Y Pwyllgor Cyllid - Y Bumed Senedd
Finance Committee - Fifth Senedd12/03/2020
Aelodau'r Pwyllgor a oedd yn bresennol
Committee Members in Attendance
|Alun Davies AM|
|Llyr Gruffydd AM||Cadeirydd y Pwyllgor|
|Mark Reckless AM|
|Mike Hedges AM|
|Nick Ramsay AM|
|Rhianon Passmore AM|
|Sian Gwenllian AM|
Y rhai eraill a oedd yn bresennol
Others in Attendance
|Anna Adams||Dirprwy Gyfarwyddwr, Pennaeth Polisi Strategaeth Treth ac Ymgysylltu, Llywodraeth Cymru|
|Deputy Director, Head of Tax Strategy Policy and Engagement, Welsh Government|
|Dr Long Zhou||Myfyriwr ymchwil, Ysgol Fusnes Caerdydd|
|Research student, Cardiff Business School|
|Professor Kent Matthews||Athro Bancio a Chyllid Syr Julian Hodge, Ysgol Fusnes Caerdydd|
|Sir Julian Hodge Professor of Banking and Finance, Cardiff Business School|
|Rebecca Evans AM||Y Gweinidog Cyllid a’r Trefnydd|
|Minister for Finance and Trefnydd|
|Tom Nicholls||Cynghorydd Economaidd, Llywodraeth Cymru|
|Economic Adviser, Welsh Government|
Swyddogion y Senedd a oedd yn bresennol
Senedd Officials in Attendance
|Gemma Gifford||Dirprwy Glerc|
|Georgina Owen||Ail Glerc|
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 09:03.
The meeting began at 09:03.
Bore da a chroeso cynnes ichi i gyd i gyfarfod Pwyllgor Cyllid y Cynulliad Cenedlaethol. Mae'n dda eich gweld chi yma y bore yma. Gaf i nodi, fel arfer, wrth gwrs, fod yna glustffonau ar gael ar gyfer y cyfieithu a hefyd gaf i atgoffa pawb i ddiffodd y sain ar unrhyw ddyfeisiau electronig sydd gyda chi? A gaf i ofyn os oes gan Aelodau unrhyw fuddiannau i'w datgan? Nac oes.
Good morning and a warm welcome to you all to this meeting of the Finance Committee at the National Assembly. It's great to see you here this morning. Could I note, as usual, that there are headsets available for translation, and also could I remind everyone to ensure that any electronic devices are on silent? Could I ask whether Members have any interests to declare? No.
Ymlaen â ni, felly, at yr ail eitem. Mae yna ychydig o bapurau i'w nodi. Gaf i ofyn i'r pwyllgor nodi lythyr gan y Gweinidog Cyllid a'r Trefnydd ynglŷn â threth tir gwag, ar 3 Mawrth 2020?
On we go, therefore, to the second item. There are a few papers to note. Could I ask the committee to note a letter from the Minister for Finance and Trefnydd on the vacant land tax, of 3 March 2020?
Can I say on this matter that she refers to the Exchequer having agreed to move on to the next stage of this process, but perhaps the committee reviews this process before dissolution and provides some advice, possibly, to the next Senedd on the way in which this process is actually operating in practice?
Okay. I see a lot of nodding heads. It's certainly something that we'll try to factor into our schedule for the remainder of this Assembly. Great. Diolch yn fawr iawn.
Felly, pawb yn hapus i nodi'r llythyr. Mae yna ail bapur i'w nodi hefyd, sydd yn llythyr gan y Gweinidog Iechyd a Gwasanaethau Cymdeithasol ar y Bil Iechyd a Gofal Cymdeithasol (Ansawdd ac Ymgysylltu) (Cymru). Mi nodwn ni hwnnw, ynghyd â dwy set o gofnodion cyfarfodydd blaenorol, un ar 27 Chwefror eleni ac un ar 4 Mawrth eleni. Mae'r rheini i gyd wedi eu nodi, felly.
So, everyone's content to note that letter. There is a second paper to note, a letter from the Minister for Health and Social Services on the Health and Social Care (Quality and Engagement) (Wales) Bill. We'll note that, as well as two sets of minutes of meetings held on 27 February and 4 March. Those are noted.
Iawn, wel, ymlaen â ni at yr eitem nesaf, sef, wrth gwrs, i barhau â'n gwaith o dderbyn tystiolaeth ar effaith amrywiadau yn y dreth incwm genedlaethol ac is-genedlaethol. Mae'n bleser gen i groesawu dau o'r tystion fydd yn rhoi tystiolaeth inni y bore yma, sef yr Athro Kent Matthews, sy'n athro bancio a chyllid yn Ysgol Fusnes Caerdydd, a Dr Long Zhou, sy'n fyfyriwr ymchwil, hefyd yn Ysgol Fusnes Caerdydd. Croeso cynnes i'r ddau ohonoch chi. Mi awn ni'n syth mewn i gwestiynau, os ydy hynny'n iawn, ac felly mi wnaf i gychwyn jest gyda cwestiwn bach cyffredinol i ofyn i chi ymhelaethu ychydig ar y model economaidd rŷch chi wedi ei ddatblygu i fesur effaith amrywiadau mewn trethi incwm ar economi Cymru, a beth fuasech chi'n ei ddweud yw manteision hwnnw o'i gymharu efallai â modelau eraill.
So, on we go, therefore, to the next item, which is to continue our work collecting evidence on the impact of variations in national and sub-national income tax. It's my pleasure to welcome two of the witnesses who will be giving us evidence this morning: Professor Kent Matthews, who is a banking and finance professor at Cardiff Business School, and Dr Long Zhou who is a research student, also at Cardiff Business School. A warm welcome to both of you. We'll go straight into questions, if that's okay, and I'll start just with a general question to ask you to elaborate a little on the economic model that you've developed to measure the impact of variations in income tax on the Welsh economy, and what would you say are its advantages compared to other models.
Thank you very much, Mr Chairman. The model we have is built on an input-output framework that we've worked with the Welsh Government on in the past. It's embedded in a wider what's called a computable general equilibrium model, which you have had evidence on before. So, what's different is that we can have this input-output framework inside the model that enables us to look at the allocation of resources as they affect from the macro economy into the various sectors. So, we have 21 sectors—and I'll allow Dr Long here to elaborate them if you want—but this allows us to look at how its income flows are allocated over these sectors. It does enable us to look at the short run, which is very restrictive because it's input-output and lots of factors are fixed, but we also look at the long run, where we allow various factors to be relaxed, particularly things like labour movement, migration and capital movements.
Okay. Does Dr Long want to elaborate, then?
Yes. I think, because this CGE model is based on a kind of extension of input-output models—so, compared to a macroeconometric model especially, it is based on national accounting principles, so it has the advantage of input-output modelling of linking multiple sectors, and you can observe the interactions between all the sectors. But, for the traditional macroeconometric models, you tend to have much, much fewer sectors inside. So, I think it is the most advantageous. And also another advantage is, compared to the partial equilibrium models, it is general equilibrium, so it focuses both on the supply side and the demand side. Ultimately, it will deliver new equilibrium prices and quantities compared to the base equilibrium, but, for the partial equilibrium, it is very difficult—it's just focused on a particular part of the market or some particular variables.
Okay. We're not going to get too technical, I hope, because there's a limit to how far I can be with you on that journey, but there we are. Certainly, that's very useful. You mentioned the sectors. Could you just elaborate a little bit then about how it looks at the private and the public sectors, and I imagine distinct other sectors in the economy as well?
The sectors basically are just based on the standard industry criteria to standard 7 categorisation [correction: on the standard industry classification 2007]. So, we don't disaggregate these sectors further, but just base them on the basic categorisation level, SIC, to standard 7 [correction: SIC 2007]. So, we have agriculture, mining, quarrying and manufacturing, then energy. Then we have construction, then wholesale and retail, just like a traditional input-output table. For the public sector, we have public administration and defence services, we have education services, human health and social work services. Yes, basically, that's—
Okay. Mike wants to come in.
How is it different from the ONS distribution? The ONS have got a list of different areas, haven't they, which they use when they publish. How does it differ from that?
ONS. The ONS, they've got—. The ONS is very similar to what you said. Are you exactly the same as the ONS, or—
Yes. Yes. Because this model is based on the so-called social accounting matrix. So, that matrix is—. We have all the 21 sectors, but all the 21 sectors have to be gross value added incorporated.
Okay. So, tell us a little about how this model can account for behavioural effects.
Okay. If I may talk about that, yes, the model at the—. If you like, think of the input-output as being inside an egg, and the outside is the bigger model, the CGE model. The CGE model is what drives this interaction between the public and private sectors, and it works through, first, the private sector responding to things like tax changes, or responses to external stimulus. And that works in the public sector, because, in our model, we assume a very strong, balanced budget. So, as tax revenues change because of expenditure and income changing, that then affects Government spending, and that then has a second-round effect in our model. So, in many ways, it's a very traditional Keynesian-type framework. What differs from a Keynesian-type framework, where everything is demand driven—in the short run, we assume that all the resources are fixed. So, there's no migration, there's no movement of labour, there's no movement of capital. In the medium run, we allow labour to work harder—indigenous labour, that is, people who are here in Wales. In the long run, we allow for complete migration, so that the factor prices—that's the price of labour net earnings between England and Wales—are equalized. And the same with the return on capital. So, those are the three stages you can think about.
Okay. Okay. That's interesting. Rhianon.
In regard to what you've just stated, the short, medium and long term-factors and assumptions that this model is based on, what I'm trying to understand is—and excuse me as a lay person trying to understand this, but I didn't understand it the last time around—is this an accepted model and framework, or is this a new model that you are testing out on these assumptions? Because, obviously, in terms of them being assumptions, what I suppose I want reassurance on is that this is an accepted model in terms of the methodology, and therefore we can have greater confidence in its conclusion.
Well, I can answer that. Yes, it is an accepted methodology. And that's different from saying, 'Is it an accepted model?' Because the methodology is accepted, and that is, in a country where there is very little data that we can use for econometric testing, we use what's called a CGE, where we construct a hypothetical or theoretical model and we then put parameters, values, for these. Now, where do these come from? Well, some of them come from microeconomic studies, where people have looked at particular sectors. Some of them come from where people have done similar work—say, in Scotland—where we can import that model, and use those parameters. Others are pretty much guesswork—they're called assumptions. So, we test the robustness of a model by varying those assumptions. Where there's a lot of uncertainty about a certain parameter, that's where you need to put in lots of different values to see what the outcome is. There are some parameters that we are more certain on than others; there are others where we have to keep on testing by really varying them.
Do you have anything to add to that, Long?
Alun wants to pick up.
Yes. I'm fascinated by your work. Of course, you can test a model and the methodology in all sorts of different ways. And perhaps one way, of course, is to look at real-life examples, where policy makers have made interventions into an economy, or into a sector, to actually—with particular objectives in place. Are there any places, any states, that you could point us to where this sort of methodology has given pointers, or has explained, or has reflected real-life examples—a more comparative study, rather than simply a theoretical study?
This particular model, I can say, is virgin, in the sense that it's the first one that's been built in Wales. I don't think there are—. There is a Scottish model, of course, which takes very much the same kind of framework, and there is evidence there, where parameters have been estimated for certain sectors. Sensitive parameters—one has to really look at what my colleague here called 'partial models', the sort of thing that you're talking about, making observations and trying to get evidence. The trouble with that is that it's looking at a specific thing. I'll give you an example. Studies have been done about tax cuts during the Thatcher period, and they found that there were quite strong supply-side effects for the tax cut at the high end. Now, that's partial. It's partial in the sense that you're only looking at those people, their expenditure, their supply of labour; lots of other things were going on in the world at the same time. And we use those parameters guardedly, but these are the sorts of evidence that we can use, and these are the sorts of evidence we would put in our model.
Okay, thank you. That's a good basis for us to build upon there. Thank you for that. Siân.
I symud yn ôl at y cyffredinol, mewn ffordd, ydych chi'n teimlo y byddai amrywiadau treth tymor byr ac amrywiadau hirdymor, ar draws ffin Cymru a Lloegr, yn annog gwahanol ymatebion o ran y ffordd mae trethdalwyr yn ymddwyn?
To move back to the general, in a way, do you feel that short-term and long-term variations across the Wales-England border would encourage different behavioural responses by taxpayers?
I do. In the short term, of course, it may not make a great deal of difference. I have to caution you that, at the moment, we've been looking only from the Welsh perspective, and any modelling that we look at, we make the broad assumption that there's free movement of labour in the long run, so that labour can move from Bristol and Bath over to Cardiff; that's what we would assume. But, of course, we need to look at other studies, and there have been studies that looked at migration, with actual data. With this actual data, what's not been looked at is the response of English taxpayers, because what we're looking at here is the effect on Wales.
In the medium run, there will be more labour supply in Wales, as people work longer hours and they take on unemployed resources. But in the long run, we expect labour to move into Wales. Now, of course, that's an assumption. We're assuming that, at some point in time, net earnings will be equalised from across the border over in Wales. No-one has studied what will be the actual response of English taxpayers to, say, a cut in Welsh taxes, to enable them to know what their response is. We don't expect the same response from people who are located in Wales, as people who are located outside Wales, and that study certainly hasn't been done.
Ond mae eich model chi yn dangos y byddai yna effaith penodol yn digwydd o gymryd yr ymfudo i mewn i ystyriaeth. O ddefnyddio'ch model chi, rydych chi'n dod i'r casgliadau rydych chi newydd ddisgrifio, mewn ffordd.
But your model does show that there would be a specific impact taking place, taking migration into consideration. That is, using your model, you come to the conclusions that you've just described, in a way.
Absolutely, we are making that as an assumption. We are making the assumption that people will move. Of course, that assumption is a valid one for a very long run. What our model doesn't say is how long does it take to get to the long run, and that's where studies of responsiveness to tax changes in a particular region really have to look at actual migration from one region to another.
Ond rydych chi'n sôn am gynnydd a thoriadau. Beth ydy'r gwahaniaeth petai'r trethi yn cael eu torri a beth ydy'r gwahaniaeth efo'r trethi yn cael eu cynyddu?
But you are talking about an increase and the opposite. So, what's the difference between what would happen if taxes were cut and if taxes were increased?
Again, in the short run, I would expect no movement in labour, but, yes, I would expect some movement, particularly at the high-value end of the labour market structure, because the opportunities for people at that level are much greater than at the low-income levels. So, I would expect a migration at some point, as people move across the border for similar kinds of jobs, or even further afield.
Roeddech chi'n sôn bod prinder data yn gwneud eich gwaith chi'n anodd. Sut fyddai modd gwella ar y sefyllfa yna? Pa fath o ddata fyddai'n helpu?
You mentioned that a lack of data made your work more difficult. So, how would it be possible to improve that situation? What kind of data would assist you?
I'll tell you what we need for the model first, and then I'll move on to ask my colleague to embellish on that. We would certainly need to look more closely at migration statistics. We would also need to look at trade statistics within the net flow of trade in Wales and the rest of the world. These two are most important at this stage. We would also need to look at things like what's available in general household survey data about income distribution and tax paying bands. But I think my colleague here will probably be able to tell you more about that.
Yes. In fact, because this model was conducted based on a single level of households, it does not consider the different income bands of the households. So, one of the reasons we chose to do this was because of the lack of data, because currently the Welsh data for the household—there is only one type of household and there is only one type of consumption pattern, in terms of different consumption goods. But if we want to include more income bands at the household we have to include the consumption pattern for each of the income bands, for each income group of the household, and also for the income distribution, including labour remuneration and also the company profits, like bonuses—how these are distributed on different income levels of the households. So, with these data we can consider to feed into different levels of income groups of households. And also for the trades that this takes, we have the export and import from and into the rest of the world, but we don't have the statistics to and from the rest of the UK.
Yes, of course.
And that plays a very important role in constructing input-output tables or social accounting matrix, and also it is an important role for the computable general equilibrium model. Sometimes, it runs not very stable, because these parts of the trade are very large. Also, for the migration statistics, we want the migration, not only for labour, but also for the capital, and also for both of the factors that are purely induced by the tax rate that you're interested in, not just for no reason or for non-tax reasons. That's very important to add some dynamics into the model so we can trace the stage by stage changes in the long run and how long it takes to reach.
Mae Rhianon eisiau dod mewn ar hwn.
Rhianon wants to come in on this point.
You've partially answered my query. With regard to the importance of the migration labour trade detail and with regard to the importance of a number of different factors that you've just outlined—I cook, so if I don't put the right ingredients into my cake, it won't work at the end of it—so how do you satisfy us that the assumptions that you've come up with, based on the lack of data, in a sense, that you've already referenced—? How do we approach this model in that regard, because if you're very limited in what you feed into the sausage machine, how can you assure us that what you've said to us, as the assumption at the end or conclusion, is actually of any worth?
Okay. With any model there's going to be a large amount of uncertainty. We can use various things, like the fact that numbers have to add up, balance sheets have to add up, income statements have to add up. So, we know bits with great confidence, but the trouble is that the bits we don't know are still quite large, and that's where the net trade figures come in. So, to give you an example, we know what the economic theory will tell us about the direction of various things. So, when we stimulate a tax cut, we know, ultimately, that this will have an effect on expenditure, income, movement of labour, capital, but how does that translate into actual output? So, in terms of your cooking, what does it taste like at the end? [Laughter.] Well—
We'll reserve judgment on that. [Laughter.]
Sorry to use that as an example. In our model, we find that tax cuts have a very significant, but still fairly numerically small, effect for certain tax cuts, and the reason for that is because the parts of uncertainty are to do with these trade flows. We assume that we have a high marginal propensity to import. So, when you're doing you're cooking, you need to buy all this stuff for your recipe from England, and that actually has a negative effect on our GVA. Our consumption goes up. So, households are better off, they're cooking more, but they are buying their ingredients from England, and so the total output doesn't increase by as much. Now, that's an assumption, because we don't know enough about those trade flows. And the other thing is that everything that we do know about, we put in, but all the things we don’t know about, which includes uncertainty and measurement error, go into this trade flow. So, that's the residual.
Okay, thank you very much for that. Mike.
I'd like to talk about non-tax factors. My understanding is—you tell me if I'm wrong in any of the things I say here—that the average wage in Wales is about £25,000, and a 1 per cent reduction in tax would be worth about £20 per person, per month, and that would pay for an extra £10,000 on a mortgage. House prices seem to be the driver, and we saw it in the 1980s, didn't we? When taxes were cut, house prices went up. We're very dependent on house prices. We also know that the difference in council tax between areas that are fairly close, on a band D property, are substantial. It can be up to £40 a month or so.
But we've got the other side of it: what you get for your £200,000 in Blaenau Gwent and what you get for your £200,000 in Usk—if you can get something for £200,000 in Usk—are entirely different. So, what I'm saying is, the non-tax effects—. And also, people like to live close to family because they get the benefits of free childcare. So, you have these non-tax things.
If you take a model where everything is only based on people making decisions on tax, then I can understand it. But there are these other non-tax things, which I've just mentioned, which have a huge effect. Can I just finish the question on this? We've seen the effect of doing away with the tolls on the Severn bridge. We've seen a huge increase in house prices in those areas that are close to the bridge, which, in many cases, is substantially more than a 1p or 2p cut in income tax.
I agree with the sentiment of what you're saying, because there are two ways to think about this. One is that in a general equilibrium model, we allow everything to change in response to a change, say, in a tax cut. So, we allow house prices to change. Now, of course, we don't have the level of disaggregation that you're thinking about, because house prices will differ from Cardiff and other parts of Wales, but we have a general feeling about house prices. In our model, that will change.
But the other things that you talk about—the non-tax factors, the quality of a place, the desire to be near your family—these don't change, and so we're looking, really, at marginal effects, i.e. keeping all those exactly the same. What's the marginal effect of a change in tax? We're not looking to change everything else, and, of course, if you were able to pull a lever and change those non-tax factors, we could build those into the model, but those are not measurable.
You may not agree with me, but I actually think that house prices are the huge driver of where people live, much more than marginal rates of tax.
And jobs and family. But you can commute a reasonable—I come in from Swansea every day; you can commute a reasonable distance, depending on your willingness to commute, but that also adds cost, doesn't it? So, if you look at people who are prepared—and we know they are—to live in Monmouthshire, in Newport and in Torfaen, and commute into greater Bristol, because house prices in those areas are—. Even the expensive parts of Nick's Monmouth are cheaper than the expensive parts of Bristol.
I completely agree, but that's the great thing about the general equilibrium model—we allow house prices to move. When I worked in London as a professional economist, I was surprised that, in my office, people would come in from Northampton to spend the day in London. We do allow prices to change, and that affects incentives.
And the final point on this is, of course, the ability to travel. The ease of travel. House prices in both Southampton and Bedford went up substantially, didn't they, when the rail links into London improved dramatically? The Bedford-St Pancras line increased house prices substantially in Bedford, and throughout Southampton, with the Southampton into London line. So, you have these things as well, which play a major part, don't they?
I don't disagree, but those are not things that are part of this model.
And another point—. The last point I want to raise is on—. We're going to keep teachers' pay the same and NHS pensions the same in Wales and England. Are there dangers, if there are substantial changes in income tax, that somebody can decide whether to live in Chester or Wrexham and work in exactly the same place where the people might actually decide, on those grounds, especially those at levels of the highest rate of income tax, that, actually, moving to one or the other is beneficial? Though, again, I put the house price variation in there.
The short answer is 'yes'.
Okay, thank you very much. Alun.
I think that perhaps we should be having a psychological conversation rather than an economic one. [Laughter.]
In terms of where we are at the moment, the model that you've used here has introduced a 5 per cent cut in order to stimulate the impact that you've tried to measure, and in the model, as I understand it, that's led to a greater reduction in revenue in the short and medium terms. The proposition that we're testing, of course, that we're seeking to understand, is that, were one to introduce such a cut to income tax levels, then that would stimulate economic activity to the extent that it would actually lead to increased income tax take because of stimulating the market and stimulating the economy. And that's the proposition that we're largely seeking to test. You don't believe that; your model doesn't show that.
That's right. Yes, the model doesn't show that because in our model, we have a very strong balanced budget assumption. Now, if you're doing—. If you think about the logic of what the purpose of the tax cut is, if the purpose of the tax cut is to stimulate growth for the longer term, the obvious way to meet that gap of funding is through borrowing. That's how the UK Government would behave and that's how local governments used to behave when we had local authority bonds, because they could then smooth that out. But, in our model, following the strict rules as to if you have a tax cut and the revenue falls, Government spending has to fall pari passu—one for one. Now, that has a very strong second-round effect, because Wales has a much stronger Government sector than the rest of the United Kingdom. So, that immediate effect leads to a fall in GVA.
Of course, as we relax things over the longer term and we get migration as these tax cuts come in to bite on labour movement, we have stronger effects. My colleague will, perhaps, say something about this because what we find is that small changes in the tax rate don't have such a big effect because that's what we're used to in models—to look at small changes. We don't like to do huge shocks to a model because the data and experience that we have doesn't really warrant that kind of thing. But, if we were bold and went for much larger tax cuts, we'd do that, actually; larger effects and proportionally more in the longer run than that small effect. Would you like to come in?
Before Dr Long comes in, I think, Nick, you wanted to pick up on something and maybe we could address them both at the same time.
Yes, thanks. You've just said something that's really fundamental in terms of the changes to the tax rates. You said that because of the way the Welsh economy is structured and the greater proportion of Government jobs, then if you drop tax rates to, hopefully, in the future, stimulate economic growth, actually, because of the way that the Welsh economy is structured, that may not necessarily deliver the same sort of dividends that it would if it was done across the UK, but if you have a right-wing Government on a UK basis that wants to reduce the state, then, actually, they're reducing less than they would be in Wales. So, does this go back to a point that I think was made by an economist some time back when we began on this tax journey in Wales, that, actually, it is much easier to leave the tax rates where they are in Wales than it is to either reduce them or increase them than it would be on a UK basis?
You're asking me to really make a political judgment here rather than what's in the model. If you're asking me, 'Is it easier?' yes, it probably is easier, but, 'Is it better?' I'm not so sure. My own personal prejudice is that Wales has got an amazing opportunity here to be able to experiment with tax changes.
When you say 'experiment', do you mean up or down?
Well, down. [Laughter.] Because, if one looks, for instance, at how Ireland was able to develop—and this is nothing to do with the model, by the way, this is just my view—part of it was to do with changes in taxation, particularly corporation tax. Now, that had the effect of bringing capital in and with it, then, high-quality labour. I don't see any advantages that Ireland has that Wales doesn't have, except, of course, that it's an independent Government—but, subject to that. I think there is the potential there for Wales to do something that would be actually beneficial in the long run. Now, that requires some bold thinking about what you do in the short run, because while you have the balanced budget, that makes it very difficult politically to be able to deliver that. But, if you had the ability to issue bonds—Liverpool used to do that in the 1980s when I lived there—you can actually consumption smooth and then get around that problem.
Coming from the centre-right of politics, there are a lot of things that you say there that resonate with me, and there is a certain sense with these powers of excitement of what you could do to stimulate the Welsh economy. But, my concern is that, because of the different nature of the Welsh economy and if we're not planning some fundamental change to that, which I don't believe we are and that will be very difficult, then, yes, there are opportunities but there are enormous risks as well, aren't there, of reducing taxation rates on the basis in the future of some sort of stimulus, but at the same time, what could the costs be? As I say, I like some of the things that you're saying, but I'm just thinking, 'God, this is a leap, isn't it?' and I'm just not sure—
Well, we're not writing manifestos today but, certainly—[Laughter.] As soon as politics get involved, you can't keep them quiet. Very briefly then, Mike, because I am conscious that we have about 20-odd minutes left.
Wales can issue bonds, but it counts against our borrowing ability so, we've got the capacity but it doesn't give you any extra money than borrowing at a cheaper rate from the Government.
Fine. Statement made. Sorry, Dr Long, we sort of went off on a tangent.
This model is based on a 5p cut—
It's based on what?
A 5p cut, so 5p per pound. So, after that, because we just want to make this evidence simple so, we just report a 5p cut. But later, we have conducted some larger shock. So, for example, if we do a 20p per pound cut, so four times larger than that, we have a 1.2 per cent rise of gross value added and also a 2.6 per cent rise in employment. Also, because we don't have an explicit migration mechanism in this model, we just let the price go back to the first initial equilibrium so, release the labour and capital stock. So, this rise in employment, we don't know where it will come from; it could come from the rest of the UK or even the rest of the world.
Mark wants to pick up on that.
Yes. I'm concerned about that divergence between 1.2 per cent projected increase in GVA, and the 2.6 per cent increase in employment. I mean, that suggests that it's a fairly productivity sapping impact; why would that be?
Yes, I think you're right. The implication of that is that it's actually low productive labour, isn't it? Because, you're looking at a faster increase in employment than in actual output. But, don't forget that in the model here, we've got this high marginal propensity to import and that actually depresses GVA. If you had looked at domestic output rather than total output, you'd probably have a much higher productivity growth. So, don't necessarily go on these figures. There are two things that I could say about these figures: one is that it's an aggregate figure that takes into account imports and our assumptions about imports and two, I don't think you should read too much into the actual numbers that are coming out here; it's the direction that I think is important, and what we find is that it's positive. We need much more research to actually give greater precision to those numbers. But what's interesting is that you're getting proportionately more for a 10p cut than for a 5p cut, so there is a certain amount of not—asymmetry or non-linearity, if you think; it's not just multiplying the 5 per cent by two to get to the 10p. That does surprise me, but that does tell me that the model does have a certain amount of non-linearity or threshold: a threshold effect is one where you need to get a bit of a biff, before it actually begins to have a real effect, and the model is kind of exhibiting that.
Okay, good. Happy, Alun?
Could I just—[Inaudible.]
Go on, then.
In terms of—[Inaudible.] I'm interested, therefore, if we increased taxes, if we said, 'Put 5p on tax', rather than reducing it, do you, from your work, what impact would you anticipate that to have?
Again, two things. It really depends very much on what you do with that revenue, okay? If it's purely going to fuel Government consumption, maintain a standard of living or whatever, then, in the longer run, I expect that to be negative, because those increasing taxes will lead to a migration of labour—outward migration.
If it's a short-term effect, to be able to, say, make things better for people to live in Wales, it will attract people in here, i.e. infrastructure spending, spending on schools, or whatever you think that people are interested in, that could have an effect on the supply side. Now, we don't have a supply side in our model; what we do is we have a short-run response, and we assume that, in the long run, that supply is infinite, in the sense that if people can earn more in Wales, they'll just travel to Wales. If there are better investment returns in Wales, they will invest in Wales. What we don't have is actual responses. Now, that response—to the gentleman over here, who talked about the non-tax factors—those are the things that will matter also, depending on what you do with that revenue. So, the answer is, really: I don't know. If you were to just simply carry on as you are before, it'll be negative.
But in the short-term, of course, you can employ more doctors, you can employ more professors or you can employ more teachers, or what have you, and you know, those people are paying an income tax, and they're contributing to a very different sort of society as well.
That short-run effect is taken into account in the sense that we do have an increase in Government spending. The medium run will tell us more about that, because in the short run, there are no more doctors; you just make the doctors work harder here. It's only the long run that you can attract more doctors, right? Just having vacancies alone in our model doesn't do anything.
We're going to have to make progress now. Thank you very much for this. It's very stimulating, very interesting. The more we get into it, the more we want to ask questions. So, Rhianon.
Yes. I mean, it's been absolutely fascinating, so thank you. It would be very interesting, going back to where I started, if we had that more longitudinal, more in-depth study and obviously, we need the data around that and I think that there is an issue there for us all.
You've already mentioned that obviously, any larger-scale significant change to the tax rate would give you a larger behavioural response, and vice versa, but with regard to the fact that from your data and from what you've already said, a 5p cut would cause less revenue for Welsh Government, and we all know that Welsh Government would need as much as possible in the short run, I'm interested in that projection for the 10 per cent cut in terms of what that would then mean.
So, would it be more positive in terms of the trajectory that you talk about for Government, in a sense, to say, 'We're going to go for the 5 per cent', or the 10 per cent? I know that may sound like a political question for you in terms of a tax cut, but with regard to what you've just said, I think it's really important that we underscore your view on that. Because as you say, it's a positive trajectory, but we'll have less revenue.
Yes. My view is that it is a positive trajectory, but there's huge uncertainty around it. This is a central—. It's not a prediction; it's a central simulation. The big uncertainty is about what happens to GVA. If GVA is important, then of course, a 10p cut is not going to bring us the same kind of benefits that you'd think you'd want, and that's largely because of the import propensity in our assumptions. On the other hand, it's very clear on what's happening to domestic demand. In the model, household consumption and investment by firms increased dramatically, even for a 5p cut. So, it's just that you have to buy your ingredients from outside the border, but actual productivity and actual household welfare increases. I think, even for a 5p cut, it's almost nearly a 0.5 per cent increase in household consumption. Am I correct?
And possibly something like a 0.75 per cent increase in investment. These are, I think, quite interesting figures. So, a 10p cut will give you more of that, and possibly even proportionately more, but of course, in terms of GVA, it may not be that great simply because of our imports. Now, I think Long has some numbers here if you wish to—
Very briefly, but before you do—I'll come to that very briefly, if I may. With regard to Welsh Government strategy and direction in terms of procuring locally and in an economic contract that steers business into buying locally, in terms of your importing from over the border in England wouldn't necessarily be, from a Government public sector buying corporate ability, so much of an issue, but you're saying this would be household.
Okay. Right. Sorry, did you want to continue with your figures?
With the 5p cut, in the short run, household consumption rises by 0.5 per cent and, in the long run, it's 0.6 per cent. Also, for investment, which is gross fixed capital formation, again, it is little bit more than 0.5 per cent in the short run and, in the long run, it is almost 1 per cent.
Okay. And balanced against that, you would have less revenue.
Thank you. Mark.
How difficult are the challenges respectively for the short-term versus the long-term modelling? Which would you be more confident in?
Well, I think this is a limitation of the model, because the model is static. It's like a snapshot. It's looking at a snapshot, the short run of what is, and then a snapshot, short run, of what could be. What we don't have is actual—. We need data to model the dynamics—how we get from here to there. That's what the politicians are interested in: how long does it take to get from here to there? Unfortunately, at this stage, we are not able to give you anything definitive on that.
Well, we will have to wait, I think, at least until next year to see what, if any, experimenting the electorate would like to be undertaken. Can I, though, try to clarify something a little more about the model? You mentioned the migration flows from Wales to England and England to Wales. I'm not sure if I understood it or not, but I thought you were saying that you measured what happened to people in Wales. Were we also estimating what the migration was from England to Wales? Surely that's really significant. Is that there?
We're not modelling that, no. We're assuming that the labour will come from somewhere. Whether it comes from other parts of the world or from England, the labour comes from somewhere so that there is no need for any further migration, because in some ways there is an equalisation of net earnings across borders.
Okay. How does that happen without a housing supply response? You're saying this is in the short term.
The model is general equilibrium, so it takes into account investment in 21 different sectors. I'm presuming housing is somewhere in those sectors as well. So, yes, it actually allows for that interaction. That's the benefit of that model: it actually doesn't assume that it's going to happen, we can actually look at the flows. Now, if those flows don't meet demand, we allow prices to change. So, house prices will change, rentals will change. That will create hardships in the short run, but it also creates incentives for investment and for building and all the other things that you need for a model to actually capture all these things.
You mentioned Ireland earlier and I just wonder, perhaps compared to a more normal approach, whether this assumption of, 'Well the supply just comes and it's from the rest of the world', and this and that, and there's quite a lot of that that can happen compared to the domestic economy, so it's a valid assumption—did you consider Ireland specifically in terms of making that assumption for Wales in terms of its appropriateness?
No, we didn't. This is a standard assumption in this kind of model building, so Ireland just happened to be something that's a good case study, but, no, it's actually a modelling assumption that we use, and it's an assumption that the Scottish model uses as well.
Okay, thank you.
Diolch. The committee's heard that given that the Welsh rates of income tax enable the Welsh Government to benefit from the full mechanical effects of the tax rate change—that's the evidence we've heard, anyway—but only a proportion of the behavioural effect, then it's highly likely that any increases in tax rates, even at the additional level, would increase Welsh Government tax revenues. Do you have any view on that?
Well, yes, in the short run, if the tax increase with the full benefit of the mechanical effect, which is assuming that nothing else changes, but just the tax change, and that income and expenditure don't change, or, rather, income doesn't change in the short run, then, yes, you'd get that benefit. But regarding the behavioural change, there's a huge amount of uncertainty here, because we really don't have enough information. What we have are assumptions. I believe these assumptions are valid in the sense that I don't expect behaviour not to change. I mean, I do expect behaviour to change. What I'm uncertain about is how much it will change, i.e., to use an economic term, elasticity. If I expect a good strong response, I expect it to be a very elastic response. I expect labour to respond very well to it. If you believe that labour is very inelastic in its response to tax changes, then you could get the full benefit of taxing people more, and they will continue to work just as much. I just don't believe it. The evidence of the work with GHS data for the Thatcher tax cuts shows exactly the opposite. There's some work that's been done for the Baltic states, where there has been flat taxes, and these, too, show some revenue increases. More has to be done on that, but that is an area that we would want to look at, for instance, to experiment: what sorts of taxes would give you optimal revenue.
So, you're basically saying that even with a base the size of Wales and an economy like we've got, punitive tax rates would actually have a behavioural change in the medium term?
Punitive—. I think even small tax changes will have punitive effects, never mind punitive ones. But in our model, there are—. We make through our assumptions—
I should have said more of an effect than they would on the UK level.
Less of an effect.
Yes. Well, that's what I believe. The model would say that, but then the model is a bit—. You know, what you put into it is what you get out. So, in our model, we have positive effects from tax cuts, so the more tax cuts you get, you're going to get more positive effects. How much is a question of something that we need to investigate more strongly.
What plans are there for refining the model, and what are your timescale for finalising it?
Okay, let me start with that, and then pass on to Long. We need to get data on trade flows, so that we can look at exactly how responsive is net trade to domestic demand in Wales. We need to get migration data and labour supply data. I would hope to do meta-studies with labour economists from Swansea or Bangor, but mostly Swansea, who'd be able to tell us about studies that have been done for other countries and other small states where they've done tax experiments of this kind to look at these responses, to see how we can push that forward here in Wales.
And finally, we want to get a handle on the idea of dynamics: how long does it take to get from here to there? That is obviously not something that's going to be easily done, but we can look at where people have estimated dynamic models of this type for regions in Spain, to see what information we can get from that and try to simulate those.
I'll pass you on to Long, if there is anything else that I've missed.
Before you do that, could I just ask—we have the Minister in our next session this morning—what could the Welsh Government do to help you, in terms of getting some of the data that you need to do some of this, then?
Well, I believe the trade survey is on its way and that will be of great value. At this moment, I'm thinking very much about what kind of information you have about income structure and tax paying and migration. These are the most important at the moment. Labour economists will be able to give you a much more definitive answer to this, in terms of micro studies, because you can't—. Wales is never going to have the kind of data that we have in the United Kingdom to build an econometric model, but we can build bits of it and that will be very important, and labour studies are important. So, I do believe that the sort of information you're getting in the general household survey on a regional basis would be interesting.
Sorry, Long, did you want to add anything?
Nothing much, but I just want to say briefly about the structure of the model, because this model is just about from one equilibrium to another one. It is not about the forecast, it is about how the economy is starting—. For example, my model is based on the 2013 data. So, from the first equilibrium of 2013 to an alternative one, given the shock—. So, given the shock, what the economy requires to satisfy what the shock needs to actually reach another equilibrium—that is how the model works. And so, that's why we can, in the long run, just release the labour and capital stock, although we don't know where it comes from. If we can have the migration statistic it will be very helpful.
In terms of what Professor Matthews mentioned—
Sorry to interrupt—are you having problems with the data that's available, because we often hear on committee how there's an issue with the availability of data in Wales? I know that HMRC didn't necessarily keep all data specific to Wales before it became necessary.
I think this data exists, having worked with HMRC before, but it's a question of confidentiality in getting that data. But I don't see any problem, hypothetically, in getting that data. I think it's certainly—. It's feasible, should I say? It's just a question of getting through protocols and all the sort of stuff that you—
How many of you are working on this overall? [Laughter.] It's a fascinating area at a fascinating time and you're obviously doing something quite innovative, but—
Well, the thing is, it's not my main job either. I do other things. I'm a professor of banking, not a professor of taxation.
Can I just ask one brief question, before we conclude, because I am conscious that we're more or less out of time? Some of the evidence we've received has suggested that under certain scenarios there'd be behavioural effects, in terms of self-employed people maybe incorporating their business, businesspeople drawing their income in other ways, which would have an effect on the Welsh rate of income tax. Is that something that is accommodated in your model or could be or—?
It could be. The more complex a tax system is, the more incentives you provide for avoidance. In some way, that's the simplicity and beauty of flat tax. It may look very regressive, but it has positive effects, in terms of revenue generation. And so, that's the sort of thing that we could simulate in the model, because, remember, in this model we're not using the real world, we are giving a representation of what we think the real world could be—it's a model. So, we could model precisely flat taxes, high taxes, low taxes. It just depends on what assumptions you want to put into it. So, in that sense, the model is quite general.
Last word to Alun.
Could you do that? [Laughter.]
Yes. It sounds fascinating. You mentioned that the Baltic states, of course, have used this, and I'm interested in—. We have opportunities in Wales to do things, and I think the more information we have about a richer policy environment in the debate would be something that would be very useful to us.
Well, it's on our radar, but, of course, the important thing is to first understand the dynamics, because I can already tell you, from the assumptions in the model, what the long run will be with a flat tax. It will be good. But the question is how long? And that's where political trade-offs come in. How long is the long run? Well, that depends on how long the election cycle is—
That's the tricky bit that we have to work on.
Well, Professor Matthews and Dr Long Zhou, can I thank both of you for an excellent hour this morning. It's been really stimulating and it's set us up beautifully, I think, for our session with the Minister in a moment. So, thank you both. You'll be sent a copy of the transcript, by the way, to check for accuracy and, of course, if there are any issues, then please let us know.
We'll just break for two short minutes and then we'll reconvene to take evidence from the Minister. Thank you both.
Thank you very much.
Gohiriwyd y cyfarfod rhwng 10:00 a 10:05.
The meeting adjourned between 10:00 and 10:05.
Croeso nôl. Welcome back to you all to the Finance Committee this morning, and we move on to our fourth item on the agenda, which is to receive another evidence session on the impact of variations in national and sub-national income tax. I welcome Rebecca Evans, the Minister for Finance and Trefnydd once more; and joining her this morning, Anna Adams, deputy director, head of tax strategy policy and engagement at Welsh Government; and Tom Nicholls, economic adviser with Welsh Government. Croeso. Welcome to each and every one of you.
Can I go straight on to questions, if that's okay with you? I'll just start by telling you, of course, that the committee has heard that international studies measuring the impact of behavioural responses to cross-border tax variation can't be easily applied to Wales, as you can imagine in relation to the Wales-England border. So, I'm just wondering if you could tell us a little bit about what work you might have undertaken to measure the effect of behavioural responses on the Welsh rate of income tax revenue, as a result of tax diversion across the border.
Well, good morning, Chair, and committee. As income tax devolution is a relatively recent phenomenon in the UK, actually we're in a similar position in the sense of having a lack of applicable data to enable us to estimate any migration effects that might occur within the UK. But, ahead of any data being available, Welsh Government has undertaken a detailed literature review, so looking at the kind of case studies that I'm sure committee has been exploring as well, and we published some of our thoughts on that in the 'Welsh Government: Welsh Tax Policy Report 2018', and perhaps I'll ask Tom to add a little bit of information on that. It's really important that we take account of the best available evidence, and that's one of the reasons why we're trying to expand the base of evidence that we have, and also expand the expertise that we have in Wales. So, I understand that you've just heard evidence from one of the individuals whose PhD Welsh Government were sponsoring in order to help us widen that tax base, but also take the opportunity to widen the skills that we have in tax analysis and so on here in Wales as well. So, the information isn't there at the moment because, as I say, we're in early days but we're keen to learn from other places, although, as you say, things aren't easily applicable to Wales.
Sure. And did you want to come in?
Sure, thank you. So, I think in terms of general behaviours, tax-induced behaviours, I think the UK has quite a good evidence base. The UK has quite a long history of tax changes, and from that we have been able to observe—when I say 'we', I mean UK Government as well. So, HM Revenue and Customs, the Office for Budget Responsibility and the Institute for Fiscal Studies as well, so even outside of Government have been able to study what the effects of these changes have been to analyse what the behavioural effects would be. So, there's a good UK-wide body of information on behavioural effects.
But, with tax devolution we now have a new dimension of behavioural change and that's within the UK migration, and so, there is no evidence base on that for the UK. That's somewhere that we're conscious that the information needs to be improved, we need to improve the evidence on that, and that's acknowledged by HMRC and the Office for Budget Responsibility, and also our colleagues in Scotland in the Scottish Government and the Scottish Fiscal Commission. So, tax devolution is new to the UK, but having 'sub-national' as it's kind of referred to in the literature, sub-national variance of income tax is not unusual in other countries, and so from that, what can we learn?
There's quite a lot of economic literature that's been written on some of these countries, and that's where we've looked to do this detailed literature review to understand what we can draw from, looking across the studies on when tax rates have varied, what's happened as a result. And, particularly looking at this migration, because it's that particular behavioural response where there is, potentially, the largest current evidence gap. And, from those international studies, what can we try to infer what could happen in the UK if there was a tax change within the UK? And obviously we're most interested in what would happen for Wales. So, we published, as the Minister said, in 2018 what, at the time, our latest review and what our latest views on that evidence said. We're continually monitoring that. I mean, studies drip out now and then and update that evidence base. So, we have one or two newer studies that have added to that but don't really drastically change the picture in terms of what our views are on that.
In terms of what they're saying and what we can draw out of it, which is, I guess, why we're here today, is, it is highly uncertain. Behavioural effects, I'm sure, as something that has been mentioned many times in these evidence sessions, are highly uncertain; they're not exact. But there are some general things: if tax goes up, then people are more likely to migrate out, and vice versa. So, I think that's the general finding, and I think that's not an unsurprising one. But these effects aren't very large, particularly for the general tax base. They're very hard to actually identify and actually find in the data. But they are larger at the top end of the income distribution. For the real top earners, then they are more likely to move, and I think this is something that really does come strong from the data.
Now, how strong that move is is very uncertain, and the studies vary enormously on that. There are some studies that say, 'Actually, there are no findings', and there were some quite famous studies done in New Jersey, for example, looking at tax rate changes there, that millionaires didn't move. However, there are other studies that have been done in Europe that find that people do move, particularly ones done that are looking at the Swiss cantons, and they actually do find some slightly larger behavioural effects in those places. There's a judgment there in terms of how much we think we're more like New Jersey or more like cantons of Switzerland. And then, there are other studies in between looking at Spain and Italy. So, it's trying to draw inferences from those types of studies.
Sure, and we'll be drilling into some of those issues as we proceed. Can I just come back to the evidence session that we've just had, really, and the reference you made to the work that Government is sponsoring at Cardiff University? Have you taken anything from that? What have you learned from that work, really? Have you used it in any way?
Well, the work began looking at land transaction tax to explore what the behavioural impacts might be there. But then we sought to expand that tax, then, out into income tax. I know, again, Tom is much closer to that work than I am.
Yes. Thank you. That's correct. So, this is quite a long-standing study. It was a PhD funded, and as you can appreciate, they're not instant. So, we started this funding scheme a few years ago with the intention to look at the economic effects of land transaction tax, and then, following that, we had the powers over income tax, so we were able to turn the project, once it was set up, to look at land transaction tax and then start looking at income tax. It's only just finishing and they're only just reporting the results. I think their first paper has only, literally, just come out. So, I think the inferences from it and what we're learning from it is still at the early days.
I think the purpose of that project was, first and foremost, whether it is doable. That's why it was a PhD project, and I think what we've learned from that is, 'Yes, it is.' But I think, as you learned in that previous session, there are certain caveats to this work—strong caveats—and we knew that was always going to be the case, even if it was successful. So, I don't think that's too much of a surprise to us.
I think it was purposeful to build the capacity, as the Minister rightly said, in Wales, because this is a new policy area. I think this is a really interesting research area, and I'm pleased that academics in Wales are engaged in it. So, I think that's what we've learnt, that there is appetite for this in Wales for this kind of work. So, I think that's a major win.
In terms of direct inferences from the study and what we can learn from it, I think it's very early days. I think there are a lot of assumptions built into it, and they acknowledge it themselves, the researchers. This isn't for us to go away and directly build policy on; it's adding to the evidence base. It's also more of a platform with us to potentially build onto. There are a number of potentially quite large obstacles in the way in terms of data or caveats that we'd need to resolve to improve the data, but I really think it is a positive step forward.
And based on the evidence we have, I think we'd concur with that as well. Mike.
Well, we have evidence closer to home, don't we? We have Northern Ireland and southern Ireland, which are, by definition, on the same island, where people commute across the border. People pay tax in whichever jurisdiction they're in, the tax rates are different. I mean, have you seen anything from those?
So, generally, what's quite difficult to look at is when you have people moving across what are called tax jurisdictions, because to do really good studies, what you need is all the data in one place for you to analyse. It's very hard for you to get the data from different jurisdictions into one place for economists, researchers to analyse that data. So, the short answer is 'no' because it's very hard to get data from Northern Ireland tax records, Irish tax records, and see where people have moved in terms of how it's affected the tax base. I think it's a really interesting example, but that's where the sub-national stuff's a bit more easy, because you can look at Switzerland, and there is a national tax system, so researchers generally can get access to the tax data. So, it'd be a really good study to look at, but we're not aware of any that exist because of those data issues.
Data was the next thing that I wanted to raise with you, which you've already highlighted. There's always a continuous appetite for more data. How is the Government trying to address that? Are you proactively looking to facilitate more of that kind of stuff?
We will, as time moves on, get more data in terms of the sub-national effect, because, of course, the 2019-20 data for Welsh rates of income taxes isn't actually due until 2021. Even when that data does come in, it will only be very limited in its use to us, because our income tax rates have been the same as they have been across the border in England, so we won’t be able to tell much from that particular data. What we can do is look towards Scotland for evidence and data, but, again, it will probably be a year or two before useful data comes from Scotland in terms of being able to help us understand the issues.
But I'm wondering, more generally in terms of data, how we might be engaging with some of the stakeholders that we've heard from to understand from them what they need, because we heard earlier this morning that there was an appetite for more data around migration, around income structure and household surveys that are already there, but it's just that they want to get their hands on them.
Well, there's also a lot of appetite, I think, for more data to gain a more complete understanding of how the Welsh economy works in terms of those cross-border business flows between Wales and England. Again, that's an area where there is a gap in evidence. Work is under way to understand how feasible it would be to get a greater understanding there. So, we're aware of these gaps in evidence and we're exploring ways to close them.
And that work that's under way is internal work by Welsh Government, yes?
Yes, there's already been a survey in the field to try and—a one-off survey to see how viable this is to do, because the brilliant thing about the UK is it has literally frictionless borders. So, as a result of that, businesses don't really record Wales and England—
No, so you don't capture it.
So, you have to go out on surveys, and that can be quite onerous on businesses. Some businesses don't even know that. And so, this has always been a major restriction for this type of economic modelling in Wales. Scotland have done it to some degree, but, by assumption, they've assumed that trade between Scotland and England is not massive. I don't think we can make that assumption in Wales; I think we have to assume almost the opposite. Therefore, we're very reliant on these kinds of surveys, which are, as I say, being explored, because there are other needs now with new trading relationships that are going to happen. There are multiple policies in which we're going to need better information on this trade, so it's not just in the tax domain. I think there are many. Whether it's viable, what the cost of that is and the burden on business is stuff that we are actively exploring, and we're going to learn more, I think, in the next few months, hopefully. But that is something that the Welsh Government is very conscious of.
Thank you. Alun.
I'm interested in the whole field that we've debated in terms of the availability, the nature and the integrity of the data that we have. There are certain things, themes, if you like, that seem pretty well accepted, for example, I think it was the IFS in their evidence to this committee that said that it is clear that higher income individuals are more responsive to changes in a tax regime, whatever that may be. So, knowing that, or assuming that we know that, the first question is: does the Government accept that as a truism? Secondly, if it does accept it as a truism, does it influence policy?
So, I think that we do accept that the evidence tells us that the higher earners are more susceptible to behavioural changes in response to tax, and income tax particularly. In terms of influencing policy, there's a real tension, isn't there, in terms of seeking to encourage high earners to come and live in Wales, but, at the same time, maintaining our very progressive approach to policy and tax policy.
So, we have five principles by which we shape our Welsh taxes, and the first is about raising taxes in the fairest way possible, so that naturally leads us towards more progressive ways of developing and making decisions on taxes, and then it has to be about delivering our other Welsh Government priorities, particularly supporting jobs. So, we want high earners who are coming to Wales and able to create jobs. They have to be clear, safe, stable and simple, and also developed through collaboration and contribute to the Well-being of Future Generations (Wales) Act 2015.
So, if we were to look at how we would attract those high earners to Wales, it would be about reducing the very highest rate of income tax to encourage them to come to live in Wales. We could, for example, reduce land transaction tax on the most expensive properties. The question is: is that the space that we want to be in in terms of delivering progressive taxation in Wales? So, I think there's a tension there, and they don't pull in the same direction.
Oh, we understand the tension, Minister. My question is: what do you think?
So, our approach is to have a progressive tax regime. So, it's not about introducing tax cuts for the highest earners and the most wealthy. But these won't be the only reasons why people decide to come and locate in Wales. So, it could be about generally having lower house prices here in Wales, so it'd be getting more for your money, and obviously we take for granted the environment we have here in Wales as well. So, there are other reasons—
But this inquiry's looking at tax in itself. But I don't disagree with your wider analysis. There is of course a proposition—you referred to fairness, and I've got no disagreement with that, clearly, but what is 'fairness'? And 'fairness'—I think you've interpreted it in that answer—is a tax regime where there are different rates of income tax whereby people who earn more will pay more tax proportionally on those additional higher levels of income. That's a very traditional view of fair taxation. It may well be, and I think the proposition has been made, that, were we able to reduce taxation on higher incomes, then we would actually raise more income and therefore it is fairer, for example, for people who rely on the national health service or whatever, in order to raise additional income, to fund those services, for argument's sake. So, fairness may be less absolute, in that sense. Has the Government been having these sorts of discussions and debates?
I think there would have to be quite a large response in terms of higher earners coming to live in Wales to take advantage of lower tax in order to gain the kind of benefits—. So, at the moment, we only have 6,000 of those additional rate payers here in Wales, so—
But higher rate.
Yes. And we have a larger—
Additional rate payers.
Yes. Additional rate payers—6,000 of them.
How many of them?
That's an estimate.
So, those are individuals earning £150,000 or more.
Nick wanted to come in, if that's okay.
I just—I remember Gerry Holtham giving some evidence to this committee some years back where he said, if you look into the data, it's actually quite a small number of people—the very high earners— who are contributing quite a high percentage of the whole upper rate tax take within Wales. He also suggested that, of course, you could seek to reduce the additional rate, the very high rate, to attract some of the top earners, but, at the same time, you could use stamp duty land tax, you could use council tax and other taxes to try to claw back, I suppose, if you want to use that term. So, is that something that's been looked at by the Government?
So, just looking at income tax just for now, in our 2018 tax policy report we did look at—building on the work that the Wales Centre for Public Policy did, looking at, if you were to reduce income tax, could you recoup the revenues just from that alone. I do accept your question was about other taxes, but if we just look at income tax—. We then compared how does that compare to the international evidence, and that really was at the top end, even beyond the top end, in the international evidence, to try and get that behavioural response to give us more revenue—to answer more your question in terms of could we reduce the top end there. And that is because, as you say, building on what Gerry said, most of, actually, those additional rate taxpayers currently in Wales—I know we'd be attracting them from England—are actually around the £150,000 mark. So, you would really be looking for the top end to try and generate that. But it's also because we only actually gain 10p of the £1 of their tax rate, not the full 45p rate. So, that means that you do need a really large behavioural response to get them in terms of income tax.
I was just wondering if I can—just to complete this section. So, in terms of where you are, Minister, you're looking at—you're building this evidence base and understanding the wide horizon of different pieces of data, which will complete a wider picture, and I think the committee would welcome that. But, in terms of policy, and in terms of your policy making, what is it that your priorities are in terms of—? I'm thinking of income tax—we understand what you're doing with vacant land tax, but I'm thinking in terms of income tax. What is it that you are giving consideration to at the moment in terms of potential future opportunities to increase or decrease or make changes to the Welsh rates? Where is it that your emphasis is placed in policy terms and policy development?
So, I think we're all in a similar place at the moment in the sense that the Welsh Government has said that it won't raise income tax rates over the course of this Assembly. So, the kind of discussions and thought I'm giving to it now is where we take things for the next Assembly and what our manifesto commitments might look like. So, obviously, I wouldn't want to go into too much detail. But I know that all parties will be using the work that this committee does in terms of considering the way forward. But I think that if we start from a base of fairness and progressivity—I think that that would be where we are.
Can I abuse the generosity of the Chair? You talked about those four or five principles, and I don't disagree with them, but I think they're limited. Have you given consideration to a review of that? I think that's work that Jane Hutt did as finance Minister, I seem to remember, and she produced a paper on it in the last Senedd. Have you considered reviewing those principles? From my point of view, for example, I think the fact that sustainability isn't there is an extraordinary omission. So, are you looking at that again at the moment, or are you simply working on the basis of the paper that was produced?
Well, sustainability, actually, is one of those core themes that we're looking at when we're looking at potentially developing a social care levy or would we use Welsh rates of income tax for that. So, that's very much at the heart, because clearly we can't introduce a short-term policy for social care.
I was thinking in terms of climate policy, rather than other sustainability, but—.
I think climate policy is probably well captured by requiring us to consider and work within the Well-being of Future Generations (Wales) Act 2015, but I'm open to considering things further. These are really early days for us in tax here in Wales, so I think it's only right that we should keep everything under review to make sure that we're doing the best we can.
Okay. Thank you. We mentioned the higher rate tax and some of the evidence we've had has suggested, and I think you reflected that, that they would be the people who would be most able to—or you'd probably see a greater effect in relation to, when it comes to changing the rate. One of the questions, maybe, is: how prevalent do you think tax avoidance or planning could potentially be amongst that group if there was a notable change? Or is that something that you're aware might occur but we're not there yet kind of thing?
Obviously, that's an existing issue already in the income tax system. This would be an issue for HMRC to deal with. So, this is obviously—. They have a relative risk in terms of who they would look at in terms of income tax and, with tax devolution, they now have a new dimension to that. So, if tax rates were to vary in Wales, as they have in Scotland—. This is now more on their radar. So, this is something that HMRC will be looking at. Obviously, we also have a very strong, keen interest in that, because we are the recipients of that revenue. So, that is definitely a feature of that. That all stems, really, from the work that's going on with getting people to have the right, C, code—to have it on their tax. If you don't have a C code, then obviously you're not paying the Welsh rates of income tax.
So, there's quite an extensive bit of work currently going on, because we're in the first year, but that will be ongoing with the Welsh rates of income tax, with HMRC, to give people the correct, C, code. There are various bits of work that HMRC are already doing by corroborating it with other bits of data to make sure that people do—. Some of it might be just error, but, as you said, there might be tax-motivated incentives to do that, going forward, and that is something that will be taken forward, going forward.
Okay. Rhianon wants a quick supplementary and then we'll move on to Nick.
In regard to the evidence that the Organisation for Economic Co-operation and Development gave us a session or two back, if I recall correctly, they were very clear in their view that any drive towards incorporation from those who were self-assessed or the very high rate or additional rate potentials are already outside of Wales in the first place. They were very clear about that. Have you got a view on that?
I think that's an interesting view. That would be a UK tax base issue, I think, and so that would be something that would be more appropriate for HMRC, whether they think that. I think, for us—
But I think, in regard to the round, we can't, in a sense, self-isolate around that particular point.
No. So, we're—. Yes, so, I think this one—. There are two answers, there are two bits, to that. I think there's—. That probably is an issue. It may be—. Sorry, it may well be an issue for the UK tax base, and that's something that—HMRC are obviously very interested in those people. But—
But the point would be—to interrupt, if I may, and obviously you can come back, but—. I think the point would be that there would be a negligible drive in terms of taxation planning around higher-end drive to avoid.
I think it would be a slight behaviour, but I think migration here potentially is the bigger unknown now, because it's so new. The behaviours you're describing are not unknown to HMRC and they're well monitored as well, because they're something that already affects the UK tax base. For Wales, with the system we have in our fiscal framework, we're moving off a platform from that base, so we move off a platform from whatever our revenues are in 2019-20 to growing our tax base from there. So, if that has happened, in some ways—it's a bit callous to say this—we don't always care, because it's in our baseline, so, if that has already happened, then that's fine; it won't affect us going forward so much. Now, it might if tax rates vary, and, as we described, HMRC would be interested in this as well, but maybe it would be dampened because maybe most of that behaviour's already happened, because there already is the differential in the tax rates.
Okay. Thank you.
And it's already factored in, isn't it, to the HMRC work and the forecast from—[Inaudible.]
Exactly. Yes. Okay. That's good. Nick.
Diolch, Cadeirydd. The committee heard that seeking a broader tax base or encouraging the growth of existing tax bases would be preferable to increasing tax revenue, rather than changing income tax rates. Do you feel you have the appropriate fiscal levers to achieve this, and what are your strategies for growing the base, the tax base?
I think that it's good that we have both tools. So, we have Welsh rates of income tax but we also have the opportunity to grow our tax base through the right investment and policies. Changing tax rates, of course, would give us more of an immediate impact, whereas growing the tax base is a much longer-term endeavour. The difference now I think is, really, that we've got more skin in the game in the sense that we're exposed now to the growth and to increasing the take of tax in Wales, so that's the difference, I think, now that we do have both of those levers.
And have you—? It's similar to what I asked earlier, actually. Have you looked at the potential consequences of attracting more taxpayers to Wales in areas such as house prices and rental costs, and how could these impact on people currently living in Wales?
So, I think there would have to be really significant changes to migration and commuting flows in terms of the basic rate of income tax. I think such an effect would probably require large changes, because 1.3 million people are affected, and changes to the other tax rates then would affect far fewer people. So, there would have to be huge movement in order to have any real impact on house prices and rents in Wales. You can look at areas where Government policy has changed or had an impact on house prices. When you think about the removal of the Severn bridge tolls, people there, for example—. The cost to a commuter for using the Severn tolls was in excess of £1,000 a year but, in contrast, a 1p change to the basic rate in Wales for the median earner would result in a change of tax of just £100 a year. So, I think that the level of impact it would have on those individuals is very different.
Yes, it puts it in context, doesn't it, really. Okay. Mike. Sorry, Nick, was that fine?
There was—. I've forgotten was going to ask you then—[Interruption.] I'll have a think and then we can get to Mike. It's all mind-boggling.
You gave an answer to Nick there that is contrary to the information I have, so perhaps you can correct me. My understanding is the average median wage in Wales is £25,000 and a 1 per cent change would increase their income by about £10, which would be £120.
One hundred and twenty pounds?
One hundred and twenty pounds a year.
Yes, about that.
Yes, just over £100.
Thank you. Because you said £100 earlier, and so I—
Yes, I should have said just over £100.
I was just trying to work out whether my numbers were right. House prices, we've mentioned it, and you mentioned the toll, which—we can't say it's a cause and effect, but it's an effect following the change. But aren't these non-tax factors, unless you're going to make huge changes, far more important to people than minor changes in income tax, especially for those on the two lowest bands? And can I just follow up? For people on the highest bands—you do know, but I doubt you've got it with you—how many of those are actually employed in the public sector in Wales?
Tom might know. [Laughter.]
I don't know by band, I'm afraid, no. [Laughter.]
In the top band, you've said there are 6,000, how many of them are employed by the public sector? If you haven't got it with you, which I didn't expect you to—you didn't expect this question—could you send us a note on that? Because that does have an effect.
Sure, yes. We can't give you the estimate for the current year, because that kind of level of detail is quite backward looking, but we could give you what the latest data says of that year, and an estimate for the Welsh rates of income tax, assuming they are Welsh residents. Because what we also don't know in that year is whether they're actually a Welsh taxpayer, but we can look at their residence and from that we'll assume. We could do something like that.
I just think it's an important issue, because there's a difference between people in the public sector and what they can do. The chief executive of this Assembly, or the chief executive of local authorities—many are, in fact, probably all are top-rate taxpayers—they can't incorporate, for example; they can't be paid in a different way. So, the effects on people in the public sector and private sector are different.
The only other point I wanted to raise is: we talked about keeping wage rates for teachers to be the same, NHS pensions the same across England and Wales. Can I say I agree with that? But that could have an effect on where people live, especially in those areas where movement is so easy. Chester and Wrexham are incredibly close. I'd call it one conurbation, but the Chair wouldn't allow me to, but they are very close. So, people have that choice there of which side of the border to live, and sometimes, whether they move one street or a few houses.
Mike's right, in the sense that people's decisions will be made on a wide range of things. So, income tax would be one of them, but you can offer a genuine package of things that will interest people, in terms of free prescriptions, childcare offers, lower house prices, quality of life and so on. So, lots of factors.
Jobs and having free family childcare actually is a major advantage, as many of us know, don't we, Nick?
Our childcare offer, for example, is worth around £90 a week to families, so that is a very attractive thing to many people.
Going back to the importance of the public sector, the statistics I can say is that the Wales Centre for Public Policy, when they did some work on Welsh taxpayers, which is a couple of years old now, found that public sector employment in Wales counted for 27 per cent of our tax base for non-savings, non-dividend income in Wales, and that's compared to the rest of the UK at 19 per cent. So, that's just to back up that point. It's not by band, as you wanted, but that's the kind of information that I think is supporting that point.
The reason I asked by band is because we've been told of the ability of people in the highest band to do all sorts of things. Well, how many of those can do those certain sort of things?
Not all of them can, that's the point, yes.
If you're the chief executive of the Driver and Vehicle Licensing Agency, then what you can do is much more limited.
Okay. Thank you. Rhianon.
Thank you. If I come to my questions, because you've partially answered them. We've talked a lot about the tiny but influential amount of high tax ratepayers, and we've talked about our basic ratepayers. But in regard obviously to UK jurisdiction over current welfare benefit, what assessment or understanding is there in terms of the predicted modelling of a 5p cut to income tax in Wales through the CGE model that we talked about earlier? What assessment or understanding is there in regard to any impact or effect on those, for instance, on benefit or universal credit? Because we know—we've had it strongly stated here now—that can cause a loss of revenue in the short term. So, is there any view on that?
I think there are a few questions there. The question there in terms of impact on welfare and people relying on welfare from a 5p cut in that CGE-type model: as far as I'm aware, and certainly in the modelling that we've been shown today, that wouldn't be incorporated in there.
I think what I'm getting at is in terms of the holistic picture in regard to a loss of revenue. Obviously, there's a lot of mitigation around council tax exemption reduction, discretionary assistance fund, in terms of what we're able to put into our financial mix for all in society. What I'm asking is: is there any analysis of how that could affect those on universal credit, who are at the bottom end? You've talked a lot about those at the top end, and rightly so, but what about those?
We do have the ability to show what potential changes to income tax would have on household incomes, and that's something we first did, I think it might have been back in—I hate to keep referring to the 2018 'Tax Policy Report', it might have been even before that in the 2017 'Tax Policy Report'. We could show that what we could look at is how taxation actually affects by household income, by the distribution. What we did in that report for the first time was we showed the relative difference between income tax and council tax. That was very illustrative, just showing how people at the lower end of household incomes pay very little income tax; even the basic rate is a highly progressive rate, because you have to be earning over £12,500 until you're actually paying the tax rate. So, income tax is a highly progressive tax. So, even small changes—. The basic rate has very little effect on those on the lowest income.
I accept that. That's not my point. Sorry for being inconcise or unconcise—whatever the word is. What I'm saying is, in regard to the loss of revenue that then there would be, as predicted in the CGE model, if there were to be a 5p cut, there would be a loss of revenue to the Welsh Government. That's my point.
So, has there been a conversation around that, and an understanding as to how that would impact those at the very bottom? Because although we've got many public sector workers, we also know we've got a disproportionate amount of people who are on universal credit.
That loss of revenue coming to the Welsh Government wouldn't impact the majority of people who are on welfare directly, because that would still be paid by the Department for Work and Pensions.
I accept that.
But it might affect what we could fund in terms of other general public services.
That's what I'm getting at.
Thank you, Chair.
But any reduction in—. The Scottish experience was that, when they cut the rate for the lower rate, of course, the means tested universal credit was amended, and therefore any saving was lost. So, it's not the impact on the Welsh Government, it's the impact on the individual and their family.
But what I'm trying to get at is that the loss of revenue would also then have an impact on those at the bottom—
I think that's true, yes. It's progressive. I think that's what we are saying. Public spending generally benefits those at a lower-end income, as they are more engaged in terms of health services, or in terms of local authorities—I guess this is your point. I think that's definitely an issue, if you were to cut public services in general, then that would have a disproportionate effect. That's been shown by previous work on Wales and at the UK level, on the effects of austerity, both those that happen at the UK level and with us having budget cuts, in terms of, where it's happened, then you can see that public services would be affected, and that's something that would obviously be in our workings, going forward. If that was to be a policy, we would be able to show that, yes.
Anna wants to come in.
Just to add, in yesterday's budget, the UK Government actually changed some of the offer. I'll read it out, because I think it's easier to do that. They said:
'The government will legislate in Finance Bill 2020 to clarify the Income Tax treatment of three new social security payments. The legislation will confirm that the following three benefits introduced by the Scottish Government are exempt from Income Tax: Job Start, Disability Assistance for Children and Young People; and the Scottish Child Payment. The legislation also includes a new power which permits the government to confirm by statutory instrument when new social security benefits introduced by the UK Government or any of the devolved administrations will be tax exempt.'
So, they're trying to work on to make sure that those sorts of situations don't happen again, and the change will take place from April.