National Assembly for Wales

Well, I think they're fair. I don't know how simple the changes are to the tax on housing transactions. They've made an effort to be fair at the cost of a little bit of complexity, I think, but I don't think they've done anything that I would want to criticise at this point. I think the approach of the Welsh Revenue Authority to take a sort of collaborative approach to compliance is probably a good one. The proof of the pudding will be in the eating, but I think it's the right way to start. I don't have personal experience of it—I haven't bought a house or anything in Wales in the last three years—but as far as I know, it seems to be working fairly well.

There we are. Okay. Thank you for that. The Government, of course, are currently consulting as well, aren't they, on new tax legislation, so that they can make immediate changes to Welsh taxes in response, or to react to any events such as UK policy changes. Do you feel that that kind of legislation is necessary, and what would you say are its strengths and weaknesses?

I think consultation is very necessary. If Wales raises any tax rates, it will enter into tax competition with the big boy next door. And I think it's very important that it has achieved a level of public understanding and, as far as possible, agreement with whatever it's opting to do, so that you don't lose out from tax competition. It's not necessarily bad to have higher rates, as long as people think they are getting something for them and approve of that. So I think the consultation is very necessary. I think it's absolutely important that a small place like Wales takes the public along with what it's going to do. And, of course, there's no shortage of things to spend money on, so the hunt for other sources of revenue is a reasonable one, but I think it has to be handled very carefully.

And I suppose there is something there as well about striking the right balance between allowing the Government to be nimble and to respond to changing circumstances, but also allowing for meaningful scrutiny that isn't a tick-box exercise at the end of the day.

Indeed. And I think that's why it's important to try and anticipate things and to, if you like, soften up public opinion in advance of having to take a decision. It's not possible to take the public along in a fortnight—you have to prepare the ground, as it were.

Well, bang on cue, Mike Hedges has joined us. Mike, we're just coming to the area of questions that you were going to pursue, if we're not throwing that at you unexpectedly.

I hope you can hear me.

We can, loud and clear.

Good, because we've spent the last half hour attempting to achieve that. The Welsh Government recently published its tax policy work plan, which includes the aim to strengthen the Welsh tax base. What do you feel are the necessary tools to grow the tax base? And, surely, growing the tax base really means growing the economy.

Well that's essentially correct, yes. Growing the tax base does mean—well, certainly growing those parts of the economy on which you're levying taxes. And as income tax is the most important one that Wales currently controls, then that means growing wages in Wales, which is growing the economy—absolutely right. Do they have the tools? They certainly don't have the tools to ensure it, they don't have the tools to mandate a growth in the economy—that's certainly the case. They have a few tools that could be used to encourage the growth of the economy, but it's not going to be an easy job.

I agree with that. Of course, land transaction tax is not directly proportional to the growth in wages, but obviously, as wages go up, not only does income tax go up, but the value of houses goes up because people have more money to spend on them. Would you agree with that?

I don't think land transaction tax is a great tax, actually. It's a tax on turning over the housing stock, the property stock; it's not a tax on the stock itself. So it's not a proper wealth tax, it's more like a transactions tax. It's there, it's accepted by people, they understand it, so I wouldn't argue for abolishing it. But I don't think it's something you should look to for any further revenue. If you increased council tax by 10 per cent, you could abolish it altogether; it isn't a very significant tax.

You're coming to one of my areas now. Of course, if you increased the number of bands in the council tax, and you had some of the people in the £1 million houses, £0.5 million houses, paying more, then that would have an effect as well.

Well, as you know, I can be a bore on this subject—I've written a lot of things about the need to reform council tax. If you just made it proportionate to the value of the property—. Because at the moment, it's highly regressive. It's not just regressive in income, it's regressive in the value of the actual property that it's laid on. So, someone in the Valleys with a £50,000 house is paying 2 per cent of the capital value in tax, whereas somebody in a £1 million house in Cyncoed, or somewhere, is probably paying, I don't know, 0.5 per cent. But there's no equity—[Inaudible.]—tax, and if, say, everybody pays 1 per cent, you would halve the tax rate on houses at the bottom and you would probably multiply it by three or four at the top, which would be very unpopular, but that would be one way that you could make it fairer, and it would be a tax on wealth, on property, not on income, and wouldn't have such a discouraging effect on activity.  

I won't ask any more questions on this. I agree with you, and it's something I've both written about and argued about for some time. 

There we are. Thank you, Mike. On to Nick Ramsay, then. 

Diolch, Cadeirydd. Gerry, the UK Government's report noted the governance arrangements that continue to oversee the final changes being made, including to HMRC self-assessment systems. Can you elaborate on what the final changes are and how they'll impact on the Welsh rates of income tax? 

I'm afraid I can't help you there, Nick. I'm not really across that issue. 

No probs. Can I just go back to something you were saying to Mike Hedges with regard to council tax? I understand why you're not a fan of council tax, and it's got many problematic issues but, of course, one of the big issues levied against it before is, because it's not a tax on wealth and not a tax on income, it's related to the size of the house, you could end up with people actually not on very big earnings who end up living up in a larger house, maybe through inheriting it or whatever, who then would—it wouldn't necessarily be fair then to tax them on that basis, would it, if you did make it more proportional. It's just an interesting issue that you've just alluded to. 

Yes. I had a look at that. The proportion of people who could be thought of as not-well-off living in houses in bands G, H, I, for example, it's between 6 and 9 per cent, depending on how you define not-well-off. So, it's not a large number, and I think for people like that you could have deferment arrangements, whereby, if they can't afford to pay the tax, they don't, but it rolls up and it's then a charge on the house when sold or on the bequest, if left to relatives. So, I don't think you need to throw people out of their homes or anything like that. I think that it would be possible to get around that and it is a problem that it's possible to exaggerate. It certainly exists, but it is quite a small proportion of the housing stock or people that are in that situation. 

Thanks. I asked you previously about the Welsh taxpayer situation. Am I right in saying that an issue arose where there were problems with identifying Welsh taxpayers? And I think we even ended up with some Welsh taxpayers on the Scottish rates of income tax, SRIT, rather than the WRIT. 

Yes, I think that's true, and I think one of the problems of the devolution is that we do depend on HMRC, which is an organisation that is under pressure. I think it's suffered from economies itself. It's not got a lot of spare resources for doing anything very complicated, and it's not clear to me that it's got a very strong incentive to work terribly hard at ensuring that Welsh taxes go where they're supposed to. It's obviously got to do what it's supposed to do, but, when it comes to compliance and things like that, they are going to be under pressure and I can't see that they have got a huge incentive to knock themselves out. So, I think that will be an ongoing issue. 

Okay, thanks. 

And we can ask them directly within the hour, because they're up next in our meeting today. So, it will be interesting to ask them what incentives there are for them to get that as right as possible, really.  

I did wonder whether there wasn't a case for putting them on piece work, you know, so that the fees the Welsh Government pay to them for doing the job are somehow related to results, a bit like tax farming. I admit that working out a feasible plan in that way would be quite difficult, but it does occur to me that a flat-rate payment is not necessarily going to give what you want. FootnoteLink

Yes. Well, we'll see what they have to say about that, then. Okay, so on to Rhianon, then. 

Thank you, Chair. The Welsh Government has been critical about the current mechanism for devolving tax powers, namely that it's not fit for purpose. How important is it to have that robust mechanism in place—that's almost a leading question—to introduce new taxes around influencing behaviour, encouraging growth economically or supporting public spending. But I suppose more important than that obvious question is really what mechanism would be most appropriate or optimum? You have sort of inferred some of this, but not directly.

Yes. Well, of course, the ability to acquire new taxes in Wales does depend on the consent of the British Government, and there's a very general principle that says things shouldn't be to the disadvantage of other parts of the United Kingdom. But there's not much case law, as it were, or much detail about how that's assessed. So, I think we're in the position that you could make an application, and the judge and the jury and the witnesses are all the same person, and so it's a little unclear how to proceed. If the British Government says 'no', what do you do then, even if your request is perfectly reasonable? So, I think it is something, like a lot of things in the British constitution, that would benefit from a little more clarity and perhaps a little more codification of what counts as damage and what's permissible.

So, that's the general scenario, but to be more specific to my question in terms of an optimum mechanism—. That's the challenge, isn't it, that's the difficulty in the context and the background that we're in. But, in regard to that mechanism, assuming that those leverages have been pulled and Her Majesty's Government steps up to the plate around that, have you got any thought on that? Because I'm still no clearer from what you've said.

When you say 'mechanism', do you mean for administrating a tax or for agreeing that a tax be devolved?

For administrating—so, that mechanism. Obviously, you'd have to do one before the other, but I'm thinking of that as a question to you.

Well, I think, as far as possible, it would be in the interests of the Welsh Government to develop the Welsh Revenue Authority and to take control of the administration of the tax here, because, if you're relying on the HMRC, we've got the same issue that we were just talking about. So, I think, in that sense, if you're going to have considerable tax powers, it seems to me to follow that you'd want to have a commensurate and adequate tax administration machine and not to rely entirely on the UK.

Of course, it does depend on the nature of the tax. We couldn't currently take income tax over, for example, whereas if—one of the things that has been talked about—you were to devolve capital gains tax on property, for example, I can see no reason why that shouldn't be administered in Wales in the same way as the property transfer tax is being administered.

So, in answer to my question around a robust mechanism in place in order to be able to alter social behaviours, what you're saying is that comes secondarily, obviously, to what you've just said, but you've got no further comment on what would be an optimum mechanism.

Well, I don't think I do. At the moment—. There are two parts to this—(1) you've got to agree with the British Government on what taxes, in principle, could be devolved, and I don't think the procedures for doing that are as clear as they might be. Then [correction: Then (2)] you've got to decide how to administer the tax, and I think, certain taxes, just practicality says it will stay with HMRC. But others—and, as far as possible, more—should be handled by the Welsh Government's own revenue authority. I don't have any specific views on how that should be reconstituted or altered or reformed.

Okay. Thank you very much, Chair.

Diolch, Rhianon. Alun Davies.

Thank you very much. Professor Holtham, I was reading your paper that you wrote in April about how Wales would be financially impacted by COVID, and you were discussing the Barnett formula and the rest of it. Your view seemed to be—forgive me if I'm getting this wrong, but correct me—that you felt that Barnett, even with the new needs-based comparability elements to the formula, wouldn't properly reflect Wales's needs during the COVID epidemic, because of an older, less healthy population, if I've read that correctly. Do you think that there are wider lessons there about Barnett?

Yes. I think many people would agree that it would be better if the Barnett formula explicitly contained elements that tailored the allocations to need, and it's quite a long time ago now—about 10 years ago—that we produced a formula that was quite simple, it only had four or five elements in it, but it went quite a long way to showing how you could adjust the Barnett formula on a needs basis. And if that were the case, then expenditures currently in England would have a Barnett consequential that was adjusted for need.

Now, of course, something like COVID, although it's [correction: COVID is] correlated with other needs, so you tend to get more of it where people are poorer and in worse health. So, standard needs adjustments would go quite a long way to adjusting for the non-uniform incidence of COVID problems, but, of course, wouldn't go all the way. This is a new event, and, to that extent, you wouldn't expect any formula to actually completely solve the problem. But I think a standard needs-based adjustment would have helped. Unfortunately, we've known this for at least 10 years and there seem to be very difficult political roadblocks in the way, not the least of which is Scotland. 

Yes. It's fair to say that an extreme event such as COVID does help shine a light on things that we've known anyway and brings it into a sharper relief.

Indeed.

In terms of the wider elements of the fiscal framework, you've written on this many times over the years: how do you see it actually operating at the moment? Is it meeting Wales's needs? Is it meeting people's expectations? Do you feel that it's working?

Well, yes. I think, in a rough and ready way, the changes brought in in 2018 and the floor, and the fact that, even when we're above the floor, we're getting 105 per cent—[Interruption.] Let me just turn this off. I thought I had. I do beg your pardon.

No problem.

Yes, the fact that we're getting 105 per cent of English expenditures even though we're above the £116 floor is a—it helps. So, I think we're no longer in the position where we can say we're being clearly underfunded. In a general sense, I think it's no longer easy to make that case, but then events like COVID, as you say, do throw up instances where our particular needs will not be met.

But, of course, the fix is—if I can put it that way, we're, in the great British tradition, entirely ad-hoc. There's no real set of principles that are being consistently applied behind them. The 105 per cent expenditure is a bribe, if you like—it's a compensation for the fact that the way they're administering the offset in the block grant for the tax deduction is not advantageous to Wales. The way that they're actually pricing the tax base that they've given us is not advantageous, so, instead of fixing that, they've just said, 'Okay, well, listen, we'll give you 105 per cent on the expenditure side. Now it's all right, isn't it?' And I think the Welsh civil service ran the numbers and had a look at it and said, 'Oh, yes, well, it is all right, actually, so, okay, we'll do it.' But it was typical horse trade; it wasn't anything that was terribly principles-based, and that always creates the possibility that something will happen in the future and that funny tax adjustment will come back and haunt you. So, I think it's fine, it's a good fix, but it is just a fix. 

That's very interesting. The tax and the comparable model was my next question, in fact. When I look at it—I don't have your knowledge and experience, but, when I look at it, I don't feel especially comfortable, because there's something there that feels that it's not been done for our benefit, it's been done—you know, you can easily envisage a Treasury official on a Friday afternoon who's been told to sort this out, and just saying, 'Right, we'll do this, this and this, and that'll work for us.' 'Does it work for us?' is the question I'd like to ask.

Well, this is quite tricky to explain. I'm quite happy to put a brief paper in if that would help. Essentially, the way they've done the tax compensation—they say, 'Okay, you've got 10 points of income tax', what's that worth? We've got to charge you for that and knock it off the block grant.' And so, we're basically talking about a deduction off the block grant for the tax base you've been given, which is reasonable. But instead of saying, 'Okay, we'll relate it to percentages', what they've done is to do a Barnett-like thing where they've said, 'Okay, if the tax in England goes up by so many pounds per head, then we'd expect the Welsh tax to go up by the same number of pounds per head, and that's what we deduct from the block grant.' So, it's not done in percentages [correction: percentages of tax revenue], it's done in pounds per head, then adjusted for population [correction: adjusted], and that has the same consequence as the Barnett formula itself. If the Welsh tax base is lower and growing [correction: is growing] more slowly than the English one, you could overpay for the tax base. I'm afraid I haven't explained that very clearly, but that is the situation: it is a mechanism that could see Wales lose if its tax base [correction: if its tax base—income and population—] grows more slowly than the English tax base. And that's why I think we've been given the 105 per cent. Now, why did they do it? I think they did it because they want to clobber the Scots, and if they've got us to accept it, they get the Scots to [correction: accept the particular tax adjustment of the block grant, they might get the Scots to accept it. And unlike us, the Scots don't have the floor, so the floor bit doesn't come into play; they just have to talk about compensation for taxes. So, if they make that nasty, and then compensate us on the expenditure side, it's just net nasty for Scotland. I mean, this may be cynical, but I don't know why else they'd have done it the way they did. 

Well, certainly, it's probably the best and most rational explanation I've heard so far. If you're able to drop a note on that, that would be—I think Members would be very, very grateful to you for that. I don't have any further questions. Thank you.

Indeed, we would. Yes, thank you for that. Thank you, Alun. Siân Gwenllian.

Well, thank you. I think they should. That would be a reasonable way to decide what the borrowing powers might be. You can understand the UK Treasury saying that it's not your total revenue, because, after all, we are borrowing for that; it's the bit that you can raise on your own account is what should determine borrowing powers in Wales [correction: it's not our total revenue, because, after all, they are borrowing against that; it's the bit that Wales can raise on its own account that is what should determine borrowing powers in Wales]. So, I think one has to accept that that argument's reasonable, but if you think, okay, we could raise £2 billion a year, let's say, from income tax, what leverage would you think would be appropriate on that £2 billion? Well, what I look at is what proportion of that revenue should you be able to consecrate to tax—sorry, to service your borrowing? If you said, 'Okay, listen, we shouldn't be able to spend more than 5 per cent of that £2 billion on servicing the debt', you then have to look at what your annuity rate is, and then you multiply that up [correction: then you apply that] and you get what you get. And I think reasonable assumptions say you should be able to borrow £2 billion or £3 billion.

Indeed. I'm going to repeat myself, I think this is another area where it's all at the grace and favour of the Government of the day, and there are no principles laid down that commit anybody to do anything in particular, so that you never quite know—everything's a horse trade; nothing can be decided on a principle basis.

I think it's not impossible. I think you could—. I don't expect them ever to agree to a formula, but you can use a formula in pursuit of a particular ad hoc settlement, if you see what I'm saying. So, you might get them to agree, 'Well, all right, you can have £2 billion, not £1 billion', because you've told them, 'Well, £2 billion at current interest rates will only be 2 per cent of our expenditure, or whatever, to service that debt.' So, I think you can make that kind of argument that would be the basis for a rational principle allocation. You won't get a rational principle allocation but you can use the argument to get an ad hoc agreement. I think that's probably the best we can hope for.

Two quick points. As you're well aware, local authorities can borrow prudentially, and they set their own borrowing limits, and some English authorities have borrowed huge amounts. The second thing is: we talk about the ability to pay it back; we're at historically very low interest rates at the moment. I remember paying my mortgage at 15 per cent interest rates, and I think there's a likelihood of interest rates moving around, so what might be easy to pay back now, if you're borrowing it on a variable interest rate, could become very expensive very quickly.

That's right, and that's why I would advocate using an annuity rate, which includes paying off the loan, meaning, in practice, the Government never expects to repay the debt, it will roll it on. And as long as it doesn't get too high, it's just always there. The national debt is always with us. But to meet the point you quite correctly made, I wouldn't just take the current interest rate; I would take the rate required to pay the debt off in, say, 30 years; you take the annuity rate that might be 4 or 5 per cent, not 0.5 per cent, and that gives you some kind of assurance that you're not doing anything reckless.

Again, the Treasury's made a couple of ad hoc changes, but, ideally, we should be in a position where once an allocation has been made via the Barnett formula to Wales, that money becomes Welsh money. It's paid into a bank account—stick it in British Government bonds and have it with a board of trustees or something, and it's there. So, if the Welsh Government doesn't spend it, it exists as a reserve. I mean, at the moment, not only is Welsh borrowing limited, but we can't save either. If we don't spend the money, it tends to go back. Now, there is some end-of-year flexibility, and they've actually improved that slightly in the present circumstances, but I think ideally you'd have a situation where you control your own money, and that would permit you to build a reserve up. But I think, at the moment, there is just the short-term borrowing facility that we have, and I think it could be a bit bigger, but I don't have a specific number in mind.

Indeed. It's leading to necessarily much larger budget deficits in the UK, and there should be mechanisms in place that allow the devolved authorities to do similarly in their own territories. At the moment, my impression is—I'm not entirely up to date—that the British Government is making a solid effort to try to make sure that things aren't too bad, but it is in a very ad hoc way.

Diolch, Siân. Right, okay, we've got a few minutes left, so I was just wondering—you mentioned in your article on tax devolution post Brexit that to compensate for a potential loss in EU structural funds, the Welsh Government could seek tax expenditures rather than just receiving cash grants. I don't know if you wanted to elaborate further on that and how that work might impact on the Welsh economy.

Indeed, I think that's right. At the present time, although it comes to the same thing, really—tax expenditure is an expenditure—it somehow doesn't strike people in that way, so I was thinking, possibly, it might be easier to extract that sort of concession than actual cash expenditure. For example, it does seem to me that there is no reason why you couldn't devolve corporation tax on a controlled basis and allow the Welsh Government to give companies concessions. Now, this would have the effect—this is a beggar-thy-neighbour policy, if you like. This would have the effect of moving activity into Wales from areas where the tax will be higher, but, after all, that's the point of regional policy anyway. So, you do it to the extent that you think is appropriate for regional policy and no further—not to the point where you're unnecessarily penalising activity in other areas.

It would need to be calibrated and discussed, but I think it would be a reasonable approach, and one where Wales wouldn't have to—. Well, it depends how you do it. If you devolve corporation tax—if you say, 'Okay, we're going to let you keep corporation tax in Wales, and we're going to let you reduce it up to this point and no further, because beyond that we think it's inappropriate', Wales would, then, suffer the effect of the tax reduction, because the tax has been devolved. If the British Government said, 'We're not going to devolve it, but as a matter of regional policy we will set it at a lower rate in Wales', then, of course, that's jam, because you get the lower tax and they're taking the revenue hit, but I don't suppose they'd do that.

Of course, they wouldn't have been allowed to do that in the EU, because that would've counted as state aid. They would've had to devolve it so that Wales took the hit for it to be consistent with EU regulations. But now we're out of the EU, they would have the option, if they wished to do it, of just announcing lower rates of corporation tax in the regions.

Although potentially subject to an internal market Bill as well, but that's a whole other—

Well, yes.

Mike wants to come in and then Nick, and then we'll have to conclude, I think.

The Azores judgment. Corporation tax is almost a voluntary contribution by major corporations. One or 2 per cent is the actual amount they're paying, as opposed to the amount that they're meant to be paying. The other point about it is that, with corporation tax, during COVID no company has been making a fuss about corporation tax; they're worried about council tax and they're worried about VAT, because those are taxes that are not easy to avoid. Corporation tax is, to many large multinationals, the amount that they decide to pay. 

Well, it's true that a large multinational with activities in many countries has got a lot of latitude in deciding where to pay the tax, and therefore how much to pay. But I think there are quite a lot of companies in the UK that are not in that position, so I think it would have an effect if you had a lower rate of corporation tax.

Nick, then—final question.

Thanks, Chair. Yes, it just seems to me that corporation tax would be one of the more useful taxes to be devolved to an area, to a country like Wales. We've been discussing this for so long with you, Gerry, formally and informally, actually, over the years, and successive Governments have been very reticent to devolve corporation tax. What do you think is the core reason behind that reticence and do you think it is movable in the longer term? As you say, you could control the rate, the amount by which Wales could alter the rate of corporation tax so it wouldn't be too harmful to the English economy, and there are other factors at play as well, aren't there, other than just the corporation tax.

Well, you'd need formulae to determine where the company was located. You know, they could put a brass plaque anywhere—put it in Jersey. So, you have to have a formula that said where are people employed, where are your sales, where are your investments, where is your plant, where is your head office. There would be a lot of factors like that and you can imagine that could lead to a lot of litigation and administrative nonsense, so I think that's why they don't want to get into that. It's much simpler just to have everybody pay the same rate. But these things aren't insuperable. That sort of thing happens in the United States, where taxes have to have exactly that kind of formula to determine where tax liabilities occur.

There we are. Okay. Well, time has beaten us, I'm afraid, so can I thank you again, Professor Holtham, for joining us once more? We appreciate the evidence that you give us, as always, and you will be sent, of course, a draft of the transcript just to check for accuracy. So, with that, thank you very much. The committee will now take a short technical break, and it is very short. It will be about four minutes, because we're reconvening at 15:20. Diolch yn fawr.

Thank you. Good afternoon. I'm very happy to do that. I did want to start by attempting to greet you in Welsh—prynhawn da. I do hope I've got that broadly correct.

Put in very simple terms, HMRC's role in supporting fiscal devolution is very much to implement the agreed arrangements for Welsh rates of income tax, and I'm very pleased with the progress that we've been able to make to date on this. I listened with interest to the comments of your previous witness about the incentives we have to get this right, and certainly appearing today before you, I do feel very strongly incentivised to be in a position to be able to report that we are making good progress.

I'm particularly pleased that we've been able to stick to the milestones for implementation on Welsh rates of income tax. We made the necessary changes to PAYE ahead of the 2019-20 tax year, and we've since completed the changes that were required to the self-assessment system as well. I do think that the success we've had so far has been very much underpinned by the excellent quality of collaboration with colleagues in the Welsh Government; both my teams and those officials are in regular dialogue. We are certainly challenged and it's right that we are challenged, but it's a constructive discussion with strong working relationships. This is very much supported by the governance arrangements that we have in place. We have our Welsh rates of income tax board, which is co-chaired; Jackie McGeehan chairs for us alongside colleagues in the Welsh Government. As we go through the evidence, if you'd like Jackie to give you more of a flavour of those discussions, I'm sure she'd be very glad to do so. So, I hope that's a reasonable summary to get us started.

Thank you, Chair, and thank you to our expert witnesses. How do you feel that the actual transition to the Welsh rates of income tax—excuse the harp in the background—in practice went? And how effective were the governance arrangements in supporting its administration?

Thank you very much. As I say, we are pleased that we've been able to make the good progress that we have. Key milestones have been met and the PAYE system was changed in good time, and throughout that period, we had very strong dialogue with Welsh Government officials and our governance arrangements have been operating well. We're now in the business-as-usual phase with things now running as we wanted them to be, and in the process of continuing to improve matters.

That's not to say, of course, that everything's been completely perfect; there have been teething issues, there have been transition issues and we have more work to do with employers in particular to make sure that they're consistently applying the tax codes that we issue to them correctly. This is, however, a UK-wide issue. Obviously, it's particularly important that we get that right in the devolution context.

Okay. Thank you. So, with regard to your administrative and IT systems, I would gauge from what you said that you feel that that has worked, overall, in a productive way. But I'd also like to get your comment really in terms of your position around the service level agreement and the memorandum of understanding. How do you feel that is bedding in? Obviously, as you say, there are teething issues. And then, if I may, what do you understand to be any additional costs that have been incurred by Welsh Government?

I'll hand over to Jackie to answer that question. 

Thanks, Ruth. So, I think there's—. Can you hear me?

Yes, we can. 

Great. I think, in terms of governance, we had governance initially to implement the Welsh rates of incomes tax, and as Ruth says, we've moved into business as usual, and we have a new governance structure set up around that. The service level agreement has been in place since 2019, so for the first tax year where Welsh rates were applied, and we've updated the service level agreement subsequently to apply to the current tax year. The update was really around finance, and particularly focused on ensuring that there's ongoing dialogue with Welsh Government about the costs that we are incurring, the choices that might be available to Welsh Government. Where there are options, then those are laid out; we have regular discussions. But also bearing in mind that Welsh rates of income tax are administered through the UK income tax system, so there are limits to how far we can move away from that.

But we have a governance board, the Welsh rates of income tax board, which, as Ruth said, I co-chair along with Anna Adams who is my opposite number in Welsh Government. We meet quarterly with Welsh Government and HMRC officials to look at, essentially, how we're doing against the service level agreement—are we delivering, are we identifying Welsh taxpayers, are we collecting the right amount of tax from them, and are we making sure the Welsh Government gets the right amount of tax, and are we providing data where we should. So, I think our ongoing discussions in that board have been very positive. I think we are delivering against the service level agreement. The service level agreement itself is signed by Andrew Jeffreys from the Welsh Treasury and Carol Bristow from HMRC, who is director of individuals policy. They meet every six months, again to review how we're doing against the service level agreement, and the meetings have been very positive again in terms of assessment of our delivery against the service level agreement.

I think the other thing that Ruth mentioned is employers. So, I think that's the only area where I feel—. It's not that we're not delivering against the service level agreement; we are doing absolutely what we should be doing in terms of making sure we've got address data right, making sure our records are right. But there's a bit to do in terms of engaging with employers who are still finding their feet and getting used to devolved income tax. But we are delivering against the service level agreement. 

Okay. So, in terms of that last point—and I would expect you to obviously be delivering against the requirements of the SLA—but in regard to who's job it is to engage with those employers, how would you describe the work stream that's ongoing? Because it's a big issue in that regard, isn't it? That's a difficult one to crack. 

It is, but I would say it's not a very big issue. We've made huge progress in bringing the number, the proportion of employments where employers applied the wrong code down. It's come down quite dramatically, from something around just over 10 per cent initially, down to less than 3 per cent of employments where the C code has not been applied correctly. We've been engaging with the employers themselves to educate them about the importance of applying the code that we give them. So, that's worked well. As I say, the numbers are coming down dramatically. We're keeping an eye on that. It's not just about Welsh codes. Employers will get the codes wrong for all sorts of reasons, but we do keep on top of them. We spot where employers are repeatedly getting it wrong, and our teams will go out to them to see if they need any help. But it is coming down—the rate of mistakes is coming down—and we expect that to continue to go down. 

Okay. And, finally—my last point—what are the additional costs to Welsh Government that have been incurred?

I can set out the position on costs, and I should start by saying that our commitment is absolutely to make sure that we're monitoring costs extremely carefully, securing good value for money, and making sure that there's a fair and accurate recharge to the Welsh Government, which of course is also subject to National Audit Office scrutiny, which we feel is important too.

The costs fall into two main parts. There are the implementation delivery costs for the IT changes that we've been discussing, and then there's the ongoing running cost. So, in respect of the IT delivery, we have further now narrowed our forecast, and I'm expecting the total cost to be in the range of £8 million to £9 million. You'll have seen from our annual report from last year that £7.5 million of those costs have already been incurred. So, we're now looking only at our final delivery, which is some changes to the annual tax summary, to reflect Welsh rates of income tax, and then we'll be into the project closure phase. So, we're right at the final stage of that process, and we'll be working very closely with Welsh Government officials to finalise those implementation costs by early in the new year, I would expect.

And then there are the ongoing running costs. For the current financial year, we're estimating those to total £700,000. At present, we are on track to be within that forecast. We share real-time—

Can you articulate, for clarity, when you say 'several hundred thousand pounds', how many severals?

I was attempting to say £700,000. Thank you for checking—£700,000, yes. Absolutely, we need to be precise about these things. So, we're currently within that forecast. We share information monthly and discuss that with Welsh Government officials, so we give lots of opportunity to challenge those costs. And these are the net additional costs that we incur in order to meet the obligations set out in the SLA, which Jackie detailed earlier. We're also expecting that those costs will stay at the same level next year, so we're expecting to hold that forecast flat for next year.

Finally, if I may, Chair, just so that there's clarity, there's £1.5 million additionality to the implementation costs. Is that correct, or have I read that incorrectly?

So, in the implementation costs, in the annual report, we said that we had already spent £7.5 million. We're estimating the total costs for the project now in the range of £8 million to £9 million, and we will finalise that early next year, in the early part of next year.

And I suppose the question really is: that will fall to Welsh Government, or not?

Yes, it is for the Welsh Government to pay those net additional costs. Yes.

Okay. Thank you, Chair.

Thank you, Rhianon. Nick Ramsay.

Diolch, Cadeirydd. Good afternoon to you all. The UK Government's report noted that the governance arrangements are continuing to oversee the final changes being made, including to HMRC's self-assessment systems. Can you elaborate what these final changes are and how they'll impact on the writ?

Yes, of course. So, those changes have now been successfully completed. And they were simply to ensure that we can make sure that all Welsh taxpayers are paying the right amount of tax, and it's correctly allocated, and those changes didn't need to be in place at the same time as the PAYE changes because of the lag in the self-assessment system. So, for the last tax year, people who need to submit returns, the deadline for that is January 2021.

There have been issues with identifying Welsh taxpayers, haven't there, for some time now? And I understand that some Welsh taxpayers were, in error, paying Scottish rates of income tax. Is that something you can comment on?

Yes, of course. That was one of the specific teething transitional issues that I was referring to earlier. And I will explain how that happened, the fact that we put it right quickly, and I will then hand to Jackie, who can set out for us the important ongoing work we're doing to make sure that we continue to identify taxpayers correctly on an ongoing basis.

So, for the arrangements to operate as they need to, it is of critical importance that we hold the correct addresses for Welsh taxpayers so that we can identify them as being Welsh taxpayers, and so, before the Welsh rates of income tax went live, we did a major exercise to cleanse our data, to corroborate that data with third parties, to achieve a high level of accuracy. Having done that, we then put a 'C' to identify all Welsh taxpayers and shared those tax codes with employers. We are then reliant on employers to operate those tax codes correctly, and, as we have been discussing, there have been some challenges with that, but, as Jackie outlined, we've made really significant progress since last April.

Now, what happened with the particular problem you're speaking of is that we discovered a problem with some of the software that some employers were using. As soon as we identified this problem, we worked very closely and quickly with employers and with the software providers to put it right. So, that was a transitional issue. It's regrettable, but it has now been sorted out. But I would like to hand to Jackie to explain how we keep those address records up to date. Jackie. 

Thanks, Ruth. As Ruth has described, before Welsh rates of income tax came into being, we did a significant amount of work to get our address data improved, to complete incomplete postcodes and make sure we had identified as many Welsh taxpayers as possible. But obviously that population doesn't stand still and there is no single list of Welsh residents we can check our data against. So, we do a lot of ongoing work to ensure that our data is as up to date as possible. I think the Office for National Statistics estimate that something like 69,000 people leave Wales for the rest of the UK each year, and 60,000 people move into Wales from other parts of the UK. They're not all taxpayers, but that gives you an indication of the population moving in and out, and, also, people move house within Wales, so we need to keep track of them. 

Research that took place in relation to Scottish rates of income tax showed us that around three quarters of people actually tell us without prompting when they've moved and what their new address is. But we need to make sure that we keep our address data as accurate as possible. We will be running another exercise, similar to one run in 2018, where we compare our address data with a third party, and look at where there are discrepancies, writing or trying to contact the taxpayer and ask them about that to confirm their correct address, so that we can corroborate addresses where we can. We did that before 2019. We could then corroborate 98 per cent to 99 per cent of addresses, so that's a very high level of accuracy. We obviously want to maintain that. So, we communicate with our customers, where we can; we can put things on our tax summaries, on their self-assessment return asking them to confirm their address and update their address; we get information from returns from employers on employee addresses, so we can check against our own records and update where necessary. So, it's an ongoing process, but a process that we think is maintaining a very high level of confidence in the address data that we have. 

Just to come back on that briefly, Chair, that's a very difficult thing to do, isn't it, to keep that database? Given the amount of movement that you have across the UK, it's a very difficult thing to keep that database up to date. 

Certainly, yes, it is an ongoing challenge, and it's a really, really important part of the SLA and of our ongoing work. So, as Jackie was saying, it is good that the majority of taxpayers will, without being prompted, let us know if they do change address, but we need to keep focused on this and make sure that we keep the records up to date. And, as Jackie was outlining, there is a range of different ways in which we do that. In particular, we have a range of digital prompts. So, if you go into your personal tax account, you will be prompted to check that your address is still up to date, and it's also the case whenever we submit a self-assessment return—those taxpayers who need to—we are asked to confirm that our address is up to date. 

There'll always be an element of error, I'm sure. Do you have an acceptable level of coding errors in a file somewhere? And how close are you to that level, really, because there will always be changes coming and going, but it's about minimising it as best you can. But is there some sort of acceptable level that you have that you use?

So, we don't have an acceptable level threshold—we're always working to get these systems as accurate as they possibly can be, while recognising that the position for employers, for employees, for people is changing all the time. But, you know, we think we have strong systems in place to stay on top of the situation.

Okay, thank you. Moving on, then, to Mike Hedges.

On the communications strategy, how successful do you think it's been that people know they're paying a Welsh rate of income tax as well as a British rate of income tax? To the average person, it doesn't matter very much, because it's exactly the same amount.

It's certainly the case that, when we launched the Welsh rates of income tax project, when it went live at the beginning of the last tax year, there was a very, very strong communications push. I know this for a fact because it was all over all of my social media feeds, which was absolutely fantastic to see, in both English and in Welsh. As you say, at the moment, with the rates being the same, possibly not everybody focuses on it as much as they might. For us, the absolutely key part of our communications approach, for all the reasons we've just been discussing, is to keep reminding people of the critical importance of keeping their address details up to date. And, for us, that is the key communications plank to make sure that the whole system is operating as we need it to.

The main way in which we monitor our overall approach is to track contact levels—to track the numbers of people who are contacting us with queries or any confusion about the arrangements. We report those levels of contact in both our annual reports and the quarterly business intelligence report that we share with Welsh Government officials. And those levels of contact and confusion are very, very low indeed, which suggest to us that we've got the approach right for now. Obviously, as and when issues arise, Welsh Government officials raise those with us and we work those through.

Okay. Thank you.

Thank you, Mike, Siân Gwenllian.

So, thank you. A very important question, and we have lots of contact and lots of activity in this area, and there are different parts of this process. So, we have arrangements in place to monitor the current performance of the PAYE arrangements. We have arrangements in place to calculate the outturn and we have arrangements in place to feed into forecasting, and I'll run through those. We have in place devolved income tax analytical working group, which works through all of this, and we have an agreed work plan in place that's overseen by that group.

So, on monitoring, the key information that we share is real-time information data from the PAYE system, which we share with the Welsh Government on a monthly basis. That provides real-time information about employee liabilities, and provides, for example, a breakdown across the tax bands. So, that's an incredibly important source of information. However, we can never expect that to be 100 per cent accurate in real time, primarily because until the self-assessment returns are completed at the end of the year we won't have the full picture, and also, of course, there are ongoing in-year adjustments, as we've been discussing, as well.

In terms of calculating the final outturn, I'm very pleased that, last week, in our 2020 annual report, accounts and trust statement, we published our first provisional estimate for the 2019-20 outturn. That figure stood at £2.0 billion, which is very close indeed to the Office for Budget Responsibility's estimate, which they provided in March 2020, and we will now be working, as ever, very closely with Welsh Government officials to produce the final outturn position in our trust statement for 2021, which we expect to publish next summer. These outturn figures are, of course, very carefully looked at by the National Audit Office. So, I think that part of work is going well.

Then, moving on to the data that we share to support forecasting, the main source of information there is a survey of personal incomes, and there we need to work within the constraints of our legal obligations to taxpayer confidentiality—so, we provide what's known as a public use tape, which anonymises some of the wealthier individuals within the cohort to make sure that no personal confidential data is visible there. But, in essence, that information provides everything that the forecasters need, and it's exactly the same information that we share with the Office for Budget Responsibility and the Scottish Fiscal Commission.

Alongside that, I just want to—. Sorry, this is quite a lengthy answer, but it's a really important area. Two other things I wanted to mention, because they're important. One is that there have been ongoing discussions on the Welsh rates of income tax ready reckoner, which our analysts provide. We're aware that we can make some improvements to that to make sure that we're providing for potential behavioural impacts as rates potentially start to change, and we've now agreed a way forward to progress that work.

And finally, we are also in discussion with both Welsh and Scottish colleagues on a new project to put in place a longitudinal data set to make sure that we can track the impacts of introducing devolved income taxes over time. So, we have a really comprehensive work programme in this area. Thank you.

Thank you. So, from our point of view, we've worked through this area in significant detail with Welsh Government colleagues; we have obligations set out in the service level agreement, and we're of the view that we are providing all the data that is needed both to monitor income tax as we go along and also to feed effectively into forecasting. As I say, we are constrained by taxpayer confidentiality, so we do need to work through these issues very, very carefully. Our assessment is that we have comprehensive plans and we're providing what's needed. However, if there's any sense that that's not the case, we remain open to further discussions. 

Sure. Thank you very much indeed. As HMRC's additional accounting officer for Wales, I completely understand the importance of making sure that we spend the money very carefully and wisely and keep costs down to an absolutely appropriate level. As I was setting out earlier, the running costs for this year, we are forecasting to come in at £700,000 for the whole year, and, as I set out, we have arrangements in place to make sure that we share the detailed underpinning information with the Welsh Government every month. We are open to be challenged on those estimates, and then, before any money at all is paid across, it's scrutinised by the Welsh rates of income tax board. We share the detailed breakdowns in advance, we encourage challenge, and it's only once the board has agreed to the statement of costs that we then actually raise the invoice for them to be paid quarterly in arrears. Where there are any new costs that would be incurred that exceed £50,000, we have a special change control board to make sure that any increases of that scale would be subject to a further level of scrutiny.

There has been one recent incident that I've become aware of. This is a positive incident—the term 'incident' can sometimes sounds like something negative, but I do know that, in the final stages of the changes that I mentioned to the annual tax summary, my delivery teams have actually gone back to challenge suppliers on an initial estimate that they submitted to us, and we've been able to negotiate those costs down. So, at this stage, I'm feeling reasonably comfortable that we have strong arrangements in place to control the costs, but, obviously, we'll want to keep that under really close review.

Good afternoon, and thank you for the invitation to give evidence to your inquiry. My responsibilities in relation to the Welsh rates of income tax are all set out in the Wales Act 2014, and it's essentially rooted in the fact that in my role I'm the auditor of HMRC in its UK-wide role and, of course, it's HMRC who collects the taxes across the UK, and has a role, then, in allocating them to Wales through the Wales rates of income tax system. So, the reason that I have an involvement there is because I think it was thought sensible that there's one auditor of HMRC, and then we play our role in giving you assurance about the way in which the amounts are identified that relate to the Welsh rates of income tax, and also the compliance activity that HMRC apply to making sure that that's an accurate figure and the system operates as it's intended to. So, my role is to produce reports, and so far I've produced two reports from 2017-18, 2018-19, and my report on 2019-20 is due to be published in January, in a couple of months' time. Obviously, I'd welcome the opportunity to give evidence to the committee on that report after that's been published too.

Overall, our view is that HMRC has made good progress in implementing WRIT, and largely that's because it's using well-established processes that it applies UK wide, and those UK-style systems apply in this case too. So, obviously, we do a lot of audit work as part of our annual audit of HMRC on all of its processes, so we've got good knowledge of their effectiveness and reliability, and it's from that evidence base that we have an underlying level of confidence in HMRC systems. Given that the Welsh rates of income tax currently remain the same as the rest of the UK, the divergence from the UK-wide regime is limited, and that reduces the risk, in some cases, of different levels of control applying in Wales to the rest of the UK. But, obviously, the most significant area, and one that we've focused on in our work, is HMRC's administration of the system in identifying and maintaining an accurate record of the Welsh taxpayer base, and that's clearly very important, even when the tax rates don't diverge; it becomes increasingly important if there is future divergence. So, that's where we've focused a lot of our work in the first two or three years of this system, understanding how HMRC has established that separation of the Welsh tax codes from the rest of the UK, using the C identifier in front of the tax code to signify 'Welsh' or 'Welsh resident', and I'm very happy to talk about the work we've done on that. But, I think that's the key area of compliance that we've focused on in the first two or three years of our audit work.

Thank you for that. If I could just pursue that a little bit further, are you content with the governance arrangements in supporting the administration of the WRIT, and do you feel that the arrangements are working properly and that the Welsh Government have appropriate input into the process?

So far, the evidence that we've seen suggests that there are effective overall governance arrangements for WRIT. There's a joint project board, chaired by both HMRC and the Welsh Government, with representatives from both working together in the membership of that project board. We've seen the kind of programme management governance that we would expect for something as important as this, with the right level of seniority, the right level of resourcing and so on. So, whilst it's still early days in this system, our view is that the governance they've put in place so far is appropriate and fit for purpose.

Good, thank you for that. I think you mentioned in your report that there are several tasks that are still to be completed by HMRC in 2019-20. Could you perhaps outline what those tasks are and where that leaves us?

Yes, I'll bring my colleague Darren Stewart in here, if I may, who can give you more of the detail on that. Quite a lot of the work in 2019-20 has been on improving the accuracy of the data, particularly on postcode and address data, but if I may bring Darren in, he can elaborate.

Of course. Thanks, Gareth. So, when we mentioned this in the report, there were two specific areas that we were pointing at, some of which, I think, you've heard evidence on from some of your earlier witnesses. So, one is the relief at source project, known as RAS. This is a project that seeks to introduce automatic confirmation of the residency status of pension scheme members. Currently, or previously, that was applied using a manual process, which, clearly, is subject to a higher degree of propensity for error and things of that nature. So, it's really beneficial to try to automate that process and avoid those errors occurring in the first instance. As we understand it, and it's something we're picking up as part of our 2019-20 work, that project is due to be signed off in 2021.

The other, which Ruth Stanier of HMRC spoke to you about earlier this afternoon, is the OTRSA, so, the online tax return for self-assessment project. This is, essentially, establishing a mechanism where Welsh taxpayers can, in a digital way, update their residency status, again framed around trying to prevent errors from occurring in manual processes that would allow errors to creep in. That is something that we are specifically picking up as part of our 2019-20 work, given that it was rolled out in April 2020.

I'm content with that, thank you.

Again, Darren may want to add, based on some of the detailed work that we've done. Overall, our view is that HMRC has adopted a sensible approach to identifying the correct country and therefore the correct tax code to use. We noted in our report that, actually, the challenge of accurately identifying properties on the border in Wales is arguably a bigger challenge than has already been encountered in Scotland, just because of the nature of the border, with a more densely populated border and a longer border. So, we're just acknowledging the scale of the task being larger, actually, in the Welsh context. But we did look at the range of activities that HMRC had been following to identify and maintain an accurate record, including its data scans to ensure that the postcode information that it holds is correct, and testing that against other third-party databases of postcodes to identify anomalies and correct errors. Darren may want to say a little bit more about the detail of that exercise.

Indeed. Thanks, Gareth. I think that gives a sense at a high level of the type of activity that HMRC undertake to ensure that the Welsh taxpayer population is complete and accurate. So, starting with their own database of contacts and ensuring that they are properly identified as Welsh taxpayers, but then also using external data, so for instance credit reference agency data and the electoral roll, to essentially triangulate the accuracy of the data that HMRC hold. I think the acid test for this is where—. I think you're familiar now with some of the issues with employers, in some instances, having applied the incorrect Welsh tax code and prefix. I think the acid test for this is how HMRC was actually able to then respond in pretty quick time, actually, and seeing the overall level of Welsh taxpayer discrepancies actually having fallen quite considerably since that original instance of 174,000—I think, we reported on—incorrect tax codes in our 2018-19 report. That has steadily reduced and got to a much more stable level now.

Darren, would you like to—?

I was just going to say that I think it's probably a combination of both. So, I think, certainly, employers not getting it quite right is how it has played out in the early days. There were some software issues, which HMRC has directly engaged with the software provider about, alongside the employers themselves. I think, to a certain extent, when you roll out a project of this significance and complexity, some of those early teething problems are probably to be expected.

I'm just trying to get to the crux of the problem, because we are told that it's the border and the fact that there's a long border and that that, somehow, is making the situation different to the situation in Scotland, but then we're hearing that, actually, it's because the employers are putting in the wrong codes. So, I don't see how the actual border issue comes into it.

I think it's because of the volume that employers are having to deal with as people move across the border, as they commute across the border. That's just a very significant number of people and therefore it puts the employer, and particularly the employer software systems—because any large employer uses payroll software, and I think the evidence is that some of those systems have not coped well with keeping track of the residency tax code. The fact that the border is more populous than the Scottish version just means that there's more pressure on that system.

I think the exercise you just mentioned in November 2018 is the last major communication exercise that we’re aware of from HMRC. Clearly, they’re better placed than we are to give you an update, and I know you’ve taken evidence from them today. So, I’d have to defer to them on the detail of their plans for further communication. But, in our report, we did stress the importance of regular exercises, including direct communication with taxpayers, all designed to maintain the accuracy of the tax base.

I'll certainly invite Darren's view on this. Darren is the director in charge of our HMRC audit, so obviously works very closely with the department.

Our view is that HMRC takes its statutory responsibilities very seriously and doesn't need kind of commercial incentivisation in order to implement the will of Parliament, including its responsibilities in respect of the Wales rate of income tax, actually. So, there's a level of professionalism in HMRC that I'd be pretty confident applied to this challenge as well as lots of other challenges that it faces. But, Darren, do you have any comments on that point about incentives?

Only to echo what you've said, Gareth. I think that there is a professionalism baked into this and I think the way we've observed, HMRC is actually engaged with this as a process. I think there is something about hearings and forums such as this, the joint governance arrangements with the Welsh Government, that if HMRC were to get this significantly wrong, then it would play out in quite a public way. So, beyond just their general professional approach to administering the system, I think there is an incentive of sorts around getting this right, actually.

Diolch, Cadeirydd. We've talked a lot about pay-as-you-earn. A number of self-employed people and others will be putting in their own tax returns; some of them will be putting them in with a postcode that is—and I look at Nick Ramsay and some of his constituents—an English postcode, and there are also those in the north-east of Wales who will put them in in England with a Welsh postcode. Are you convinced that things are done well enough to identify those people who are filling in their own tax returns and that they're actually being allocated to the right place?

Well, again, I'll invite Darren to comment. I think HMRC's system for checking the accuracy of self-assessment returns, for example, is well established. And clearly, in the case of the Wales rate of income tax, they are focused on the accuracy of the Welsh tax code as well. So, I think their overall processes give you a strong control environment there. But, Darren, do you have anything to add by way of specifics?

I think if we look at the combination of measures that HMRC implements to ensure the completeness and accuracy of the Welsh tax-paying population, both using the information that it already holds through the pay-as-you-earn system, as an example, but also through the self-assessment system, but then triangulating that with other third-party data, I think you draw a level of comfort from the combination of all of those activities. I think it's fair to say that there isn't one single version of the truth around what the actual Welsh tax-paying population is. So, I think it's looking at those as a package of measures and processes and controls that gives HMRC the comfort that it has the right starting point.

Thank you.

Okay. Thank you, Mike. I know, auditor general, that you've looked at the increasing divergence between Scottish rates of income tax and UK taxes, and I'm interested in the lessons that we can take from that here in Wales and how those lessons could be applied to WRIT, of course, if there was greater divergence in a Welsh context.

I think, so far, the evidence that HMRC has gathered in the Scottish context is that there is relatively little evidence of deliberate misstatement of residence, for example, in order to avoid the different thresholds and starting rates that have been implemented in Scotland. I think their view is that the differences so far are quite modest and therefore it's not the economic incentive to potentially not just non-compliant behaviour but potentially illegal behaviour on the part of taxpayers—the gains don't justify the risks, if you like, in that case. But I don't think there's any complacency either, and HMRC are confident that the wide experience they've got of tackling inaccurate declarations by taxpayers will be useful in this context too. So, I think what they have done is, first of all, make sure they communicate the importance of accurate declarations in tax returns and use their general anti-avoidance and communication with taxpayers to reinforce those messages. So, I think, so far, the evidence is that it hasn't presented a big problem in the collection of accurate tax, but, yes, it's the kind of thing that both HMRC and ourselves will keep a very close eye on. If there's any evidence that problems are starting to emerge in a more systematic way, clearly we'd not just want to be alert to that in Scotland but for any implications for Wales too.

Absolutely. Okay, thank you. Alun Davies.

Thank you very much. The NAO published a report last year, I think, 'Investigation into devolved funding'. And the conclusions seemed to be that there was a lack of clarity in some areas of that, particularly consequential decisions and non-Barnett additions and the rest of it. In terms of looking at that today, it would be useful, I think, if we could understand the processes that led yourselves to that conclusion, why you found that to be the case, and how you believe that level of uncertainty within a structure then creates potential difficulties for the Welsh Government.

Yes. I'm joined by my colleague, Lee Summerfield, who led the work on the report that you mention there, so I'll ask him to comment as well, if I may. But, just to start, this was a factual review on our part, from 2019, where, essentially, we thought it would be very helpful for Parliament and for the devolved administrations for us to set out how the system is actually working now. Because, clearly, as you know very well, it builds on the Barnett formula, which is a very old formula, has been there for some time, and has clearly had to be adjusted in various ways as new streams of expenditure have come on. And so what we've got now is a pretty complex mixture of formula-driven allocations, exception categories that don't apply—so, the Barnett formula is not applied to—and others that it does apply to.

So, we thought it was important that we set out how that was actually working, and, maybe not surprisingly, our conclusion was that it is pretty difficult to follow. There are practical consequences of that as well—it's very hard for, for example, the Welsh Assembly to be clear on what will happen if the UK Government decides to do x, because the answer depends on lots of complex applications of the rules. And sometimes it appears to be very unclear how those decisions should be taken in a transparent and fair way.

So, our job is not to comment on policy, and not to tell Government that there is a better way of doing it, but just to be—. We thought it was important in our role to be clear that the system was complex, lacked transparency, and crucially, I think, lacked predictability so that, in making spending decisions in Wales or other devolved administrations, it's very hard to have sufficient clarity and notice, and therefore with implications perhaps for value for money, because decisions that can't be planned a long way in advance are often—it's more difficult to ensure that they are value for money. But maybe Lee would like to add to what I've said.

Thanks, Gareth. I haven't got a huge amount to add. I think that point around uncertainty is absolutely key here, and the timing of some of those announcements. And it would depend on the extent to which that was already pre-empted within the statement of funding policy, which is what drives the decision as to whether something results in a consequential for each of the devolved nations or not. So, therefore, big announcements that might, on the face of it, seem like they were going to result in a significant consequential for one of the nations may not turn out to be the case because, in fact, the money wasn't new money, it was repurposed money, and, actually, it was part of a reserved rather than devolved responsibility. So, it really does create a level of uncertainty. 

Each of the nations time their budgets differently and, of course, that will be out of sync, for various reasons, with when some of these announcements are made. So, you were talking about waiting for some of the big fiscal events, such as the UK Westminster budget, which tends to be in late autumn, of course, in normal circumstances, in order to get any certainty as to how much money and consequentials might be coming towards the devolved nations. So, I think Gareth's point around uncertainty is fundamentally around the timing of some of those announcements.

Which is exactly our experience as a committee, and I think we've made comments and recommendations in some of our reporting on these matters. When you have, I think, what are called now fiscal events taking place on a semi-regular, but unplanned basis—. I accept this year's probably been an exception, but, even outside of this year, there's been an unstructured, it would appear, element to fundamental decision making within the UK Government. And what that means is that, as a legislature, we're unable to scrutinise our Government, because our Government doesn't understand exactly what funding is going to be available to it from different sources right up until, sometimes, after that announcement's been made. And that means that our democracy doesn't work. Is that fair?

I would point you to the report—. Following our report, the Public Accounts Committee in Westminster held a hearing on this topic and took evidence from Treasury officials, and their report, following that inquiry, recommended that the Treasury should share timely information with the devolved administrations on how it was going to fund projects and programmes and be very clear on the basis for that in a timely way to tackle the issue that we're discussing now. And the Treasury, in its response, did agree with that recommendation.

I think, clearly, this year is probably not a fair year to judge successful implementation of that, given the speed at which they're having to make enormous spending decisions on the pandemic. But I think it's definitely one to keep in view and to come back to, because there's a formal response from the Treasury accepting that recommendation from the Public Accounts Committee and it would be good to come back and see what that means in practice when it's maybe a slightly more normal cycle.

Yes, thanks. I'm grateful to you for that. It would be useful, Chair, if we did see that report, actually, and looked hard at that, because I think, from our position, a structure, rather than simply timely announcements, would work best, because a timely announcement, of course, is a very subjective judgment, isn't it? And where those announcements take place on an annual structure, then you can plan around that. What you can't plan around is the Chancellor simply deciding to do something for whatever reason and then there being a knock-on impact around the UK. But, look, I'm grateful to you for that. I don't have any further questions on this matter.

Okay. Diolch yn fawr. Thank you, Alun. And it would be useful if you could maybe circulate a link for that report so that we can have a look at it—that would be useful. We'll move on to Nick Ramsay.

[Inaudible.] Your investigation found that the availability of information and transparency at the UK Government end has negatively impacted on the devolved administration's ability to plan. How do you think the devolved administrations, such as the Welsh Government, can manage their spend more effectively and do you think there needs to be more transparency from the UK Government side of things?

Yes, I think so. I think it's a very similar point to the one we've just been discussing, really. I think the scope for similar-looking spending programmes to be treated in quite different ways in terms of the impact on the funding for the devolved administrations is one of the points that struck us in that work. And, therefore, that's where our current recommendation for clearer and more transparent decision making and more timely communication—that's where that really came from. And we made the point that, if you're on the receiving end of funding allocations in that manner, that makes achieving value for money more difficult, as I said earlier. So, I think the report had sufficient evidence of the problems caused by the current approach for us to make the recommendations we did for a more effective communication of the approach being followed.

Do you think that, at the moment, it's too much that the Treasury—HM Treasury—tell devolved administrations, 'This is what you're getting. This is the way we're doing it', and there isn't enough of a back and forth, a discussion, so that the Welsh Government has it's own—you know, is confident in the allocations made and can confidently explain it? Do you think that that could be a way of proceeding, so that there's much more of a balanced agreement, a balanced arrangement?

Obviously, we're very careful not to stray into policy territory in our role as auditors. So, I think what we're pointing out is plenty of evidence that the way the communication works at the moment has significant consequences. So, what the solution to that is I think has got to be a matter for Ministers and parliamentarians on both the UK and the Welsh Assembly side. But I think our purpose in reporting it in the way we did was to point out what's now the complexity of the way the approach plays out in practice and some of the consequences of that. I think it has to be for the elected politicians in both administrations to take a view on whether there's a better way of doing it.

Thanks.

Thanks, Nick. Could you explain to us then a little bit about how you audit HMRC's costs relating to WRIT and how accurate you think those costs are?

Yes. We followed—. As I say, we have a good understanding of how HMRC records its costs in general and so, obviously, we were then able to look at how they applied it in this case. In our most recent report, which was 2018-19, HMRC recorded £5.8 million in costs relating to the implementation of the Welsh income tax rates, and, based on the work we did, we concluded that that was recorded accurately and fairly, and obviously we're now working on the equivalent figure for 2019-20, which will be in our January report. And so, as always with audit opinions, that doesn't mean that we've audited every penny, but we've gathered enough audit evidence to essentially form a view that that's a true and fair figure, which means materially correct. And so I think you can have confidence to that level that what's been reported is true and fair.

And no issues in disaggregating some of the costs away from UK-wide.

No. We look at the way they do that. They have a service-level agreement with the Welsh Government in HMRC and, as part of that, an agreement to set out the recharging arrangements, and we looked at that, looked at whether the stated approach to allocation of costs had been followed in practice, and the evidence suggested that it had. So, no issues to report to you on that.

There we are. Okay. Thank you so much. Do Members have any further questions? No? No. I think that probably reflects the fact that we're content with what we've heard so far. So, can I thank you—the four of you—for joining us this afternoon? You will be sent a draft copy of the transcript as well, just to check for accuracy. But, with that, I'll thank you very much for joining us and, in fact, that concludes, again, the public part of this meeting. So, the meeting will again, as agreed previously, move into private session. Diolch yn fawr. Thank you.