genedlaethol: Sesiwn dystiolaeth 1
3. Impact of variations in national and sub-national income tax: Evidence session 1

Iawn. Ymlaen â ni, felly, at yr eitem nesaf ar yr agenda, sef cychwyn ein gwaith a'n hymchwil ar effaith amrywiadau yn y dreth incwm genedlaethol ac is-genedlaethol. Rwy'n croesawu David Phillips atom ni, sy'n gyfarwyddwr cyswllt y Sefydliad Astudiaethau Cyllid. Croeso atom ni unwaith eto.

Awn ni'n syth i gwestiynau, ac mi wnaf i gychwyn, os caf i, gan dynnu sylw at y ffaith bod eich tystiolaeth chi, wrth gwrs, yn cynnwys adolygiad o waith ymchwil rhyngwladol ar ymatebion ymddygiadol i amrywiadau mewn treth incwm. Felly, a allwch chi, yn fyr iawn, grynhoi i ni y canfyddiadau o'r ymchwil hwnnw, ac efallai dweud ychydig, yn eich barn chi, ynglŷn â pha mor ddefnyddiol rydych chi'n credu yw'r gwaith hwnnw i ni mewn cyd-destun Cymreig? Oherwydd rwy'n credu ei bod hi wedi bod yn anodd iawn ffeindio tystiolaeth sydd yn union berthnasol, efallai, i'r sefyllfa rydym ni'n mynd i ffeindio ein hunain ynddi.

On we go, therefore, to the next item on the agenda, namely the beginning of our work and our inquiry into the impact of variations in national and sub-national income tax. I welcome David Phillips, who is the associate director of the Institute for Fiscal Studies. Welcome once more.

We'll go straight into questions, and I'll start, if I may, by drawing attention to the fact that your evidence includes a review of international literature on behavioural responses to income tax variations. Perhaps you could summarise your findings briefly and perhaps tell us, in your view, how useful this work is to us in a Welsh context, because I think it's been difficult to find evidence that is exactly relevant to the situation that we're going to find ourselves in.

Yes. So, bear with me. I'll try my best to summarise this in a couple of minutes, but obviously there's a lot of evidence here. So, I think the first thing to say is that there's pretty good evidence that high income tax payers are more responsive than people at the bottom and the middle of income distribution. That looks like it's to do with the fact they have greater scope to make use of various tax planning and tax avoidance things—perfectly legally, I should add—also to migrate to other jurisdictions—it's easier to migrate if you're a millionaire than if you're someone on an ordinary salary—and also perhaps greater award of effort and performance. So, if you're a top executive, maybe you can really push the boat out a bit more. Your earnings will depend a bit more on your performance—although not always—than someone on a more regular salary. So, high income tax payers tend to be more responsive. It's usually particularly responsive for certain groups like people who are in retirement, young mothers. So, they're particular groups you might want to think about the impacts of tax policy changes on.

Capital income—so, dividends income—that seems to respond a lot more than income from employment. I think that's relevant for Wales, because, obviously, the Welsh rates don't apply to dividends income. So, you might think, 'Well, actually, the bits we've got are going to be less responsive to these tax changes', and that is true in one sense, but also, because the Welsh rates only apply to the non-dividends income, if you were able to shift your income between dividends and non-dividends, that opens up another opportunity to avoid the Welsh tax. So, for example, you could see self-employed people incorporate and then get their income through dividends, which would go to the UK Government, for example. So, I think, how those play out, the fact that the stuff that is actually devolved is less responsive in itself versus this new opportunity for, sort of, shifting, I think that's very important. I don't know what the answer is to how that's going to play out, unfortunately.

I think it's also important to mention that people that are linked or from an area tend to respond a bit less to tax differences. So, if someone is from Wales, they often have long-standing family connections, cultural connections, it's a bit less likely they're going to move for tax changes, whereas people that have come from elsewhere, well, they've already shown that they're more mobile types of people, especially if they've come from overseas. And, again, easier for them to move. So, you tend to find foreigners in particular, but, actually, probably people from elsewhere in the UK a bit more responsive. And that, again, could matter because, often, some of the people at the top of the distribution tend to be in these groups.

And I think the last thing, and this is, sort of, maybe—. I don't want to be defeatist, but, sort of, realistic. What you tend to find is that people respond a lot more to big changes than small changes, and that doesn't mean that they just respond more because they're bigger. It's that the difference is bigger than you'd expect, given the difference in the size of the tax changes. And that's because for small tax changes, it's often just not worth bearing the adjustment costs, finding out about taxes elsewhere, paying the movers, paying the accountant to tell you. So, you don't really respond at all. Once you get beyond a certain threshold, suddenly the responses are much more worth it and you see more of a proportional jump. And the reason I mention this is that despite me saying that I think that it is important to do research in Wales on these issues, the research you can get from a small tax change may not give you the impact of a big—. You can't just, kind of, scale up the figures and say, 'Well, this is what a big tax change will be.' I think that's an issue.

So, a bit of a whirlwind tour there, but I think there is stuff that can be learned. There are more likely to be responses from the additional rate of tax than the other taxes lower down, if you think about what's happening to small businesses in particular and the dividends, and recognise that this is a difficult area to get good estimates of responses on, and you can't just generalise from one reform to the impacts of all other reforms.


Before I bring Mike in, just to ask, then, because you have suggested—. [Interruption.] Sorry, Rhianon, yes. So, just to—. Because you do suggest that the Welsh Government should do some sort of research or undertake some work themselves, but from what you've said, there are pretty clear messages coming through. So, where do you think the Welsh Government should be focusing, really, and how do you think that they could make sure that that work that they would undertake does contribute in a robust way?

Yes, I think, where there is a bit less evidence is on two things. First is this issue about the incorporation and dividends income. So, I think we don't have very good evidence about the extent to which the fact that you've got these two effects—. Employment income is less responsive, but this new margin for response by incorporating, we don't know how that plays out. So, I think, if a gap opens up between tax on non-savings non-dividends income in Wales and the dividends tax rates, doing an investigation of how that's affecting the incorporation rates, how that's affecting revenues, it won't necessarily be easy, but I think that's something to look at.

And also the migration effect. So, we have very little evidence about how people respond to changes in taxes in the UK. There's evidence from the US, from Switzerland, from Scandinavia, but quite often, migration patterns are quite different in different countries. I mean, traditionally, Americans are way more likely to move in general than people in the UK. I'm not sure about the other countries. But I think what we're not quite clear about is actually how important it is that we have a lot of—. A lot of people live on the Welsh-English border. I think it's something like half of the Welsh population are within 25 miles of that border.

I think that's a Gerry Holtham statistic, yes. Sorry—I should have attributed that. [Laughter.]

But, I think, looking at that sort of migration issue is another thing. But I would say, again, in all of this, I think it's very important that Welsh Government and the Welsh Assembly recognise that you can get estimates that are as robust as possible by engaging with academics and researchers and thinking about the design, but these estimates always come with often quite wide intervals of uncertainty, and the basic reason is we never—. When you want to look at the impact of a policy you need to know what would have happened in the absence of that policy, and you never observe that. You need to sort of find a—. You need to estimate that counterfactual, and that's always the tricky part. So, my advice would also be: take anything with a pinch of salt, at least.


Okay. Thank you very much. We'll come to Mike and then to Rhianon.

I think you drifted into two areas I would like to question.

The answers did rather than the questions. But the things we know: we know that council tax can vary substantially; we know that between Blaenau Gwent and Newport it varies by up to £650 or so on a band D property, but we also know that properties are cheaper in Blaenau Gwent. So, you have that balancing effect. For exactly the same property, you'll pay substantially less.

And we also know that in America, a lot of the movement is due to age and weather rather than around tax.

The third thing we know is that an awful lot of the sports people end up living in Monaco, no matter what country they come from. So, we know those things, but small variations—people have talked about putting a penny on, taking a penny off, which would be the 5 per cent change in tax rates—do you foresee that as having a substantial effect on anybody, including the very rich?

So, I think you raise various different points there. I think there's some work that tried to look at whether we can learn from the variations in council tax what might happen for income tax. I'd be a bit cautious about doing that because they're quite different taxes. And there's pretty good evidence that if you raise council tax in an area, what happens is that property prices fall. So, it doesn't really affect migration, because it's the owner who currently owns it is the one who bears the cost of the tax not whoever ends up living in it. So, I'd be cautious about inferring too much there.

The evidence from America is that these other factors are much more important—jobs, climate, family, et cetera—but for certain groups, tax does matter. So, the evidence suggests that, actually, for some retirees, that is why some people go to Florida and Arizona and other things. They're lower tax than New York and Pennsylvania. There's always been nicer weather.

And this thing about the sports stars, you're right there as well. There's evidence that sports stars are particularly mobile. There's been a big study of footballers. There some research by the LSE, and footballers definitely don't like paying tax, as we know, with what happens in Real Madrid, I think it is. And star scientists. It seems that this group—. The people who everyone is trying to attract are the ones that seem to be most mobile. So, I think that is something that is worth bearing in mind, although I should say there are not many of these people in Wales in the first place, so there are not many of these people to lose and maybe there are some to gain.

This point about this 1 per cent thing. I read in the questions that were sent in advance that some people suggested that a 5 to 10 per cent difference in the tax rates will be needed. I don't know whether they meant 5 to 10 per cent or 5 to 10 percentage points because they're quite different; I don't know what they meant there. As I was saying, I think, for very small differences, I wouldn't expect there to be a big immediate response. It's not worth people immediately changing their behaviour for this. But I don't think there'd necessarily be no response, because sometimes, opportunities for moving or for changing jobs just kind of come along. And maybe whilst it's not worth you saying, 'I'll actually move to avoid this tax,' if something comes up, maybe it's another thing that enters your decision making. So, for these smaller ones, like I say, I'd expect there to be very little response in the short term; it might actually grow a little bit in the longer term.

When there are bigger changes there, then it becomes much more worth while actually making those big, active efforts yourself. And I think I'd expect a bigger response but also a more immediate response when the changes get bigger. So, my short answer is: I don't want to put an exact number on it, and I think it'd be a lot smaller with these smaller tax changes, but it could grow a bit in the long term.

Thank you. You've obviously prefaced all this in terms of small changes versus pipeline—who knows where that could go? In regard to your comment at the beginning around exchanging from dividends to non-dividends, not portfolios, but whatever your terminology is, how can Welsh Government and its agencies try to attempt to limit that? That's my first question. And is there a need for a longitudinal Welsh research study around taxation flow?

Yes. I think that the role of the Welsh Government in this area is probably relatively limited, I should say, because I think most of the rules around the tax base and these things are not devolved. I think, ultimately, what we should be moving towards in the UK as a whole is a system where we haven't got these big tax differences between dividend and non-dividend incomes. I was at the Scottish citizens' assembly at the weekend, and I was telling them that if you're on £40,000 a year, if you are an employee, the total tax on that with national insurance is about £11,500. Actually, I think it's £11,800. If it's £40,000 as a small business, through dividends and other things, and corporation tax, it's £7,500, so a £4,500 difference in tax. That's just nonsense and that's what's driving these behaviours is the differences in taxes. We shouldn't have these differences. So, I think what the Welsh Government maybe should be doing is lobbying or pushing the UK Government to start addressing these issues—


So, my question then, would be, to follow that up: do you think, in terms of having that current ability that we do have in terms of the limited functions that we do have around income tax, do you think that it would be absolutely commonsensical then to have those other levers that you're talking about?

So, I don't think it's a—. It's a no-brainer, I should say. On the one hand, you need to devolve quite a lot to deal with all these issues, you need to devolve not only dividends income tax, but also national insurance, as well. That would be a substantial increase in the Welsh Government's powers. The good thing with that is you could maybe tackle some of these problems and have more flexibility and freedom in other ways to change the tax system. I think the flip side of that is that it would expose you to more volatility and risk, a bigger proportion of the budget would depend on the performance of the Welsh economy, you would be exposed to the full ups and downs in the dividends income, as well, which is more volatile, and is more prone to responses.

I actually do think, as well, that—. One of the things to think about with tax devolution is we do have a redistributive system in the UK, and the way tax devolution has worked so far is on a no-detriment principle. 'We'll only take off what we're giving you.' If you start devolving more and more, you could get to a state, effectively, where areas like Wales that have relatively low, small tax bases get to keep all their taxes and then get a chunk of money on top from Westminster. Do we think the English will necessarily be willing to let Wales keep all its taxes and pay money on top? So, I think I would say, if you think about it, it's kind of a trade-off: a trade-off between having more flexibility to fix problems and sort things out versus some additional fiscal risks.

I probably would, personally, myself, devolve the dividends and savings income. I think, initially, it wasn't devolved because we had a lot of interest on bank accounts being taxed, and that was apparently difficult for the banks to sort out. That's not the case: bank interest is more or less tax-free nowadays. Dividends income is all via self-assessment, we know where they live, it's the same with self-employment. I think that can be done now and I think that is probably worth doing. Devolving national insurance and other things, I'm not so sure about that.

Yes, so in terms of the research study—

Yes, sorry.

I think that if I was the Welsh Government, I would, in advance of any policy changes in this area, be thinking about who I could contact, which researchers or what questions I had about this, and put in place a sort of research design, before I actually implemented the policies. You would not necessarily tell the researchers before you did it or before you announced it, because you need to be careful about what you announce in advance, but have definite plans in advance to do research.

Whether it's an ongoing research project, I'm not sure. I think one thing that HMRC would do in what's called KAI—the knowledge analysis and intelligence division—they do a lot of research on an ongoing basis, and I don't know whether they're doing that for the Welsh Government as well as doing it for the UK Government. If they're not, maybe that's what I'd investigate to find out, actually, are they doing this investigation specifically for the Welsh, and if not, why not?

Well, that's a question we can ask. Are there any scenarios you think where we could see behavioural changes for low and medium-income people and any impacts on Welsh rates of income tax revenues or not?

Yes. I should say that the strong evidence is these people are a lot less responsive to changes in income tax, but because there are far, far more of them, a small change can turn into a bigger cash amount. So, I wouldn't say that you should ignore the effects. In pounds millions, it could be bigger, actually; it's just a much smaller percentage of the revenue you get from them. 

Where would I expect there to be differences for these groups? Well, if you have a look at the evidence about who is most responsive, it could be—people with childcare responsibilities seem to be a bit more responsive to differences in tax rates and benefit policy—mothers with young children. There's some evidence that when people move to retire that can be a factor there. And there's some evidence, although this is a bit unclear—some say it's fine, some don't—that young graduates can be a bit more mobile, because they haven't set up their families and their careers yet. So, they may be a bit more willing to move around. Evidence from Switzerland suggests there's some of that going on.

So, those are the groups you might want to think about. I know the Welsh Government's been thinking more generally about what to do with young graduates who come to Wales to study, or leave Wales, and there's a bit of an in-flow of older people and an out-flow of young graduates. It could have an impact there, but probably relatively small.


Okay, thank you. Can you elaborate on how the elasticity of taxable income can be used to model the responsiveness of individuals to income tax rate changes and how this can be applied to Wales?

Yes. Let's see if I've got some notes for you on this one. So, what I would say on this issue is that the elasticity of taxable income is sort of a summary measure. It's a good way of summarising the overall effect of a change in the tax rate on the amount of income people report to the tax authorities. So, it is what is used by the Scottish Government, HMRC, the Office for Budget Responsibility, the Scottish Fiscal Commission, and I think the Welsh Government as well, in their modelling.

So, you take this summary measure of responsiveness, you have the amount of people's incomes, you've got the tax system, you put them all together in a little formula, and it will tell you if you change the tax by 1 per cent, incomes will go down by 0.2 per cent, and that will offset 10 per cent of the direct mechanical effects of my tax change. That is really useful and it's a good summary measure; we do that in the IFS, and others do it.

But it only tells you part of the story. It tells you the overall effect on the tax base, but you could care about other things as well. So, if you're thinking about the wider effects on the economy, you'd care about: is that people's real response? Is it them migrating or changing how much they work? Or is it through tax avoidance, tax evasion, those sorts of things? That's when you need more detailed work to look at. So, I'd say it's very useful to look at the revenue effects, but for thinking about the wider economic effects you need to do more analysis, more research. I would hope that any research the Welsh Government commissioned or did internally or other things would look at those other things as well.

Okay, thank you. We've touched on how influential the non-tax factors are. We've not touched on housing costs, necessarily, but it's those kinds of things, isn't it? Is there anything you want to add? Living costs, housing costs—we've covered some of that—the weather, even.

Yes, living costs, housing costs, the availability of jobs, cultural similarities seem to matter, family connections, many of those things.

And they tend to be more influential, or would potentially be more influential than the economic factor in the Welsh context, unless the changes are more substantial.

I think overall they're more important for these kind of things, but that doesn't mean that there's no effect at the margin. But yes, they are, as I say, very important.

And I think we've probably covered some of the stuff around the subsets as well; older people, and we've mentioned the professional footballers, potentially. Okay—yes, go on then, if you want to pick up.

Sorry, Chair, on that question in terms of young mums and retirees and more mobile graduates, I completely understand why we're focusing on those groups in terms of an ability to do so. In a sense, it's the missing middle I'm more interested in, and how we can actually—. It goes back to the earlier question. In that regard, how do we measure the responsiveness of those who are outside of those particular groups that we are very focused upon, who are more likely to escape the system?

Yes—the sort of analysis you'd want to do to analyse the impact of these reforms would depend exactly on the specific changes you're looking at. But the basic strategy is you look at the group that's seeing a tax change, you find a similar group that's not seeing a tax change, and you compare how their behaviour differs before and after the tax change. That's the basic approach. And you can apply that to any of the population, provided you can find that control group, as we call it. It's often easier to do that for specific groups, because it's more easy to find a control group for them, but—


Yes. There's nothing stopping you looking at regular people on incomes of £20,000 to £40,000; there's nothing stopping you looking at the impacts on those.

A couple of points. One is: you said about giving additional money earlier, over and above the income tax, isn't that what happens in Sicily at the moment, or used to happen in Sicily, and what was happening in the Azores as well? With Portugal and Italy, they actually gave more than their taxes back. So, the two questions I've got are—. You talked about mobility, but you can be mobile and not have to travel very far. If you accept that Wrexham and Chester are effectively one conurbation, which Llyr probably won't—

They're effectively one conurbation in terms of population movement. If you want to work in Chester Hospital or Wrexham Maelor then you could live either side of the border. Then, surely some small changes can have some effect? It's the same with teachers teaching on either side of the border. The third point I was going to make is that for big movements you haven't factored in, for young families, childcare costs, which can be free from family and exceptionally expensive when not.

I think all these factors matter. There's good evidence—well, I say 'good' evidence, there is some evidence that migration is a bigger thing when they're smaller distances. So, where there are differences in tax rates between small areas, like Swiss cantons or Swedish municipalities, migration seems to be more of an issue there than it is across US states or across countries. So, I think that is a really important point with the densely populated Welsh border. I think something like 8 per cent of the Welsh tax base comes from people working in England. Those people, you might think, would be particularly mobile as well. So, I think the point you raise there is very important.

On the childcare costs, yes, that's another reason why it's often good to be close to your family. So, I wouldn't expect necessarily families with young children to be migrating to avoid these taxes; I'd more expect them to be changing their behaviour in terms of how much they work and how they organise their affairs.

Sorry, can I just throw one comment in before I ask another question? You've also got the Ontario-Quebec situation, haven't you, where you've got quite a lot of movement there as well, and they run quite differently?

The question I've got is—we've also got England and Scotland, now Scotland has got its own tax rate, and it did drop £1 billion or so, which may be because of behaviour, may be because people decided where to register, or may be because HMRC got it wrong, or probably a combination of all three—are we expecting, as the tax rates diverge in terms of income tax, for there to be changes, and are you monitoring that, because that may well give us some information that could help us in the future?

We also know that house price changes and other forms of taxation don't—. You could tax housing, but all it does is affect house prices. If we put a 25 per cent tax on all house sales, all that would happen is that house prices would drop by 25 per cent. If we gave a 25 per cent bonus to anybody buying a house, all house prices would go up by 25 per cent. We've seen that with Help to Buy, which has helped to inflate house prices. So, sometimes, what you're doing is pumping money into the system and somebody, somewhere does well out of it.

I think there are several different questions there. I think that, on this issue about Scotland, it's too early for us to say as researchers what the impacts have been. It takes a while for HMRC to process the data, put them into the research lab, and for us to get our hands on them and start doing the research. So, if I was looking at the impact of Scotland, I would at this stage say, 'Well, let's get some accountants and others who advise on this kind of stuff and see what they're saying about Scotland.' I think they'd have the earlier first-hand things. I think the research will follow, going ahead.

I was actually talking to colleagues about could we do some research in this area. Talking about the housing market, I was talking about could we do some research looking at the differences in stamp duty in Wales versus England, and the differences on first-time buyers, et cetera.

On the second point you were making around the property taxes and their different incidence, I completely agree. I'd probably expect taxes just to change property prices. Income taxes—actually, differences in income taxes probably matters a bit, because people can move to avoid them, but I'd expect them to affect behaviour more.

So, I think that difference between property tax versus income tax is an important one. I think it comes back to this point that I've not made so far, but have often made, that if we're thinking about making the taxes more progressive to get more income from richer households, it’s not only the income tax system you can use in Wales, you can also use the property tax system. That’s important to think about as well.


I've argued on a number of occasions that we need more higher rates on property. The fact that it stops at £250,000—you've got a £5 million house or £2.5 million house and you're still paying exactly the same council tax at the top rate.

You'll be very pleased to hear a report's coming out on 2 April looking at council tax reform in Wales. It's being launched, I hope, in city hall.

Thank you. Yes, you want to come in, Nick, and then we'll come to Mark.

It was just on that point, and again going back to Gerry Holtham who used to support the committee a lot, he said you could look at a variation and a move towards property taxes, and reductions in income tax, potentially. It's all part of the mix, but we often forget that the property taxes are part of the system as well.

Yes. I think you could potentially do that. I think that, if you were to shift the burden of tax, especially at the top end, from income to property, that would probably lead to slightly stronger work incentives, and it would lead to probably a slightly more attractive environment for high-income people to come in to Wales, because it would be the property prices that fall. So, the property tax increase gets borne by the current owners, and the new incomers would get the benefit of the income tax cut. So, I think you could do that.

I think you do need to think about the fact that it would affect different people. So, you could have a progressive property tax, or make it more progressive in coming years, but it wouldn't be quite the same people who are paying that who are currently paying the income tax. It would tend to be the current owners of the properties. It would tend to be older people, to some extent, who often have bigger properties and their property wealth is higher than their income wealth.

We already, compared to most countries, get quite a high share of our revenues from property taxes. I think in the UK we get about 3 per cent of our gross domestic product from business rates and council tax. Equivalent taxes in other countries in Europe are about 1 per cent of GDP, so we already have quite high property tax. That doesn't mean we shouldn't go higher, because they are a good form of tax, but it would make us a bit more unusual compared to other countries. Where other countries tend to get more revenue actually tends to be from middle earners. Most other countries have higher tax rates than we do; it's people on £30,000, £40,000, £50,000 they get more money from, largely on national insurance type of things.

How likely would it be for self-employed people to incorporate their businesses as a consequence of tax rate changes, and what impact could that have on the yield from Welsh rates of income tax?

What I would say is that there's pretty good evidence that people are responsive to these tax differences between employment, self-employment and incorporation. We have seen a steady growth in the proportion of people who are incorporated over time. Some of that is because of changes in technology and the way that we work. Some of it seems to be because of these tax differences. When we changed the tax rates back in the mid-2000s, there was 0 per cent corporation tax for small businesses, and it shot up the number of incorporations. So, you do see people respond to these things.

I wouldn't want to put a percentage number on it. Again, I would probably recommend talking to some advisers about how many people are coming through the system, and actually looking at the statistics. I've not looked at them myself. But, look at the statistics. Look at them in Scotland. Do you see, in Scotland, a spike in incorporations now you've got this incentive to do so? I would do that.

In terms of the revenue effects, this actually applies to all of the behavioural effects of Welsh tax policies. The way that the devolved system is set up is such that Wales bears the full mechanical revenue effects of change in policy, so you put up the rate by 1 per cent, you get that full 1 per cent extra tax rate; you cut it at 1 per cent, you bear the full cost. But the behavioural effects, Wales only bears a small proportion of that; it bears a proportion on its share of the taxes. So, say, for the top rate of tax, Wales bears the behavioural share on the 10 percentage points that are Welsh. The other 35 percentage points that are going to Westminster, well, Westminster bears the behavioural effects on that share of income tax. So, the point I was making there is that because Wales gets the full mechanical revenue effects, but only gets a small proportion of behavioural effects, tax increases are more likely to raise revenue, and tax cuts are more likely to lose revenue than if Wales bore the whole of everything.

So, I think that could mean actually that there's a bit of a bias towards tax increases in the Welsh system, because if you put the rate up, you get all the revenue, but you don't bear all the costs in terms of behaviour change. 


No. So, why would a Conservative Government at Westminster have devolved taxes in such a way to Wales, as to skew the incentives with what you describe as a tax-rising bias?

I think there are two possible reasons for that. One is that tax devolution is seen as a way of giving greater accountability and greater flexibility to Governments to set their tax and spend envelope, to give them a bigger stake in the economy, so they have incentives to grow the economy. But you may not want to go too far, so you may not want to devolve all the rate-setting power to Wales, the whole chunk of income tax, for a couple of reasons as well. One is that that would mean more risk being transferred to the Welsh Government's budget. Secondly, the UK Government probably still wants to have powers to change rates in Wales itself, so they can still increase the personal allowance; they can still do something that effectively allows them to offer good news to Welsh voters, as well as English voters. 

Also, this point about this bias towards tax increases, that's partly offsetting another thing you kind of get sometimes. So, when you devolve taxes to smaller areas, you also get tax competition. So, if Wales got the full behavioural effects of it, as well as the full mechanical effects of the reform, if a lot of that behaviour was actually people migrating from England, then maybe the UK Government doesn't want the Welsh benefiting from all that migration from England, because it's robbing Peter to pay for Paul. So, these decisions about how much to devolve are kind of balancing off various things; they're balancing off wanting to give more accountability and incentives, and the correct incentives to Welsh Government, but also not wanting to devolve too much risk; not wanting to remove the UK Government's powers, and not wanting to have this sort of tax competition and worries about races to the bottom, I think, as well. 

In other areas, we're told that when one Government in the UK makes a decision that has a sort of fiscal impact on the other, then that's meant to be offset. One area that Welsh Government complains about is teachers' pension contributions and those not having been offset. In this area, is there any proposal to offset the behavioural impact on UK Government of tax changes by Wales, or is that just left as a dead weight loss for UK Government without the compensation?

Yes. So, I should say that I know the Scottish fiscal framework a bit better than the Welsh one. I think Ed Poole who's coming in later might be able to correct me on this, but the Scottish one, I think the Welsh one, includes a provision for this sort of compensation mechanism, so that if there's—

The no detriment. So, if there's a knock-on effect from the decisions of one Government onto the revenue and spending of another, a transfer can take place. But, the rules around that are that, unless there's agreement that that should take place, and what the size of that should be, it doesn't happen. And, of course, this is a bit of a zero-sum gain, because the Welsh Government will say 'Oh, no, this is a small effect on your revenues', and the UK Government will say that it's a big effect—how do they get an agreement? And if the default is, if there's no agreement nothing happens, nothing happens. 

I think, also, it depends on the attitudes of the people in power at the time. So, when George Osborne was asked about this a few years ago when it first was raised in Scotland, his view on this was 'Well, this is what it's about; it's about giving flexibility on taxes, and if Scotland wants to cut it and draw more business in, that's just what this is about; we wouldn't be pursuing this.' I don't know what—I was going to say Sajid Javid—Rishi Sunak thinks. I think that will play a role as well—how they interpret these rules.


And just returning—finally for me, Chair, if I may—to the specifics of the dividend tax, to the extent that there have been changes to that in the past, the UK Government has had that as well as the income tax levers, and if there appears to be leakage or avoidance, it can change the regime to keep the two with a degree of balance. Because the dividend tax is staying with the UK Government, where the Welsh rates of income tax are devolved, would we expect to see a larger response, in terms of incorporation, if Welsh Government were to raise income tax rates, given that the UK Government is maintaining the dividend tax?

Yes. I was saying at the start of the session, I think there are two effects about the fact that Wales has only the got the NSND. One is NSND income itself is less responsive, so that needs to be small responses in Wales, but—

I apologise—I probably missed this earlier. Could you just define NSND?

Non-savings, non-dividend.

The fact that that is less responsive than dividends income means that the responsiveness of the Welsh tax base to tax changes is smaller. But because the Welsh taxes don't apply to the dividends income, there's this new margin—or the margin of adjustment of incorporation is probably more important. So I think you're right that this issue is more important because it's not devolved to Wales, and I think that—. Again, in an earlier question, I said that dealing with all these issues requires a lot of devolution. It would require devolution of national insurance, as well as income tax. But I think there probably is now a case to devolve at least the income tax on dividends income. Because I think, initially, it wasn't because of the issue about savings income and banks. That's not an issue any more so much, because basically savings income is not taxed; it's just a dividends income. I think that could be devolved, and that would allow the Welsh Government to keep these rates in alignment. And if it wanted to actually close up some of the gaps that are there already—and I think we at the Institute for Fiscal Studies would approve of such a move to close up the gaps—I'm not sure whether the incorporated businesses would.

If we were just to raise the tax rates on everyone, you might expect that there could be some sort of overall reduction in the level of economic activity in Wales. So another option would be to say, 'Well, let's reduce it a bit for the employees, and let's increase it a bit for the self-employed and the—well, increase it a bit for the dividends income.' That would make it a bit less attractive with dividends income in Wales, but make it more attractive to the employee. So by having the powers you can level up, you can level down, you can level across—different options.

Okay, thank you. Rhianon, you wanted to pick up on something, and we'll come on to your questioning as well.

Actually, you've partially answered it, but it was just to probe a bit deeper. With regard to the anticipated move to incorporation, which we're discussing, and the fact that we don't have those levers at present—and I asked you earlier about whether Welsh Government, in its limited capacity, would have any ability to get involved in that regard. What actually could Welsh Government do, bearing in mind that it is, basically, a Westminster issue, with regard to that move to incorporation, in terms of being able to sit in a different basket—and should it? I know it's difficult to answer.

Yes. I think on this one, what I would say is that I think the Welsh Government should be monitoring what is happening in this area. In terms of practical things it could and couldn't do, I don't think the Welsh Government would be wanting to actively make it harder to incorporate and, you know, 'Oh, we'll be less supportive to incorporated businesses than we will be to the self-employed, in terms of grant support, and other things'. That would be distorting the economy in other ways. So I think that, if I was the Welsh Government, I'd be monitoring and mindful of this issue, and I'd be engaging with both HMRC on monitoring the impacts, but also potentially—for this savings and dividends income, is it something we should actually be pushing for now, now that the original reasons for not devolving it are not quite so significant? Should we now be pushing for this, and that would let us tackle these issues? But, yes, it would also expose the Welsh Government to the more volatile part of the revenues, as well.


Okay, thank you for that. We've touched upon my next question somewhat. With regard to small changes versus significant changes, how could this be applied to future research undertaken in Wales? I have asked you already if that should be happening. It's a bit of a no-brainer, I think, but how could that happen?

This is a very tricky question. Effectively, if you have a small change to estimate the effects off, that's what you've got; you can't turn that into a large change and then estimate the effect—

What you can do is you can get an answer for that small change, and given that, as I said, the evidence is that larger changes tend to lead to a more than proportionally bigger response, well then multiply it up and then look at the evidence from other studies and say, 'Okay, I'll multiply that by a factor of 1.5 or whatever it is.' I think that that is really the best you can do.

It's a factor amongst others.

My next question is with regard to—and again you've touched upon this—the mechanical effect that you've discussed, and the behavioural effect play-off. So, can you explain how this would impact on Welsh rates of income tax revenue, and how it could influence the Welsh Government's decisions to vary income tax rates?

Well, we have touched quite a bit on this, but if you want to add anything—

Yes. I think that all I would say on this—just to make it as clear as possible—what I said earlier is that this boils down to the fact that you have partial devolution of income tax but, if you like, the full devolution of the changes in income tax. Therefore, the revenue effects: you get the full mechanical effects, but you only get a part of the behavioural effects. And, because the behavioural effects tend to offset the mechanical effects, that means that tax rises tend to raise more in Wales than they would if we had full devolution in tax. Tax cuts tend to cost more than if you had full devolution.

That can be seen as a distortion. It can also be seen as trying to prevent this sort of tax competition effect as well. I have talked to the Welsh Government a bit about this. I think that they are mindful of this issue, but they see tax policy as, you know—it's not just about revenues; it's about the other things that they are trying to do as well. With that in mind, it's the effect on people. It's the effect on the economy. It's the effect on business. That is the key to this.

Thanks, Chair. What consideration should be given to spillover effects resulting from Welsh Government policies, and what processes should be implemented to ensure that all effects are identified? It's quite a broad question that relates, in part, to what you were just saying about the full mechanical effect versus the partial behavioural effect.

Can I check what you mean by 'spillovers'? Do you mean spillovers on England, Northern Ireland and Scotland? Or, do you mean spillovers on the economy, and spillovers on—?

Spillovers in terms of the economy. I mean, there is the issue of the effect on the rest of the UK, but thinking specifically about Wales.

Yes. In terms of the spillovers on the wider economy, what I would be looking at—. I think that there would be several things to think about. First of all, it's just looking at rates of incorporation, rates of employment, rates of income growth, rates of migration and the effects of migrating. I would be looking at all of those things.

Do you think that there is much of an effect on migrating? We discussed this a bit earlier, and some people think that there could be quite a migrational effect and others don't.

The evidence on this is mixed. The evidence suggests that most people are not that responsive to tax changes in terms of migration, but some people are quite responsive in certain groups. If, for example, the Welsh Government had a policy of trying to attract top scientists and inventors and researchers to Wales, the evidence is that these people are pretty responsive to taxes. High taxes on the top incomes may not be so conducive to getting those people to move to Wales. If that isn't a core policy, and the core policy of the Welsh Government is a more general policy of making Wales attractive to middle and high-income people more generally, well, for those people more generally, tax doesn't seem to be quite such a big factor, and maybe other things are more important. So, I think it depends on what the precise objectives of the Welsh Government would be. If you wanted to be the Monaco of the UK, then taxes clearly matter a lot; if you want to be the Sweden of Europe, then it matters less—doing other things matters. 


And, finally, how should Welsh Government policies be monitored to mitigate any unforeseen—back to the spillover—effects, both in the short and the long term? 

Again, I would just say that commissioning or undertaking research to actually estimate these impacts on wider things, recognising that you're not going to get an exact answer—. This is not an exact science, unfortunately. Economists like to—. We like to think of ourselves as scientists, and we are to some extent. But we also, you know—. So, with that in mind—    

Well, it's not really—. It's different to forecasting, because what you're doing is you are trying—. Forecasting is looking to the future with what you can do. You can try to use the data from the past to look at what the impacts would be. But because you never observe the world without the policy in place, it's only ever an estimate of what it's going to be and those estimates come with uncertainty. But that doesn't mean you shouldn't do it. I think it just means thinking carefully, working with the right people, and, yes, taking the estimates with a pinch of salt.

I think also the Welsh Government just needs to think what is its strategy as well. Some of this just comes down to what is our strategy: do we see tax as part of our strategy for making Wales more progressive, a place with better public services, or do we see the tax as part of our strategy of making it more attractive to businesses and high earners?

I'm going to ask you, then, the $6 million question, and you're going to say—I would imagine you would say—that's not really why—

I am asking, because I want to seek your answer and I think that it's—. Overall, do you think that there will be? Is your assessment that there would be big unforeseen consequences from this? And is it going to be—? Is it going to be hard work to try and mitigate it?

Look, I don't want to, again, give any precise predictions on what the impact of cutting the rate by a couple of pence would be, or increasing the rate by a couple of pence would be. Do I think there'd be an impact? I think there will be some impact on behaviour. I think it's more likely to be on the margin of incorporation, if that's not devolved to Wales. I think that we could see certain types of people who would be more likely to migrate in or out of Wales, depending on these groups. I probably—. I'm almost sure that the revenue effects of a tax increase would still be positive and that tax cuts will be negative, because of this partial devolution point.

The wider effects—. I don't want to say, because I think we just don't know enough evidence about how responsive people are, in terms of their migration across the UK. And I think at some point, yes, the Welsh Government just actually needs to have a strategy. I think what would be more important for the performance of the Welsh economy is having a clear strategy about what the Welsh Government is seeking to deliver overall with tax as just part of that package. 

You said something that a lot of people have said: that tax has been devolved to incentivise the Welsh Government in order to grow the economy. I disagree with large numbers—probably most—of the Members of the Assembly on it, but I can't believe that any Member in the Assembly does not want to grow the economy and any Government does not want to grow the economy, whether they get the financial benefits in terms of income tax or not. Do you know of anywhere where actually devolving tax has incentivised a Government to grow the economy over and above what it did before? 

So, there is evidence from a number of countries—well, cross-country studies, actually, by the OECD—that tax devolution is associated with somewhat stronger growth, somewhat better performance in the economy. I should say that this evidence is not as strong as sometimes is claimed. Cross-country progressions often aren't, because you put Sweden in it and everything looks good in these countries on tax devolution. But there is some evidence on this. There's a better study, actually, in Germany, which looked at when they changed the rules about local tax—state-level taxes—so that the state bore more of the variations in the revenues, and they did find that collection of taxes increased. So, I think you're right: politicians do care about their constituents and they do want the economy to grow. They're counting on being re-elected. But I think there is some evidence that this does have a bit of an effect, and I think one of the arguments for this is actually bureaucracies. Bureaucracies like to be big. One way for bureaucracies to be big in countries that have redistributive tax systems is to be poorer, because we have redistribution, and poorer places get more money. If you're a manager, a sort of bureaucrat, you want to have a big empire. That is the classic argument for the opposite side of what you're saying, Mike. But, yes, I completely agree with you that it's not just about the finances that matter for growth. It is, you know, the goodness of people's hearts and their chances of being re-elected that also matter.


Excellent. Okay. Well, can I thank you, David, for the excellent evidence, as always, that you've given us again this morning? It's a good start to our work in this area. You will, as always as well, be sent a copy of the transcript to check for accuracy. And, with that, I'll thank you on behalf of the committee for your attendance this morning.

Diolch yn fawr. We'll now move into private session, and we'll reconvene at 10:00 in the hope that we've established our link with Paris to take evidence for our next session. Diolch yn fawr.

Gohiriwyd y cyfarfod rhwng 09:56 a 10:01.

The meeting adjourned between 09:56 and 10:01.

4. Effaith amrywiadau yn y dreth incwm genedlaethol ac is-genedlaethol: Sesiwn dystiolaeth 2
4. Impact of variations in national and sub-national income tax: Evidence session 2

Croeso'n ôl i'r Pwyllgor Cyllid.

Welcome back to the Finance Committee.

Welcome back to the Finance Committee, and to our second evidence session, and our fourth item on the agenda this morning. We have representatives from the OECD and from Cardiff University—the Wales Governance Centre, really, isn't it? So, maybe I'll ask you to introduce yourselves, starting with the Wales Governance Centre here.

Diolch yn fawr iawn, Cadeirydd. My name is Dr Ed Poole, and my colleague's Guto Ifan. We're from Wales Fiscal Analysis, which is hosted at the Wales Governance Centre at the School of Law and Politics at Cardiff University, and it's funded through a partnership of Cardiff University, as the primary funder, plus funding from the Economic and Social Research Council, the Welsh Government, Welsh Local Government Association and Solace Wales, and this programme's continuing the work that was hosted by Cardiff Business School through Wales Public Services 2025 up until August 2018.

Thank you very much, and welcome to our colleagues in Paris as well. Would you like to introduce yourselves?

Well, thanks very much for the opportunity to be here. I'm David Bradbury, head of the tax policy and statistics division at the Centre for Tax Policy and Administration at the OECD. I'm joined by my colleague, the head of the personal and property taxes unit, Mr Bert Brys. Bert is also the head of our country team, and specialises in providing country-specific advice. And on my left is Sean Dougherty, who is the senior adviser of our fiscal network. The fiscal network is the network on fiscal matters across levels of government that deals specifically with issues relating to tax and spending across levels of government. So, it's the three of us with you today.

Thank you very much. And we'll, obviously, choreograph as best we can the meeting, given that, of course, you're joining us down the line, and please feel free to interject at any point where you feel that's necessary, but we'll also try and be a bit more clinical in directing questions as we go along. So, I'll kick off, if I may, and this is a general question to all of our guests this morning, really. Evidence has shown that it's difficult to draw conclusions about behavioural responses in Wales from other countries and tax jurisdictions. So, what data do you think should be collected to better inform the relationship between tax rate variations and behavioural responses in Wales, and how do you think that should be done? And maybe we'll start with the Wales Governance Centre.

This is certainly one of those areas in which academics have to give that most frustrating of answers: in other words, the evidence is very unclear. And we've seen that from all of the evidence that's been submitted so far to this inquiry. That's actually not through a paucity of studies—there have been quite a large number of studies in a large number of developed countries, but the problem is they show very various effects of tax rate changes. I refer you to David Phillips's and the Welsh Treasury's excellent literature reviews on this, which show the wide variation and a highlighting of the ambiguity all in this. And it's actually—. This ambiguity is true even within the same country, like Switzerland. We think that it's less likely for there to be effects where there are very hard cultural and linguistic lines that follow the sub-national boundary, as is the case in Belgium. But in countries like Wales where it's less likely for there to be that hard boundary, the evidence is very clear. So, this is just going to need a lot of work over a period of years. It’s not something, in terms of data collection, that we can do overnight. This is going to be something that HMRC particularly, as the tax collection authority, are going to have to monitor over some years.


Okay. And in relation to the OECD, you've published a number of papers, of course, looking at taxation and competition between sub-central Governments, and I'm just wondering whether you could elaborate on some of those findings briefly, but more importantly for us, whether we can relate any of that work to the Wales-England border situation, which we're particularly interested in.

Well, thanks very much. I'll offer a few remarks and then I might invite Bert to supplement those. I think one of the first things that we would observe—and we're not entirely privy to all of the discussions that you've been having—is we would always approach questions such as this by thinking about the personal income tax in its broader context. It is an important part of the overall tax mix, and, obviously, with the devolution changes, it is something that is the focus of your attention, but we think it's important to be constantly thinking about the overall tax mix. But even in the context of the personal income tax, it's important to take into account that it is often the principal form of labour taxation, but it's not the only form of labour taxation. I think it's important to take into account other labour taxes, such as social security contributions and the impact that they will collectively contribute too in terms of the overall tax burden on labour.

I think one other point that we would like to make is that, in broad terms, we would think that the sort of frame through which it would be useful to think about these questions is, in asking that question, 'What is the optimal personal income tax system and optimal set of rates?', it's worth focusing, obviously, on the revenue implications and what revenue is yielded. It's important to think about the efficiency considerations at which much of what your question is directed towards goes to, but then there is also the equity considerations.

I think one of the things that we would observe is that in the personal income tax base—and the personal income tax is one of the most progressive taxes in the toolkit that policy makers have, and, in that sense is, obviously, very important—one of the things that we observe in many jurisdictions in relation to the personal income tax is that where there are significant differences in tax rates of the personal income tax compared to the corporate rate of taxation, there are opportunities for restructuring of activities and many higher income earners are able to change the nature of their activities or the way they're structured in order to reduce the tax that they pay. Now, that's a challenge whether you're in the context of considering a post-devolution environment or not, but we would suggest that it’s particularly relevant in this case because there can be significant gaps between top rates on top earners and the corporate rate, but, equally, one of the challenges that you face is that you do not have much control over the taxation of capital income, whether it be through the corporate vehicle or whether it be in the form of the taxation of dividends or other forms of capital return. So, we think that these are considerations to bear in mind.

A couple of final observations before I hand over to Bert. I think that the question of mobility is one that we reflect a lot on when it comes to tax policy. Clearly, some taxes involve much more mobile tax bases than others. Now, I think it is fair—and the evidence would support this—to suggest that higher income earners tend to be more mobile than those further down the income distribution. That said, labour mobility varies from context to context. In my particular background, I come from Australia, and I can tell you that labour mobility takes on a very different type of proposition when the opportunity to move to a new set of tax arrangements means you have to actually hop in a plane and travel for several hours in order to relocate to a vastly different place. When that mobility question relates to perhaps moving one postcode, moving a couple of streets or a block, then it is a different question. There are, of course, a whole range of transaction costs and barriers associated with relocating, but they may be costs associated with relocation, and one of the challenges that tax administrations all around the world face is that sometimes, particularly for higher income earners, there can be a greater ability to relocate as far as their tax presence is concerned, without that necessarily having all that fundamental an impact on their lifestyle or the other issues associated with the decisions that they make.

And of course, if this is a real relocation, wherever you live, there will be different costs associated with living there, different housing costs, different levels of cost of living, and they will be factors that need to be taken into account. Of course, when you're talking about the difference between living on one side of the border or the other, a couple of blocks away, then some of those issues are obviously less significant. But we would say that those costs associated with relocation should be taken into account. They tend to suggest that where there is not a significant difference in rates, in those instances, we're less likely to see the sort of mobility issues. But I'll hand over to Bert, who I'm sure is an expert in this area and will have much to add.


Thank you, David, for that. Good morning, everyone, also from me. Before I raise a couple of additional issues, I first want to say congratulations to the colleagues from the Wales Centre for Public Policy. I've read the report on the Welsh tax base. It's really, really excellent work; it's got a lot of interesting material. I really enjoyed reading that, and with all the simulations.

Now, the questions that we received as part of this intervention are very much built around this particular report, so before we come up with some additional arguments that answer the question that was raised, I really first want to take a step back, if you allow me to do so, because I think there are other elements that need to be discussed as well, or there are other perspectives that you could take when analysing your PIT-rate schedule. And that's also what David said. Basically, the question that you now have is: what is the optimal PIT-rate schedule for Wales? The question then is—there was a PIT-rate schedule, the one before, before you started having some taxing powers over the parts of the rates, so the question is: the PIT-rate schedule that you basically had, was that optimal? Given the economic situation and the challenges and opportunities you face in Wales, was that PIT-rate schedule optimal? And did that answer, did that help in tackling some of the challenges you were facing? So, I think that's an important question to ask in addition to the analysis made in the report.

Can you clarify what you mean by 'PIT-rate schedule'? I don't know if it's the—.

So, it's the brackets and the tax rates. Whether the basic allowance of 20 per cent or 40 per cent or 45 per cent, whether that rate schedule is optimal given the economic issues, the labour force, the level of skills, the types of employment, all of these issues that are discussed in the report—whether that PIT-rate schedule is aligned with your specific economy in Wales. That is a question that I think is very important, which goes way beyond the analysis of pure behavioural analysis of the question of what the behavioural responses are in response to our PIT-rate schedule, and I think that is a question that you might want to answer as well. How do you take—? How do you try, for instance, to increase labour market participation in your economy? That's discussed in the report. So, your labour market participation has been low—the one, in general, in the rest of the UK, as it is described in the report—. You have ageing-related challenges, so how—? You know, that's a summary—that's analysing what your challenges are, and then see how that PIT-rate schedule can help to address some of these challenges, while taking, of course, also, as David mentioned, not only focusing on efficiency but also on equity and the revenue side, of course, of the personal income tax. So, that is, I think, first, a very important part of the answer.

Also, I want to stress, and perhaps we can talk about that in more detail later, because you do refer in your questions to the Scottish reform and the Scottish setting, but, as David said, it's extremely important not to see the design of your personal income tax in isolation from the social security contributions that are levied. I'm not here to—. Well, I'm not an expert on the Scottish tax system or the personal income tax rate, but I did see something that really makes me wonder—. They introduced a particular bracket where they increased the rate from 20 per cent to 40 per cent, while on that income level, Scottish people still pay social security contributions. For a large group of taxpayers, they drastically increased the tax burden, the overall tax burden, on labour income, not only—. Because it's that interaction also with social security that matters. So, that's really an issue also that you need to take into account. For that, we do have models on our side where we can actually calculate the overall tax burden on labour income when employee and employer social security contributions are taken into account—the personal income tax rate schedule—but also the earned income tax credit. So, any decision on a personal income tax rate in Wales will need to take into account how that will interact with the earned income tax credit of the United Kingdom. I think these are two key elements.

And then, just to reiterate what David said, because you asked us to confirm also what colleagues said—you can try to measure the costs of moving, because you talk about what the costs are mobility are. Well, for a typical household that moves to another country, you have the transaction costs—be it taxes when you buy a new house, or when you trigger capital gains tax. And just moving in itself creates financial costs for a household, so you can actually try to calculate these types of costs. The economics department in the OECD has, some years ago, identified and then has reviewed the overall costs of mobility in terms of relocation of households. That is, perhaps, a starting point, but that gives you a bit of—. That measure can show you, basically, the net present value of the tax gains that a household would get if it would move to a region where the personal income tax is lower. A rational household would balance these costs or compare these costs with the direct moving costs. So, that gives a bit of a margin of how you can change your PIT-rate schedule without inducing too many behavioural responses in terms of actually moving from where you live. But, again, to reiterate what David said, the most important behavioural response is just the change in your tax base, people that incorporate a business—that is, of course, a more important element.


Sure. Well, thank you for that. There's a lot in those two answers, and I know we will be coming back to some of those points as we proceed with questions. But one area in particular I want to bring back to the Wales Governance Centre is the fact that, of course, we do see from most evidence that it's the high earners that are more likely to be more sensitive or responsive to changes in income tax rates. But, of course, we have a smaller proportion of those earners here in Wales. So, what impact do you think those kinds of changes would really have in terms of WRIT revenues here in Wales in reality?

So, even though we are less dependent on the incomes of the very highest earners—. If you look at the income distribution in Wales, those earning over £100,000 puts you into the top 1 per cent of income tax payers in terms of non-savings, non-dividends income. They'll be contributing around 10 per cent of the overall tax base, so that's more than the lowest earning 40 per cent of taxpayers. The top 10 per cent of taxpayers—those earning over something like £40,000 or £50,000 a year—will be contributing more than the lowest earning 70 per cent. So, trends and migration decisions—the behaviour of those highest earners—will have an effect on the overall relative performance of the Welsh economy.

If you look at comparable revenues across the UK—so, what are the block grant adjustments that we need to keep pace with, if you like—that revenue is highly impacted by the trends among the very highest earners and what happens to those earning very high incomes in the rest of the UK. So, we do need to keep an eye on the relative performance of the very highest earners as well in Wales, compared with the rest of the UK.


Okay. We need to make progress, so I'll move on to Rhianon now. Thank you.

You've partially answered this. If I could ask the OECD first—I don't know which camera I'm looking in, I'm slightly myopic—in regard to low and medium earners, what effect could those groups of taxpayers have on Welsh rates of income tax revenue, in your view?

I think that one of the issues that we always like to focus in on is the interaction between the tax and social security contribution systems, and the incentives for low-income earners to engage in active labour market participation—to get a job and to work more hours. There can be disincentives, particularly when one takes into account the transfer payments system and other forms of support that might exist across the economy. We are generally quite supportive of things such as earned income tax credits as a means of delivering some support to low-income and middle-income earners in order to support greater employment participation.

These, I think, are some of the more fundamental questions that every tax system should be asking itself, regardless of whether you're doing that in the context of devolution or not. But I think these are absolutely critical questions, particularly where there might be pockets of underemployment or higher levels of unemployment generally.

Now, obviously, labour market activity is not just about tax, there are many other considerations, and these approaches should be considered in that broader policy context as well. But, we would certainly see that ensuring that particularly the interaction between any benefits that might be available through other spending channels and the tax system—try to ensure and to reduce [correction: system, we try to reduce] any of the disincentives that you might be creating. That's an incredibly complicated set of assessments to make. We have not undertaken any of those assessments in the Welsh context here, but we think that it's really crucial to be looking at that.

First, on the top income earners, I was surprised, reading the report, to see that a change in the top rate actually induces so little revenue and has so little revenue implications. That's a very interesting result, and that basically made us wonder if possibly many of these high income tax payers in Wales have already been incorporating their activities. So, that's something to take into account, because if a typical self-employed who normally would be taxed on labour income as part of the personal income tax—if these high income self-employed have already incorporated their activities and so take part of their money out of the business for a wage and then the rest as capital income, that might partly explain why you collect so little revenue from these top PIT rates. Because it's not only the 45 per cent rate, but it's also the 40 per cent rate. An important area is these businesses will not go back to self-employment status in response to a change in the top PIT rate. So, that's part of what we're focusing on with the question of top PIT rates.

I think on the other question, on the lower bands, on the 20 per cent rate, I think a very crucial element of further work will have to be the implications both on the revenue side and on the design side of the ageing of your population. We have ourselves a report that we can send you. We have done a report on Slovenia where we calculated the revenue implications of the ageing population in Slovenia. Typically, what people do when they think about ageing is they think, 'Well, that's going to result in additional costs'—health costs, more pensions, and so on—but what typically everyone forgets is that will also have a big impact on the revenues. That is mentioned in the report of our colleagues, because fewer people active in the labour market means fewer people paying tax, with a significant impact on personal income tax revenues. As I understood from the report, Wales will be particularly vulnerable to that. So, I would invite you to actually do the kind of analysis that we've done, to estimate the revenue implications of the ageing population over the next decades in your country.


Okay. In regard to incorporation, if I may, you've made an interesting point that's not been raised today, in terms of that could already be happening in terms of those higher-rate businesses. So, in that regard, and bearing in mind our partial devolution around income tax in Wales and that the levers are still with Westminster majoritively, would it be a consideration for Wales to have full control of national insurance?

I think it's difficult for us to answer that question. What I do think is that, when you set your rate, when you make a change in your rate, you do need to calculate clearly the gap between people who gather income in the form of labour income that is taxed under the personal income tax, and compare that with the overall tax burden of labour and capital income if you're incorporated. Because the wider you make that gap, the stronger the incentive you create for incorporation, and that's a particular issue because you do not have control over the capital income tax rates. So, in that sense, the way you set your PIT rate needs to take into account how the rest of the UK has set its capital income tax rates.

Just to make one additional point, these questions very much go to the constitutional and inter-governmental arrangements in any particular jurisdiction, and we readily recognise that they are very country and region specific and need to be considered in that context. So, we don't want to get into that particular debate, but I think the point we really do want to convey on the incorporation question is that we would expect that a significant number of those higher-income earners that might be inclined to incorporate have already done so. So, what you do or do not do in terms of your partial devolution and the ability that gives you around personal income tax rates is likely to be operating in an environment where that's already built in.

Thank you. And finally, in regard to Wales Governance Centre, you've obviously modelled evidence in regard to the revenue effects of devolved income tax, so could you just briefly explain how that model works and your findings in that regard?

Sure. So, it's a micro-simulation, it basically takes survey data of Welsh taxpayers—a sub-sample of Welsh taxpayers of around 24,000 taxpayers across the income distribution are sampled. So, we take that from the year 2016-17, the latest published data at the time, and then we make assumptions then on their growth and incomes—the growth in the number of taxpayers in Wales over time—and reapply the tax structure, the allowances and reliefs, et cetera, in the rates to get an estimate of what would be raised by the Welsh Government next year. We can then change the rate to get a mechanical effect of keeping the tax base constant, to get a mechanical effect of what a tax-rate change would imply.

In terms of behavioural response, I think it's important to note what we do and do not model in our work. We model responses in terms of the marginal effect of tax rates. So, that is: how much of an extra £1 earned in income is lost in tax. That will affect the individual's incentive to earn more or less—


So, how would you encapsulate that? It may be useful, if it can be shared, if it could come to this committee, if it hasn't already.

Yes. I think we did include it in the paper that we submitted. That sort of response is how many hours a person works, whether or not to seek promotion or a higher-paid job. We include that in the modelling because we've got some evidence at a UK level of what those types of responses would imply for the revenue implications of a tax-rate change.

What we found was, even if we assume very high taxable income elasticity—so, even if we assume that people do change their behaviour quite a lot by international standards, by the standards of comparable studies in other contexts—the Welsh Government is actually quite insulated from the effects of that behavioural change. We would expect a tax increase to cause a drop in earned income across taxpayers, but that mainly results in lower revenues in non-devolved income tax and national insurance contributions, because the Welsh Government only has a portion of income devolved. So, the flip side of that is that, with a tax cut, you're much less likely to claw back the cost of that tax cut in the form of more earned income.

Crucially, what we don't model in our work and what we leave out of that model are responses to average effective tax rates; so, overall, how much income tax you pay as a share of your income. Responses to that will be, importantly in Wales, as the case will be, a migration response. So, we don't include that in our behavioural response. What we do instead is look at the amount of migration response that would need to take place in response to a tax cut to achieve a certain level of revenue change. Because that sort of behavioural change is so uncertain, we don't have anything to hang our hat on, really, in terms of that. We work backwards, in a way. So, that's not included in the behavioural change that we model.

Thank you. I was going to pursue some of that, but I think we've covered that base, so we'll move on to Mike.

We've talked about income tax a lot and taxation, but one of the drivers of people moving and not moving is house prices. While you were talking, I did some quick look-ups using my phone. A one-bedroomed flat in Chelsea is £1.25 million. For that, in Wales, you can buy a mansion. I looked up Bristol and Chepstow. You are struggling to buy a three-bedroomed house in Bristol for £300,000. In Chepstow, you could probably get it for £50,000 or £100,000 less. So, you're talking about income tax and income, but outgoings in terms of house prices do—I think it does, but do you agree—affect mobility, in that it's expensive to move to certain areas.

Absolutely. We've seen anecdotal evidence and some real evidence that the removal of the bridge tolls, for example, has caused quite a significant effect on house prices. I was just looking this week: house prices in 2018 in Torfaen were up 12 per cent, Newport 11 per cent, and Blaenau Gwent 10 per cent. So, Welsh house prices increased 4 per cent in the year to February 2019, far outstripping growth in England and Scotland. So, there is clearly evidence that these non-tax factors are playing a big role.

I suppose when we're looking at this, when we're trying to find clues in this about what big tax-rate changes might signify for migration, I suppose the question is: given that house prices and living costs are so much lower in Wales already, why haven't people moved already? So, I think there is a question that we would have to—. On that basis, we'd expect that quite large tax-rate changes would be needed to trigger some sort of large response. Small changes at the margin—given that we already have these big differentials, these small rate changes are unlikely on their own to trigger a large migration response. 


I'm just wondering whether the OECD would like to add, in terms of the non-tax costs of living, housing costs, et cetera, how those play out.

Yes, absolutely. I'll ask Sean, who has some experience in particular around the housing cost space, but I think one point that we would like to convey is that we think that, perhaps in the discussions that are occurring at the moment, there is perhaps undue weight being given to some of the mobility threats when we consider the range of considerations that policy makers in Wales should be taking into account in determining what are the right rates and right levels of tax to be imposing upon people in the jurisdiction. In the end, the environment in a post-partial devolution context is no different to what every jurisdiction faces, and tax rates impact the extent to which people work, the amount that they work, their willingness to work and getting the rates right to get the right responses is the first set of questions that you should be asking. We do think that it's obviously important to take behavioural responses into account, but we think there is a danger that you could over-emphasise the impact of those. But Sean on housing prices.  

I need just, actually, to start with coming back to a broader point that David made at the beginning, just to say that the overall tax mix is really, really important, and I think looking at personal income tax in isolation may not be the optimal approach, and that really also thinking about the property tax design is really important. I was at a conference hosted by the Wales Treasury last summer, and I know that there's quite a bit of discussion about the devolution of taxes on property, both stamp duties and potentially vacant land taxes, and the design of those taxes. We actually often favour recurrent property taxes, which I think haven't been used so much in the Welsh context, but often we find have nicer properties in terms of both equity and efficiency, as well as stability of the revenue base, and actually also responsiveness to the ageing considerations that Bert mentioned earlier—that it's more robust to ageing, especially if you can design the tax [correction: the property taxes] in a way that addresses, let's say, a land-rich, income-poor household, which can also address certain political economy difficulties. 

In this—. We've been looking at housing in particular, and we've been trying to look across countries about the assignment of responsibilities and the interaction with the supply and demand for housing, and how that interacts with house prices. We have found some interesting findings. They're still fairly tentative, but I think it does really make the broader point that David made, that we should really look also at the service delivery, the overall supply responses and the quality and accessibility of housing for the overall population when thinking about these considerations. Property taxes and the autonomy of setting property taxes do seem to have a good interaction in terms of the devolution process and may help to alleviate some housing pressures that may exist, and that may help to move things in a good direction, even with [correction: even while] perhaps raising more revenue from property taxes.

Thank you very much. We have around about 20 minutes left and I don't think we've covered half the areas that we wish to, so we maybe need to be mindful of that as we proceed, but we'll move on now to Mark. Thank you. 

I just want to ask a bit more about the incorporation issue. My first question is: earlier, you were surprised that the impact at the higher rate was, I think, not higher, and I just wonder if you're convinced those estimates are reliable, because you seem to be going from an assumption that they are to saying, 'Ah, well, that means lots of high-rate taxpayers must have already incorporated and that's why it's not bigger.' Are you confident of the data in that sense completely?


It's a difficult question, of course, because that's more—. I'm pretty confident that our colleagues have done a good analysis using the data. But—. That's difficult to answer. But it is remarkable indeed that any change in the 40 per cent or the 45 per cent rate induces—. And that's purely mechanical, what is included in the report, in figure 4.1, which is not incorporating—if I've understood correctly, that does not incorporate any behavioural change. So, that's purely mechanical—a rate change, what that will induce in terms of revenue. So, that just means that, in that data set, the sample that has been used, indeed, there are not that many people that pay tax at these high rates. Perhaps this calls for, if you're surprised about that as well, to use more—you know, to give colleagues access to the full data that HMRC has on the Welsh income tax payers to do that analysis on the total population, as we have, for instance, done in our case in Slovenia. We had access to the two million personal income tax payers, so we were perfectly able to exactly calculate the impact of a rate change on revenues, assuming no behavioural change. 

But you, if I've understood correctly, are going from that data set to a conclusion that there must already have been a high proportion of people earning more who've incorporated and therefore low sensitivity. Isn't another explanation that, actually, there just isn't very much high income, whether through income that attracts income tax, or income within a company at the higher end in Wales? Why aren't you making that conclusion, rather than concluding, 'Well, a high proportion have already incorporated because of that'?

Look, it may well be a combination of both. We're not stating one to exclude the other. But I think we're just making the point that, wherever there is a significant differential between the statutory corporate income tax rate and the top personal income tax rate, this is an issue. That was the case prior to the devolution, and continues to be the case, potentially to an even greater extent. But it's not an unreasonable assumption, I think, to make, that some of that is already built into the base case that we're looking at here. 

In terms of the quality of the data, we've not looked into that specifically. We're taking the data on the face of it. We have no reason to not believe that that's been the findings of this [correction: to not believe the findings based on this]. 

If I may, I agree that this is indeed a very important question that you raise. Why are there not more taxpayers that pay income tax at the rate of 40 and 45 per cent? That is not—. We are not in a position—we don't have the data or the understanding of your economy to answer that question, but I think that is an important research question that perhaps colleagues could help answering, to see why is that. Perhaps one of the reasons can be that—well, there might be many reasons, but—your rates are perhaps too high, that people are not earning, because of the high tax rates, that they are not wanting to or taking entrepreneurship activities to earn that much income, or it might be that people who have a residence also in England or other sites of the UK, and have a second residence in Wales, do not declare their income in Wales, but rather in other parts of the UK. This raises a lot of questions, which I think are for further study. 

Yes. Or it may be just that Wales is a poor country. If you were to look at Greece or Portugal, you wouldn't come to the conclusions you have with respect to the Welsh tax base, would you?

Well, look, I guess one thing is that, when you look across jurisdictions, there are a couple of factors here. One is that, when you're looking at the top rate, you've got to look at where it kicks in. And that's not the same from one jurisdiction to another. So, when you're talking about Portugal or Greece, one of the things we always look at, and one of the measures that we often apply, is we look at where the threshold for the top rate begins by reference to the average wage—so whether it's one and a half times the average wage, twice the average wage, et cetera. There is much variation across countries, and, in the end, that's a political economy question as to where you think the top income tax rate should kick in. Then of course there is the rate, which varies from country to country. So, I think it is difficult to make comparisons in the abstract—you really need to look at those details in order to have an informed judgment. But it's definitely the case that, where there are standardised rates across a large country, and there are differences in the economic prosperity across different parts of that country, then there will be fewer taxpayers in those areas where there are not the high incomes in the top rate. That stands to reason.


And to complement that, your answer is indeed true that it could be that there are a lot of taxed not very rich people in Wales. It makes you very dependent on the tax rate in your basic tax rate—the 20 per cent rate—that's the one you collect that raises most of your personal income tax revenues. It raises then also the question of whether—the actual number of brackets you have, whether these are then well designed, given the income levels and the economic challenges that Wales is facing; whether the brackets themselves are perfectly aligned with your population.

And your analysis has focused, cross-national, on the higher rates and the incorporation incentive there. Are we safe to assume that's the issue we need to focus on, or is there the potential for this also at the basic rate? On my back-of-the-envelope numbers, at the higher rate, you keep 55 per cent of your income if you incorporate, versus only 50 per cent if self-employed. And then, at the basic rate, you keep 75 per cent if incorporated, versus 73 per cent if self-employed. That doesn't take into account the £2,000 dividend allowance, or the fixed costs of setting up and running the company. But I just wonder: are those such that we don't really have to worry about leakage at the basic rate, or could that actually be a significant risk, particularly looking at the far higher proportion in Wales who are paying that? And, if we were to increase the basic rate tax, would we, for instance, see a lot of self-employed people in construction and related trades incorporate who haven't already who have basic rate incomes?

Look, I think that our assumption, based on the literature—and there is fairly extensive literature in this area—would be that the responsiveness would be less in those brackets than at the higher end. But, of course, behavioural responses can take the form of economic actions, they can also take the form of political actions, and tax policy is a very basic part of the democratic process in most countries as well. So, if people feel as though those tax rates that they may not be able to avoid, because they may not have the skills in the labour market that allow them to necessarily move to another jurisdiction, or they may fear the added costs of moving to a higher cost place, then they may also respond through political rather than economic means. So, I think that we would suggest that it is less likely to be the case at that end, but where the bigger danger is is that that[correction: but the bigger danger is that this] would have adverse impacts in terms of labour market participation. And I think that's what I would be more focused in on—would be whether [correction: not whether] or not your tax system is, at that level, not discouraging [correction: , discouraging] people from being in your jurisdiction, but whether it's discouraging them from working, or working as much as they might otherwise wish to.

And to add on that, for me, the highest priority would be to analyse the impact of the ageing of your population on, in this case then, personal income tax revenues over time, and to see whether you can take measures to bring more people, people that are currently not active in the labour market—whether you can, through your personal income tax or other measures, incentivise them to actually start being active in the labour market, and, at the same time, to ensure that older workers don't flow out of the labour market too soon, to keep all of these workers in the labour market, because, in the longer run, that will be very important to maintain your personal income tax revenues. 


Yes. I have two questions. The first one follows on from what Mark Reckless was talking about. There are those who can incorporate, and they will be people who are working as contractors, consultants, and those who cannot. The First Minister here, other senior people in the public sector, cannot incorporate. So, there is a difference between who can and who can't. So, there'll be a fixed limit on that, won't there? 

That's absolutely the case. The nature of the work that you do means that for some people you will not have a great deal of flexibility in terms of how you structure yourself. That is absolutely the case. And what we find, in many jurisdictions in particular, is senior members of the public service for example, are often people who, at the higher levels, will be in the top tax rate, and positions of that nature allow very little scope for reorganisation. There are, of course, a whole range of activities where there is more flexibility and we observe much more responsiveness clearly in those cases. 

The other question I was going to raise: you talked about the Welsh economy. What we actually do is import retirees and export graduates, and that says very little about our tax system and very much more about our economy, doesn't it?  

Yes, and I think this goes back to what I think is one of the key messages that we would like to convey. As Bert said earlier, much of this discussion, as we see it—and we're on the fringes, so please excuse us if we've misunderstood, but much of this discussion is going to, 'Well, what will be the responses to changes?' And they are important questions, but the more fundamental questions are: for our economy, with all of its particularities, do we have the tax settings right? That's the primary question that I would be asking. And then, when you seek to answer that, of course let's take into account what the potential behavioural responses might be, because the behavioural responses might mean that the answer that you think that you are getting from a particular policy response is not the right one, ultimately, because of those responses. 

Okay, thank you. I want to bring you in—the Wales Governance Centre—because your modelling of a 5p cut to the additional rate shows that a very strong migratory response would be required to significantly impact on the Welsh budget. Now, in view of that, there's an obvious question as to how likely that scenario is to occur, but, as well, what level of divergence do you think would be required for us to see a palpable difference?

We looked at the additional rate taxpayers specifically, because, as we mentioned earlier, it's the income tax bracket where you can imagine the biggest response would be. If you look at the average additional rate taxpayer, a 5p cut in the additional rate would amount to a £10,500 tax cut each year. Obviously, you offset that against the costs of relocation that were mentioned earlier. That would be a big tax cut. The mechanical effect on the Welsh budget would be around £28 million—the hit to the Welsh budget. And even if you assume a large taxable income—[Inaudible.]—a large behavioural change of those already in Wales, it wouldn't have a huge effect on the amount of revenues that you lose from that tax cut.

We do say that you'd require, in order to claw back that lost revenue, about 1,000 average additional rate taxpayers from England to move. If your intention is to increase the size of the Welsh budget by attracting higher earners you'd need around 6,000 additional rate taxpayers to move to Wales to raise £129 million for the Welsh budget. That would be a doubling of the amount of additional rate taxpayers in Wales. So, that would be a—. That's why we—[Inaudible.] In considering the differences in housing costs and living costs already present across the border, that would represent a substantial migration response. How likely that is to occur—you could probably be quite sceptical about that sort of response taking place after a tax cut.


Yes. Okay. Fine. Yes. Well, there's a clear message there, isn't there? Nick.

The same—. Good morning. The same questions that I asked to the previous witnesses that we had, and that's understanding how spillover effects from Welsh Government tax policies could be identified and monitored. What consideration should be given to spillover effects of the Welsh Government policies and what processes could be implemented to ensure that all those effects are identified?

So, in the fiscal framework agreement, there are three effects. There's the mechanical effect of a policy rate change, which will be automatically accounted for. You've got the behavioural change to a policy change, and I think the wording is that that will be accounted for in exceptional circumstances when it's demonstrable and material for either Government—so, the policy of either Government has a direct effect or an indirect effect on the revenues or spending of the other Government. I think it would be quite difficult—. I can't imagine that ever being the case—the two Governments coming to an agreement with each other that a policy rate change by the Welsh Government has caused a demonstrable or material effect. So, as I mentioned earlier, the effect of a 5p cut in the additional rate taxpayers—say that encouraged people to move additional rate taxpayers to move to Wales. Because of the nature of the tax system that we have, they'd still be contributing to the UK Government. They'd still be contributing 35p in the £1 of their additional rate income to the UK Government. So, I think it's very unlikely that you'd see a demonstrable and material effect, and then it comes to the question of how you actually measure that accurately, given the absence of a counter-factual scenario.

And the fact there's no definition of what 'material and demonstrable effect' is in the fiscal framework agreement—this would all have to be negotiated between the two Governments, and, if there's no resolution, the dispute falls. So, again, it's something that's going to have to be worked out over a number of years. But certainly, the spillover effects, if they're not completely direct—

Do you think that was an oversight, not having it in the fiscal framework? Or was it something that wasn't there for a reason?

I think it's just one of those things, given that this is very, very new for the UK, that this is something that is going to have to be worked out over a number of years. We're lucky in a sense that some of the teething issues, shall we say, in Scotland, with the coding of Scottish taxpayers—well, the miscoding of the residences of a number of Scottish taxpayers, for example—some of those things the HMRC has been able to learn from and adopt. But this is going to be something that will subject to review. I know the Scottish fiscal framework is up for review next year. So, depending on what have been the issues there, this could be something that will be worked out over a number of years.

Another example that you have is, if the Welsh Government increases the basic rate of income tax, that has an effect, some effect, on the after-tax income and the eligibility for universal credit, and that will have an effect on the UK Government. So, as we mentioned earlier, it's important to look at it as a tax base as a whole and in terms of social security as well. But I think that effect, of a 1p rise in income tax rates on the basic rate, will probably have quite a small effect. I don't think the Treasury will regard that as a material sum of money and worth the effort of starting a dispute resolution with the Welsh Government. So, it's probably too small an amount. They're probably too small amounts to actually trigger this mechanism in the fiscal framework agreement.

Just very briefly—and, obviously, in regard to that definition, that's something we can come back to in the fiscal framework—could you just flesh that out a little bit more in terms of the—? We've not really touched upon this, in terms of benefits and universal credit access. We've mentioned tax credits earlier, I suppose, but could you just give us a little bit more sight on that?

Yes. So, the eligibility and entitlement for universal credit is based on your after-tax income, so a rise in the basic rate of income tax will affect some universal credit claimants. I think it's important to remember that the personal allowance for personal income tax now is so high in the UK, it's less of an issue than it would have been in 2010, for instance, when it was half the level it is now. So, I think it's less of an issue in terms of that a basic rate increase wouldn't actually affect the post-tax income of benefit claimants.


So, in terms of modelling upon that—I know it's extremely early doors on that—bearing in mind the great propensity in Wales of benefit claimants, you don't think that's currently an issue for us. Surely, it's something we need to be sighted on.

We haven't published anything on it. We've done some tax benefit modelling of an increase in the basic rate. So, the effect on the Welsh Government revenue would be £184 million of a basic rate increase of 1p in the pound. I think the increase in spending on universal credit in the model only goes up by—I think it's less than £10 million. So, it's only a couple of million pounds. So, I don't think it's a massive issue to look at at the moment. Maybe if there was a change in the tax system as a whole it might be an issue further down the line.

Yes. Universal credit is going to be one of those areas—. There is in the fiscal framework accounting for direct spillovers that, quote,

'directly and mechanically exist as a result of a policy decision (before any associated change in behaviour).'

So, a change in, an immediate increase in, universal credit eligibility in Wales that results from an increase in the Welsh rates of income tax would probably fit under that direct spillover. I think the question here, as Guto mentioned, is, for the Treasury, with the size of the UK budget—£700 billion in taxes—whether £10 million or less is going to be enough for the Treasury to pay attention. It may be, but, in terms of the consequences for Wales as a whole, it's relatively—relatively—small. Not to mean, of course, that that doesn't have to be modelled, and we should be doing that too.

Okay. We've surpassed the allocated time. Did you want to—

I'll just follow on with regard to the spillover effects. How could the Welsh Government policies be monitored to mitigate any unforeseen effects in the short term and in the longer term? Are there any other mechanisms that could be put in place? We know that, as you say, the fiscal framework didn't have the—

This is just going to be one of those things that's going to be needing continual monitoring. I've been raising the HMRC's administration of taxpayers, the database, particularly the correct coding of addresses, which started out being quite problematic in Scotland. Wales has been second to the party, if you like—has had the benefits of that experience—and HMRC, of course, will be looking at this. But I think the question is—. This is going to be something that needs continual monitoring, because, as taxpayers move out from England to Wales, are they being picked up as soon as they move in? People particularly with no fixed address or with multiple addresses in the year: is HMRC mobile and flexible enough to pick up these small changes? It doesn't really matter right now, because the block grant adjustment in the first year equalled the offset tax, but, as we go forward, and particularly if we're looking at changes to tax rates, administration is going to be really, really key, I think, given the Welsh Government's contract with HMRC over tax collection. I think that's a role for both the Welsh Government and the Senedd to participate in.

Okay. And can I invite the OECD to respond to that and maybe to leave us with any final, concluding remarks?

Yes, thank you. In terms of the spillovers, one important additional element I would want to put on the table is that, in the report, it is mentioned that, perhaps as part of the mobility and the behavioural assumption, people in the UK could buy a second property in Wales and start declaring their income in Wales rather than currently in the rest of the UK. Now, in reality, often, things are a lot more complex than that because, in income tax systems and as part of treaties, the country that has got the taxing rights depends on where the actual residency is of the taxpayer. So, while I don't have any understanding of how these rules are currently in Wales and in the rest of the UK, but one possible spillover would be that the rest of the UK would make it a lot more strict and difficult to just change residency from one country to the other, and actually to have more tests included, like what is normal practice in OECD countries—that you have to declare your income where you actually live, where you spend most of your time, where your family lives, where your children go to school. So, that idea of buying a second residency and then declaring income in Wales because the tax rates are lower there—that very likely will not work, or it will induce a change in the behaviour of the rest of the UK, so that's a kind of spillover effect, in the sense that the other countries will respond to any change you do in your PIT-rate schedule.


Okay, fine. Well, can I thank you all for your evidence this morning? Can I thank David, Bert and Sean in Paris for joining us, and also Ed and Guto here in Cardiff? We're really grateful for your evidence. It's been really valuable and it'll certainly enrich our deliberations on this particular subject. 

Diolch yn fawr iawn am ymuno â ni. 

Thank you very much for joining us.

Thank you very much for joining us.

Thanks very much. Our pleasure.

5. Cynnig o dan Reol Sefydlog 17.42 i benderfynu gwahardd y cyhoedd o weddill y cyfarfod
5. Motion under Standing Order 17.42 to resolve to exclude the public from the remainder of the meeting


bod y pwyllgor yn penderfynu gwahardd y cyhoedd o weddill y cyfarfod yn unol â Rheol Sefydlog 17.42(vi).


that the committee resolves to exclude the public from the remainder of the meeting in accordance with Standing Order 17.42(vi).

Cynigiwyd y cynnig.

Motion moved.

We'll now move into private session as a committee, so I propose, in accordance with Standing Order 17.42(vi), that the committee resolves to exclude the public from the remainder of the meeting. Are all Members content? Content. Diolch yn fawr. We'll move into private session.

Derbyniwyd y cynnig.

Daeth rhan gyhoeddus y cyfarfod i ben am 11:06.

Motion agreed.

The public part of the meeting ended at 11:06.