Cynulliad Cenedlaethol Cymru

Diolch, Gadeirydd. I guess, in a sense, there's no perfect point in the year in which to come into a job like this. The process of my selection and appointment was, understandably, quite a long one, but from the day it was clear that I was the preferred candidate, Huw and Isobel and the organisation were very generous in their time in helping me to get up to speed with all the issues that I'd face on day one, and, obviously, the preparation of this estimate was one of those key ones. So, yes, I undoubtedly had the opportunity to get under the skin of the estimate, to understand it, to challenge and to influence it, so I'm entirely comfortable with it and, yes, it does reflect what I think the organisation needs next year and in the future.  

In terms of priorities, one of the things that I have committed to do in my early days in the role is to meet face to face with every chief executive of every major public body that we engage with. That amounts to something like 75 one-to-one meetings. By the end of this month, I'll have done about 45 of them, so I'm getting an incredible perspective on the challenges that the public sector faces, what they need from us and how they view the work that we do. A lot of that will influence the shape of the studies programme that we will put forward, and we'll publish that for consultation this month. It will go to the Public Accounts Committee in particular to get a steer from them. 

In terms of the estimate itself, I think what the board and I are able to do in presenting the estimate for next year is to have an estimate that provides me with a very solid financial foundation for my first year in office, and I'm extremely grateful to the board and Huw for that. Secondly, and this has come through really strongly in those meetings that I've been having, we are acutely aware of the financial constraints that the whole of the public sector faces, and so we're very clear in our minds that we have to practice what we preach. We have to bear down on our internal costs and constantly strive to do things in a different, better way.

You'll be aware that we have committed to keeping our fees at the same level next year as they are this, and combined with the proposal for funding directly from the consolidated fund, taking those two elements combined, that's a 2.6 per cent real-terms reduction in our financing next year compared with this. Notwithstanding that, I think there's still scope within the estimate, and some of the issues that I'm sure we'll come on to, that are very much focused on what we need to do in the future. So, we face a number of challenges ourselves, for instance the early closure of local government accounts and there are some challenges for our funding that flow from Brexit that we need to tackle, but we also have a number of fantastic opportunities. We're investing money in our communications work to increase the impact of what we do, and in particular in the data analytics project that the committee considered last year.

A number of those issues we'll be picking up on in our questioning this morning. Just another one from me just to start off with. In your 2018-19 estimate, funding was described in the context of strategic priorities. In the 2019-20 estimate, they're described in the context of objectives. I'm just wondering whether you could maybe outline whether that signifies any kind of difference in approach or—you know, what the reason is for that change in presentation, really.

If I may, I'll ask Kev to expand on any change in approach. The first thing that I would say from my perspective, and I'm sure Isobel would reinforce this as well—one of the things that we will say to the rest of the public sector is that they need to be open and transparent in presenting information like this to enable scrutiny by you and by the public at large. So, our estimate does, I think, give a lot of specific detail that allocates pots of funding to important objectives that we have to deliver. As to the change of approach, if I may, I'll ask Kev to flesh that out. 

Yes, certainly, it is a change in approach. We're always looking to improve the clarity and transparency of what we do. So, with the estimate this year, in developing it, we were really keen to align it with the new approach we've adopted with the annual plan that the committee considered earlier this year. The aim, really, is to help stakeholders to understand where we're spending our resources and then what impact that spend has on the delivery of the auditor general and the Wales Audit Office's objectives around being a well-run organisation, around promoting improvement, providing assurance and offering insight. 

So, within the annual plan, we've got 10 areas of strategic priorities and they're all aimed at supporting the delivery of those four objectives. So, we've just done exactly the same thing with the estimate and aligned our spending bids with the delivery of those objectives. Hopefully, this committee finds that revised approach helpful.

Okay, that's very helpful, thank you. David. 

Thank you, Chair. In your figures, paragraph 42 to 46, you talk about an additional £85,000 required for the preparedness for Brexit report, which you've identified as possibly equivalent to two national reports. I'm just wondering why you've come up with that £85,000. You've done two reports previously. Did they contribute to your identification of the extra £85,000 needed in this report? Looking at a report in the summer on preparedness for Brexit when Brexit will have happened by then—I just wonder whether that's value for money.

This is an area of the estimate that I have influenced. Our thinking on what work we should be doing in and around Brexit was at an early stage when I joined. I've very much supported this and put my weight behind this, because I think it's really important that we give a perspective from our position on the state of readiness that we face across the public sector.

So, let me take these in turn. In terms of the precise figure, the £85,000, we estimate that the scale of work that we have in mind for this Brexit piece is broadly equivalent to a couple of medium to large value for money studies—£85,000 is about a quarter of that cost. We did have an internal debate about what we should present to you. I could fund that solely by diverting funds away from the rest of the studies programme to fund that element of work. I could have come to you and asked for the full amount, because it's on top of the routine programme of work; £85,000, we thought, was an appropriate balance between the two. So, I'll be diverting a significant amount of funding away from other aspects of the studies programme to support this. The £85,000 is a significant contribution to the cost.

In terms of the point you make about its timing, there is so much uncertainty associated with Brexit that picking the point at which we seek to do a piece of work is tricky—go too early, and you end up not saying very much because organisations don't know enough. We think we've probably got it about right, so we've initiated a call for evidence amongst all public bodies just this week. That will present us with a baseline of information from which to work. We intend to follow that up with some more detailed fieldwork in the spring of next year, with a view to publishing in the summer, provided, of course, that there is a deal and we're into a transition period at that point. And I think, if we are in that scenario, then a contribution from us in the summer of next year will be of considerable value. And I wouldn't anticipate that that would be the end of the work that we could do, but precisely what follows will depend on the outcome of the first.

Two questions on that: I'm assuming, therefore, you'll be looking at the Welsh Government's transitional fund as part of that work.

Yes, the Welsh Government is clearly a key element of this and we have worked very closely with them up to this point to make sure that we get the balance right, not only in what we ask of the public sector, but in what we ask of the Welsh Government. They're extremely supportive of our doing this work. I think they recognise that we have a unique position and perspective that cuts across the whole of the public sector. So, this is something only we can deliver. Very much—and I reinforce this with the team driving the study itself—this needs to be a positive, helpful piece of work. This is not an exercise in trying to pick holes in what any public body has done in terms of preparedness or what the Welsh Government has done in terms of its communication. The timing will be intended to assist the whole of the public sector, at an important time post departure from the EU, to carry on its work.

And you also highlight in this that you might need to come back with a supplementary budget if there is no deal. We might well be in that knowledgeable position in a month's time.

This is a reflection of us facing as much uncertainty as anyone. If there is no deal, then, clearly, fieldwork in the spring and publication in the summer is pointless. So, we would then be faced with having to make the most of the written examination of materials that we've just initiated, and follow that up, as quickly as possible, with some further work, so that we could make a contribution at the very start of 2019 that would be of some value. The reference of the supplementary budget is a reflection of the fact that we simply don't know what that would entail and the extent to which I would be in a position to redivert resources from other areas to support that. So, we just wanted to put a marker down that there is a possibility that, if we are in that situation and we can't find a way to absorb it internally, we may need to come back to the committee at that point.

So, your anticipation is that, if there is no deal, the expectation would be that there would a greater workload and a shorter space of time to actually address some of the challenges that will become more apparent quickly.

Exactly.

Okay. And the other point you highlighted is £50,000 extra for communication—to implement internal recommendations around transforming communications. I just want to, perhaps, tease out a little bit what you actually mean by that.

What's it for? Yes, on the detail, if may, I'll ask Kev to pick it up. Fundamentally, though, this is one of three transformation projects that Huw initiated last year, and that I know the committee is aware of. The stage we're at at the moment is that that internal piece of work has been completed. It's come up with a number of recommendations that we think will help us to improve the impact of the work that we do and the way that we communicate it. That sum of money is intended to help us fund the practical implementation of the recommendations that flowed from that piece of work. But Kev can say a little more on that.

Yes, I think in terms of some of the practical things that we're looking to do, we're certainly looking to make our reports and other communication channels much more accessible to our stakeholders and have much more impact. So, examples of the sorts of things that we've been looking at are: we produced an interactive summary of our own annual report and accounts in the summer, giving easy access to key pieces of information about us as an organisation and the year that we'd just delivered. That's led to quite a significant increase in the readership of our annual accounts. We've also got products like the recent management and follow-up of out-patient appointments report, where we had a short-form summary report alongside a short video that communicated the key messages to stakeholders in an easy-access format so people could very quickly see the outputs and outcomes of our work. We're very much looking to build on that. Another element that we're looking at is our actual house style, looking to make our reports shorter and sharper more generally, but also extending the use of plain English, and of course that will involve and require staff training. 

Going a little bit further forward, we'd be looking to build on that work to inform, possibly, changes to our own website, and also our positioning as an employer in the marketplace. Certainly, in terms of our website, you'll see from our interim report that one of the areas where we need to improve is in terms of the impact and use of our website, where some of the stats at the moment are below the target that we've set. We very much see the delivery of this phase of the communications project as helping us to meet the ambitious targets that we've set. Of course, all of that work would be subject to a detailed project plan with a benefits realisation table, so that we can plan effectively to realise the sorts of benefits and improvements that we're looking to make.

So, effectively, to make it easier for people to understand what you do, and to, hopefully, get a little more people.

Exactly, yes—just really to make it easy to understand, but extend the reach of our work—greater engagement with stakeholders right across Wales.

That's something I would have thought would have been done a long time ago, in one sense, because that's important—that people understand what you're doing. I'm not arguing with you about things; this is an issue I think we need to get addressed.

Organisations are constantly striving to improve their reach and how they communicate, and one of the things that have struck me in my time in the role is the incredible amount of information and data and insight and expertise that the office possesses. And so, I think if we can unleash that—and I may touch on this when we talk about the data analytics work as well—the potential is enormous. So, I think the office is influential and has considerable impact already, and always has done. But we can go much, much further., and this is an important strand of that.

Diolch yn fawr. Nick Ramsay.

Diolch, Chair. Morning, auditor general. I'm seeing you with a different hat on, today—myself, that is, not you. A large part of the reduction in call on WCF funds is the cyclical reduction in the national fraud initiative spend. Assuming the 2020-21 estimate will require an increase of £130,000, how does this estimate ensure there will not be a significant increase in funding in future years?

If I may, Nick, I'll ask Steve to answer that, as he's our expert in that field.

Thank you. I think, quite simply, every other year that £130,000 comes in and goes out. It's what we have to pay over to Cabinet Office to support the data-matching exercise in a very secure system across public services. The committee might remember we used to charge audited bodies for this, and the committee supported switching it into our estimate in order to encourage voluntary participation, and in this last exercise, we had two housing associations in there, as well as the Welsh Government and the Wales Audit Office. I think, going forward, what we're looking at, or what we're seeing, is keen interest from Natural Resources Wales and Transport for Wales and the Welsh Revenue Authority, and so, the appeal of that free access is very evident, and, in fact, the consultation responses we had on the fee scheme confirmed that. 

I think I'll just say that it is an oddity that we have to consult on a free fee scale for NFI, and that's one of the quirks in legislation. So, the return on investment—as a finance director, I must highlight this. We spend £230,000 over two years. The return, this exercise, is £5.4 million. That's phenomenal. The previous two years it was £4 million. So, I wouldn't underestimate the impact that the funding of this has. And, yes, we'll come back in two years and we'll ask for it again, and our arguments will be around that return on investment. 

Thanks. And what about the capital side of it? The estimate anticipated a fall in funding of £30,000. In this estimate, the reduction is actually £70,000. What's the reason for the difference?

I think, if I may, I'm going to give some credit to Mike Hedges here, because a couple of years ago—

Thanks. That doesn't happen very often. [Laughter.]

He won't refute that. [Laughter.]

Mike encouraged us to put in place a five-year capital programme, and what that's helped us do, with a lot of challenge and scrutiny from the board as well, is think longer term around our capital spending needs, and then, to revisit those annually as we prepare the estimate. So, I would say that figure could well change every year as we come into the Finance Committee. And it's because we look bottom up as to what our needs are for the following year. Over the past two years, we've invested heavily in some corporate systems, so those are in place with contracts now for the next three to four years. You therefore see a lull in our spend on corporate systems for the next two years and they pick up again in three to four years' time. 

What it's delivering, for example, through re-procuring systems, is that we can occasionally stay with the same system, but just get it for a lower cost, and that's happened recently with our HR system, where the contract value over the next four years is £70,000 less than it was over the previous four. So, I think I'd say the bottom line is that that figure is likely to change every year as we come into the committee, and we try and set out quite clearly what that intended spend is going to cover. 

And it's thanks to Mike Hedges. [Laughter.] One of his comments. 

Finally, you intend to spend £25,000 a year on video-conferencing, which was, I think, up from £15,000 before. I understand why you invest in video-conferencing, and when I came on my tour of the audit office last year, I was very impressed with the facilities you've got. The previous auditor general told me that if other organisations went down that route, then there could be bigger savings elsewhere. But I thought a lot of that infrastructure was already in place, so I'm just wondering what the reason is, other than keeping up to date with most modern technology. It all seemed to be pretty modern and working well when I visited. So, what's the extra spending for?

I suppose it's primarily a combination of two factors. One is in terms of replacing kit before it fails, because, obviously, any downtime is costly to us as an organisation. But, also, we're looking to extend the use of the video-conferencing kit, for which there's been an incredible take-up and it has delivered some real business benefits. So, at the moment, we've got 15 video-conferencing units in Cardiff, in our west Wales office in Penllergaer, and up in Abergele in north Wales. And all of that is supplemented by the use of Skype for Business. So, our auditors can communicate with each other and with clients wherever they are in Wales, whether that's in audit offices or from home. 

The feedback from staff has been incredibly positive in terms of the reliability of this new kit. The days of pixelated images and sound delays are long gone—

Careful. We're still living in those days here. [Laughter.] You're making us very envious. 

The kit we've got is extremely reliable. So, our staff can rely on that, they can manage and plan their time effectively to use that sort of kit. That's helping with things like work-life balance. It's also helping to reduce travel. And in terms of tangible benefits, that's probably one of the key areas. 

Just by way of example, when we took a lot of work in-house from the firms a couple of years ago, we were anticipating around about a 20 per cent increase in our annual mileage, roughly equating to about £200,000 a year. If you actually look at our annual report and accounts for 2017-18, our mileage actually stayed at a steady state. We're actually able to manage the mileage very, very effectively.

It's not just cost benefits, though—it also helps us to manage our carbon dioxide emissions, and, again, if you look at our annual report and accounts, in the environmental statement, our carbon dioxide emissions, actually, reduced in 2017-18. And that, of course, helps us with our Green Dragon accreditation. So, there are a number of tangible benefits as well as the work-life balance and other benefits I've mentioned.

I'll just add, Nick, it is a real feature of the way the office works. We have 270 or so staff, the vast majority of whom are spread across Wales—they're not based in any one of our three core offices. So, we have lots of people working in lots of different locations, and it is just normalised and routine for one-to-one conversations to take place via Skype and for people to join meetings with colleagues from other locations. It's just normalised and built into the everyday operation of the office. It's really good.

I think it would probably be helpful if—. I visited, I'm not sure that many Assembly Members have. So, when we talk about this in isolation, it's not the same as seeing it in practice. Most organisations are so used to the old way of doing video-conferencing, which isn't entirely reliable, and you have got that delay. So, unless you see it in action, as you say, a more modern way of working—. I'm not surprised that it's reduced the carbon dioxide emissions, because it does seem to be much easier to do remote working than it is in other organisations.

It is 10 per cent of the capital budget, though, isn't it? So, okay, you've got a spend this year on renewing and keeping up to date. How long will it be before you need that spend again, or is this going to be for a good few years? 

Certainly, with our capital programme, we're projecting a similar level for the next few years. Our IT replacement strategy is, essentially, replace kit before it fails. We estimate the life of the VC kit to be around about eight to 10 years, so that's the sort of replacement cycle that we're looking at.

Okay.

Diolch. Rhun.

Thank you. I'm going to go back a bit, because, basically, it's so important that the board works together with the auditor general and the senior executives to constantly horizon scan, look at what's coming up, how we can increase our impact. And that gave rise to the concept of WAO 2025. And through that there were three big areas of transformation—the data analytics, the communications and the ways of working—that we see as critical in this period to push ahead with, to ensure we can continue to have the impact we wish to have. So, you're aware of those three projects. They do have a lot of interrelationships, but I thought I'd just explain, if you like, those stages that each of those go through to give you some assurance of the governance around them.

These projects are unusual in that they're formed of cross-organisational teams. So, at every level, they apply [correction: at every level of the organisation, people apply] to be part of it. They report directly to the auditor general, or they have done in the past, and the auditor general helps them scope it, but there is real ownership across that team of the project. So, that's the first stage.

They then move into the second stage where they're encouraged to think bold, think outside the box, go and visit private sector organisations, other leaders in the field on [correction: in the field; it's] research and information gathering—it's the ideas.

And then they corral all that together in the third stage, which is to present their recommendations to the board. Now, these are often people who've never been in front of the board before. The ideas can be quite radical, and it's quite challenging, but also very uplifting for the board to receive those and to show the enthusiasm and care and recognise the effort that's gone in. We really enjoy those sessions. Inevitably, and this is quite hard—to manage expectations, there needs to, often, be a bit of refinement or an adjustment of some of those ideas. Working closely with the senior executives, we come to an agreed set of recommendations and then ideas about what we want to implement.

We then encourage more governance around an implementation plan or, in the case of data analytics, which is a very big investment, a full business case. That's when we come to you to seek any funding we need to make that happen. Assuming we get sufficient funding, we move into the next phase, where we encourage the production of a proper business plan, a benefits realisation framework and adopting an appropriate project management approach.

We have learnt a lot because, as a group of auditors, project management for the more creative projects wasn't embedded, but colleagues can talk about that. The progress monitoring was so important. These quite creative, innovative projects can't be just allowed to go forth. They are monitored—they're monitored both by management committee, they then come to the board, and we report to you both through the annual plan and we give you feedback on how that investment has been used through the estimate.

We're really excited about them. I know my colleagues can talk about them, particularly the data analytics. On the data analytics, you saw the business case itself, and we're now into the project plan and benefits realisation. On the comms project, we now know what we want to implement and we're seeking funding. The ways of working are [correction: The ways of working project is] at the earlier stage. The team has been formed, they're going forth and getting those creative ideas, and they're due to report to our board in March. So, we have got the right governance around them, but we're allowing that creativity to blossom. I'm happy to answer any other questions. 

That's a pretty comprehensive overlook and we'll look forward to updates from you as your monitoring continues. Indulge me a bit, on the data analytics, you discussed it, you say, in this committee before—I'm a new member of the committee—so just give us an idea of what the data analytics transformation is about and the kinds of savings that you are able to make out of that. 

Adrian's very excited about this project, so I'm going to let him take you through that detail.

At its heart, data analytics is about our being able to draw in large volumes of data direct from organisations, be that financial data for our financial audit work or performance data sets for some of the VFM studies that we do. The project itself, which the committee has supported in the past, is a three-year exercise to test the art of the possible and to pilot bits of software to see how much data is in a format that we can utilise, and so on. So, it's been going for six months or so and you'll have seen in the estimates some concrete examples of brilliant pieces of work, I think, that we've done already. But we're just scratching at the surface, I think. This is, for me, one of the most exciting strands of work that we've got under way anywhere, because if we can get to a position where we are mainstreaming this into everything that we do, it will give us some significant efficiencies, because we're just able to draw in data much more swiftly and easily. But it's a piece of work that is about much more than just driving efficiencies; it opens up the scope for us to analyse, present and visualise information in a far more engaging way.

The team gave a presentation to the board a month or so ago that was just fantastic. I'd be delighted if the committee were interested to arrange something similar for the Finance Committee. If we were in a position, along with your research service, to enable you to look at spending patterns of Welsh Government in specific areas, to drill down into that, to look at trends over time, I think it would be a huge benefit to scrutiny and accountability. So, I'm really excited about it. It will take a long time, I think, for us to get to a position where we can do that routinely. 

Are you already making efficiencies through it, though, albeit at an early stage, and is the data analytics work contributing towards the £574,000 savings in 2019-20?

Not at this stage. At this stage, we are relying on additional investment from the consolidated fund to fund the work. It is too early in the programme to be able to say with any confidence what efficiencies we can squeeze from this, but one thing I have been crystal clear about with the team from the start is that, unless we recognise that efficiency and cost saving is an element of the benefits that we want to draw from this—on big projects like this—that has a habit of drifting down the priority list. So, the benefits realisation plan that we have in place has cost saving and efficiency very clearly as one of the benefits we want to drive. But, as I say, I think the benefits are far wider than that and far more exciting than that alone. I would hope that, if I'm in front of you in a year's time, we would be in a position then to give you a much clearer picture of the scale of the efficiencies and wider benefits that we can draw from this and the timescales.

If we can turn to the efficiency savings that you have identified for 2019-20—the £574,000—talk to us briefly, if you can, about the main projects that contribute towards that saving and the basis on which you arrived that figure.

Steve.

I think I'd start by saying that it's not a figure that comes out of the blue. So, we have a medium-term financial plan that looks ahead five years, and every year we refresh that and look at the year ahead. Within that medium-term financial plan, we have an efficiency and effectiveness programme. We call it 'Smarter, Leaner, Better', and there are four pillars in there, those being around people, processes, procurement and property. If I can just say a few words on each, the 'people' is around our skills mix and staffing levels and how chargeable we are in terms of the hourly rates. We touched on travel costs and how Skype and video-conferencing can help in that. So, if someone's not travelling two hours for a meeting and instead they're on Skype and then back to work, then that impacts on their chargeability and helps us contain hourly rates. It's also those sorts of activities that have led us to reduce senior level posts by six over the last two years. So, that's how we mix our ways of working with our overall staffing levels.

The next element is processes. We've touched on data analytics, we've touched on communications. We do ongoing reviews around just making sure that our ways of working are as slick as they can be. This week we started discussing our performance appraisal round for next year, for example, and the lessons we learned from last year are that there were some cumbersome steps in there and how we used our IT system, and we'll look to get rid of those for next year. So, again, it actually frees up time then and comes back to the chargeability point. We've also got a project under way around our performance audit work and how to ensure that we've got a system that not only captures data securely but that the knowledge is then able to be shared across the performance audit practice in a way that doesn't take time, and, at the risk of repeating myself, that comes back to the hourly rate and overall chargeability levels. So, it's just making sure that we're as slick as we can be.

For procurement, I've already mentioned the HR system and what we get from procurement in terms of value for money. So, we regularly use the Crown Commercial Service frameworks and the National Procurement Service, because they've already done the value-for-money testing, so we save time and then we save money in that.

Then, finally, coming to property, we've got three offices: Abergele, Penllergaer and here in Cardiff. When we moved from Ewloe to Abergele, we made sure that savings featured as part of the business case, and, as we look over the next five years to the end of those three leases, we'll be looking again to see what property needs we need going forward and how we can get savings from any future needs. Isobel mentioned the ways of working project that we've got, and we're going to wait to see what comes out of that project to feed into that future look.

So, I'd say that it's a culmination of activities across those four areas that will lead us to achieve the savings target, and it is so important to us because we can't achieve our impact unless we achieve the savings. It's so important that the board regularly sees it as part of my director of finance updates, and it's a key performance indicator in our suite of KPIs as well.

Okay. Thanks for that detailed answer. It's a five-year financial plan. Am I right in thinking, though, that you only publish the figure for expected efficiency savings on a year-by-year basis, or do you have figures looking ahead over the next five years for projected efficiency savings? And, if you do—you're nodding, suggesting you have—what's the profile of those savings?

Yes. It becomes increasingly challenging over future years. In fact, this £0.5 million target is the highest we've set yet. We have to be transformational as we look ahead into how we're working, and these three projects are going to help us do that.

So, what's the profile? Is it gradually going down—are you able to project?

Yes, it's increasing, so the gap between our funding and our expenditure is increasing annually in the years ahead. It's the job of the board and management committee to be sighted of that and to make sure that we're taking action to overcome those gaps.

Okay, but then you actually make the decisions on how to make the savings year by year, in detail.

Inevitably, we refine them more closer to the time, but we certainly look further ahead than simply one year.

And it might be that, in future years, we are not able to contain our fees to the same level, but it's one of the ways that we can offset those incredibly large gaps.

One thing we know almost for certain is that we will most likely lose a significant stream of income that comes from the audit work that we do of European agricultural funds. Provided we have the transition period, we have a little time to plan ahead for that, but that will have a significant knock on our overall level of income, but we know that's coming and so we can plan and make arrangements for it.

What is that sum?

I don't want to stifle discussion on this, but we are up against time. We have 20 minutes and a lot of ground to cover. That is an interesting point and maybe we need to pick up on that later. Mike.

I'm moving on to costs, pensions and the pressure of pensions right the way across the public sector. Schoolteacher pensions increases are causing consternation to a number of local authorities, for example. The potential increase in employer contribution rates that you may be facing—do you want to say anything about that and how you're going to deal with it?

Steve.

I talked about our medium-term financial plan and how we're constantly looking ahead. Well, in early September, another central Government devolved body here in Wales circulated an e-mail and said, 'Have you seen this, guys?', and we hadn't. It was a letter from the Cabinet Office saying that, potentially, we'll be facing an increase in our pension contributions of 7 per cent. For us at the Wales Audit Office, that equates to £0.75 million. We simply can't afford to absorb that. So, I had conversations with other central Government bodies here in Wales and they were actually all equally unsighted. So, this communication came out from Cabinet Office on 7 September and no-one was aware of it before then. That very week, we happened to be meeting our finance counterparts in the other UK audit bodies. The director of finance at the National Audit Office was equally as shocked as to why they weren't aware of this significant increase. So, we are being very open with the committee that this has come out of the blue; we've not been able to plan for it. Our options are that we either come back to the committee with a proposal to increase our call on the Welsh consolidated fund or we look to pass it on through our fee charges, and we'd need to come back to the committee to agree a revised fee scheme, and we'd be having conversations with the likes of local government as to why their fees are going up to fund our pensions. And, as you can imagine, that's not going to be a comfortable conversation to have.

Certainly not with local authorities, which are having to fund schoolteachers' pensions and have been hit in exactly the same way. If I was going to be naughty and political, I would say that it's a bit smoke and mirrors. They gave additional money and then they take it back through increasing pension contributions so the net effect can be negative rather than positive. But I won't ask you to comment on that. 

You've also got pressure on pay and prices. Is there anything you'd like to say further on that?

Again, it's about the assumptions we make. So, at the moment, the assumptions we've made around pay and other inflationary costs are affordable and are set out in our estimate. If there are significant changes to those assumptions—the pensions is one example, and Adrian's commented around the unknown risks around Brexit as well—. So, if costs appear to escalate because of that then, again, we'd need to come back and have a conversation with the committee about that. But what we would be doing is being very clear about what our assumptions were and what the changed assumptions are and what the cost impact of that is.

Okay. So, watch this space. Okay, thank you.

Thank you. Neil. 

Sticking with fees for a minute, your note says that there'll be

'an increase in fee rates to be offset by reductions in audit days and skills mix to keep our fee scales unchanged from 2018-19.' 

I presume that there is no chance that there'll be any compromising of the quality of the audit as a result of reducing audit days.

No, absolutely.

Can you explain why that won't be the case?

We are continuously refining and tweaking our approach to our work in general and also to specific audits to particular bodies to reflect individual local factors. So, that's not an unusual thing for us to be doing. The work is led by the most senior financial audit staff in the organisation, and audit quality is of paramount importance to them, as it is for me. There's been a wider debate about audit work in the public sector, and certainly in England it's an issue of some debate at the moment. I have no intention of it becoming an issue of contention here. So, audit quality as delivered by us must be of the highest standard. So, the refinement of our approach is something that we do routinely, and it's a way that we can ensure that fees are managed appropriately for individual bodies, but we won't do that at the expense of the quality of what we're doing.

So, how have you squared that circle—with greater use of data analytics and so on, which makes the use of your time more efficient, as it were?

Yes. So, there are numerous practical things that we can do on the ground to make any individual audit a smoother and more efficient process in terms of our relationships with audited bodies—the materials that they can provide us with, the timing of that and so on—the mix of staff—so, yes, we can mix up different combinations of senior and junior staff, which doesn't compromise the quality of what we do but does help us to manage the overall charge.

Right, thank you. You had a consultation on the fee scheme. Only seven of the 90 bodies you consulted actually responded, which is a surprisingly low figure, but perhaps if you were proposing to increase the fees substantially you might have had more interest in the consultation, I suppose. So, are you concerned about that low response rate or—?

Not concerned about it. I think you're right; it wasn't the most exciting question: 'What are your thoughts on "no change"?' In the conversations that I've had with chief execs that I mentioned at the start, the subject of fees comes up reasonably often. Hand on heart, I can say that not one of them has said that they object to the level of fees. Many of them have said that they think the fees are fair and they get value for money from them. I suspect that you're right: if we were proposing a 10 per cent increase in fees, we'd have had rather more responses.

So, do you think you need to do any more to engage with stakeholders to increase response rates to such consultations, or is this a one-off situation because of the specific circumstances?

The formal consultation exercise is exactly that: it's statutorily required and we have to go through that process. We have senior staff who were assigned to every audited body explicitly with the brief to be the engagement lead and to have conversations with chief executives and the most senior staff in organisations all throughout the year to make sure that our relationships are productive. And so this is a topic that will be coming up in conversation routinely throughout the year. So, the formal consultation exercise is not the be-all and end-all.

Okay. Fine. Sorry, did you want to add something?

I just wanted to chip in, actually, in terms of—. About two or three years ago we were having a lot of concerns raised by audited bodies about, I suppose, the transparency of our savings and efficiencies and their uncertainty around whether we were providing value for money. So, our response was to be very clear, or much clearer, in our documents around how we see us delivering that, and the latest stakeholder survey responses have told us that we've got better at doing that. But, to give some context, our hourly rates have changed marginally over the last four years. The fees we're charging have pretty much flatlined over that time as well. But, this year, as the committee knows, we've renewed our own external audit contract to audit our accounts. The cost of that's gone up by 20 per cent, and when we look at the Crown Commercial Service's fee rates—so, if we were to contract out through the Crown Commercial Service—some of the hourly rates have gone up by 200 per cent. So, I look at that data and then I look at our own, and I think this is why we're now hearing much less concern from our audited bodies about it. But, at the same time, the pressure is still there for us not to relax and to keep on the efficiency and effectiveness drive.

Because you're statutorily precluded, anyway, from charging more than the actual cost of the services you provide. Last question, then: how are changes to the fee scheme agreed internally? How do auditors and audit managers feed into the fee-setting requirements—arrangements, rather?

Okay. I can take this one. Well, it starts—. The whole process starts in the spring, where the medium-term financial plan's refreshed—the assumptions you've heard about, a bit of horizon scanning, and the senior leadership team is all asked, 'What are the known pressures in the system?' And that all comes to Adrian as chief executive, who looks at all of that. The board then picks it up in June, and we look at all those assumptions, those pressures coming forward, and we give a very clear steer about what we expect the fee regime to look like. So, this year, as last year, we said, 'We would expect, if you can, no overall increase in the fee rates.'

Then, over the summer, the executives take all that information, they look at the cost base, the costing model, allocating costs over chargeable time, the skills mix that Adrian mentioned, whether the savings target is feasible and deliverable, and those investment requests. I stress it's a bottom-up approach: what do we actually need? And then it returns to the board in September as an updated MTFP, an updated estimate, and the fee scheme, as you say, to recover no more than the full cost, the consultation we've also talked about, and then the final agreement is down to Adrian and I—Adrian as auditor general and myself as chair of the board—to lay the fee scheme that you have in front of you. So, that's the process; it's very engaged, with people able to input, particularly around their assumptions and known pressures.

Good. Thank you very much.

Yes, thank you very much. We've got about seven and a half minutes, to be precise, left, so I'll hand over to Jane and then I'll conclude afterwards.

Just very quickly on your equality report, we do note your concerns about gender pay gap and also an ethnicity pay gap—quite considerable ethnicity pay gap. Can you just say if you're addressing— how you're addressing these?

I'm just going to kick off by saying that the board has really grasped this. We had a three-hour strategic debate in September, and I'm going to invite Kev to explain the specific push factors that we've now decided need to happen.

Yes. Thanks, Isobel. There are two key pieces of information that we considered to inform that debate: one of those is our new strategic equality plan, which we published in April this year, which sets out nine specific objectives, and one of them is around the need to address the pay gaps that we've identified. They predominantly relate to gender and ethnicity, as you've said, but also age. So, those are the three key areas that we've said we need to focus on.

Alongside that, we considered the progress report that we've only just published, which was the final progress report on the original four-year strategic equality plan, running from 2014 to 2018. That gave and gives more detail about the pay gaps that we have. We very much want to take a holistic approach to this, so, rather than picking off any one particular area, we think that a number of the solutions, say, to improve the gender pay gap, may well be useful in addressing the ethnicity and age gaps.

The board has a very strong view that there are moral, legal, and business benefits to having a more diverse workforce. So, we looked probably at three key areas: one was, 'What have we already done?' So, for example, one of the initiatives that we've got is to advertise all of our senior manager and director positions externally as well as internally. So, that happens simultaneously. The problem with only advertising internally is that you recruit from the same staff mix and you perpetuate what's happening. So, that's one thing that we've already done. Something else that is under way is—you'll see, in our estimate and in our interim report, reference to our people strategy. Under that people strategy, we have a number of commitments, and one of those is to provide opportunities for coaching and mentoring for all staff, but with a particular focus on underrepresented groups.

The third key area is that the board looked to see, 'Well, this all sounds good, what we're doing so far, but what more can we do?'. So, some of the further actions that the board specifically endorsed at this September session were providing more targeted and focused recruitment so that we could increase our representation from currently underrepresented groups, making specific reference in job adverts to the fact that we invite and we welcome expressions of interest from people from the black, Asian and minority ethnic community or from people wanting to work part-time or to job share, and also offering guaranteed interviews to candidates that meet the minimum criteria and who are from underrepresented groups.

Thank you. I'm sure we'll want to keep hearing progress on that in this committee. Also, just looking at progress against the annual plan—because you have, of course, the interim report, which notes:

'There has been no need to deviate significantly from the planned work programmes',

but does this suggest that there has been some deviation, and in what areas would that be? Is it anything significant that we need to understand?

No. [Laughter.]

That's fine. That's fine. Well, it's obviously important that you get through your remaining work for 2018-19 as per the annual plan. Confident that you're going to meet your performance targets?

Yes. Yes, absolutely. I'm very confident that we're on track or have in hand measures in respect of some of the indicators where we clearly need to do more. But there's nothing that is concerning me overly.

And you haven't—. Since you came into post, auditor general, you haven't amended the work programme in any way, presumably.

We're tweaking things all the time, but I haven't arrived and given a fundamental change of direction. As I said earlier, our forward programme of national investigative studies will see the light of day this month. That will be the first opportunity for me to give a more fundamental steer in terms of the nature of the work that we do, and I do have some thoughts on that.

For the future. Thank you.

Thank you very much. A final question from me, then: your predecessor, as auditor general, provided us with a great deal of information around the potential benefits that will be realised from altering the fee regime and from amending legislation, even providing us with a draft Bill. I'm just wondering what your opinion is of that proposal and whether you see that as a priority.

I've seen enough now from this side of the fence to recognise all of the concerns that Huw had. By and large, the burdens of the Act are things that fall on us, I will admit, but they do constrain us quite considerably, internally, in our flexibility and our ability to take innovative approaches or to seek new income streams from contracted work, for instance. There are considerable obstacles to the practical operation of the board, which I suspect were probably not intended at the time. And, I'm conscious, too, that, for you as a Finance Committee, there is a relentless stream of work that flows directly from the Act. So, do I see it as a priority? Absolutely, I can see that, if we were able to tweak some aspects of the law in the way that Huw outlined, it would undoubtedly give us greater flexibility, but I don't think it would prejudice in any way the spirit and intention that lies behind the Act, which is perfectly proper.

Having said that, I'm probably more aware than most about what comes with taking a piece of legislation through this place, no matter how uncontroversial one might think it is, so I think you and Isobel and I need to have a conversation as soon as we can about the committee's appetite and capacity to do that. If there is anything more that we can do to facilitate that process, of course we'd be very happy to do that.

Okay, thank you. Well, the message is clear and you're right, I think we need to consider that further, then, and as a committee reflect on what the possibilities are. Great, diolch.

With that, can I thank you all for joining us this morning? You will be sent a transcript of today's meeting just to check for accuracy and we as a committee now will move into private session so that we can consider and reflect upon the evidence and the estimate that's been presented to us.

So, could I propose, in accordance with Standing Order 17.42(vi), that the committee resolves to exclude the public from items 5 and 8 to 10 as well, when we'll reflect on other items on our agenda this morning? So, are all Members content? Yes. Diolch yn fawr iawn. We'll move into private session. Thank you.

Welcome back to the Finance Committee this morning. We move to our sixth item on the agenda, and that's to scrutinise the Welsh Government's draft budget for 2019-20. We welcome this morning Robert Chote, who's chairman of the Office for Budget Responsibility.

I'll kick off, if I may, with a couple of questions. The first question, clearly, is: what discussions have you had with the Welsh Government relating to your role in providing independent tax forecasts for 2020-1, for that budget, and beyond?

Thank you. First of all, thank you very much indeed for the invitation. It's a great pleasure to be back. Yes, we've had, obviously, detailed discussions about how best we can contribute to the new framework, and the requirements that are in there, obviously, for independent forecasting capacity. In doing that, we're very much building on a very good working relationship with the Welsh Government, and I hope with this committee as well, in the past, in the way that the task of scrutinising and forecasting the devolved taxes has gone to date.

In the specifics of the new role, from our point of view, we have no change in our overarching legislation, and therefore the formal, legal responsibilities. In effect we are being asked, and are very happy, to zoom in in greater detail on a relatively small area of our existing forecast that allows us to fulfil that requirement. We're not in the business, as with Scotland, of producing a separate Welsh macro-economic model and macro-economic forecast. I think there's an interesting question there, as there was for the fiscal commission in Scotland, as to what sort of capacity you need more broadly. But I think it's an issue that probably spans more widely than the specifics of doing the particular forecasts that are required here. There are some particular elements in terms of what goes on in the economy that are very important as drivers, but you don't need an all bells and whistles, fully developed set of national accounts and macro-economic forecasts to do that. But you may feel that it's important to have that for other reasons and want to decide how that gets best done.

We have some agreed principles for how we work with Welsh Government and the Welsh Revenue Authority in pursuing this task. I think we're very much looking at the first year as being a test. There's no point in chipping this into stone at this stage. We need to see, for everybody's benefit, how well this works on the first basis of doing it, and to learn from that. In a sense, if you think about the parallels with the way in which we do the UK-wide forecast, Welsh Government officials take on the element for us of the revenue and customs, Department for Work and Pensions role in that context, in terms of the fact that they're working on developing models and analytical approaches and in some senses turning the handle based on the assumptions and requirements that we independently put into that, and I think that sort of shared relationship is one that works well. It's a good source of internal challenge and rigour, and I don't feel that it compromises our independence, in that sense. We'd want to be as transparent about these sorts of things, and, obviously, as we can come here and explain to you what's going on in the numbers, and be responsive to those sorts of questions, I think that's fine.

Second, obviously, to the role of producing the forecasts there's the issue of certifying the costings of individual policies that come to us from the Welsh Government, and there the parallel is pretty close with the way in which we do this with the UK Government, in that the Treasury, when they decide the list of measures that they want to include in a UK-wide budget, will come to us with some draft costings of those. If it's a tax measure, it's basically produced by Her Majesty's Revenue and Customs, but the Treasury will have been involved in that. We give that detailed scrutiny. I think we had something like 80 measures in the last UK budget. It's very rare that the numbers that are originally provided to us are the ones that get through without touching the edges. We're almost always asking for some sort of changes and probing the analytical underpinnings and the data that undergoes that. I've been very impressed at the UK-wide level. I had feared, when I started the job, that you would come in and this would be a naturally adversarial relationship, that the Government would have got all its ducks in a row and have a particular line that it was pushing, and we would be fighting over the lines of that. It's not that case. In this sort of work, you're dealing with enormous uncertainties about behaviour, about the nature of the data you're dealing with, and I think there's a shared desire to get the best understanding that you can out of that.

In terms of broader spending issues, outside the specifics of the sort of measures you get in the budget, we have no new responsibility there. The Welsh Government's broader spending comes under the departmental expenditure limit approach that we don't do—a sort of bottom-up forecast that we're not trying to assess at the UK level, you know, how much the education system is going to cost, how much the health system is going to cost; you take those plans and make judgments about over or underspending against them.

I think, coming forward, as I say, we're very much treating the first year as a test to see how all this goes. I think one thing that would be useful to get a sense from you on is what additional engagement on the civil society side you feel would be helpful. I think we're hoping to, as it were, make an additional visit in having discussions with, for example, colleagues at Bangor who've been working on the forecast to date, but there may also be people you feel would be helpful for us to engage with more systematically, given where you see the particular areas of expertise that we may miss in some of those forecasts.

Then, finally, on the publications, the aim is to produce something alongside the draft budget and final budget, pretty much on the line, I hope, of a relatively transparent explanation that we provide in the devolved tax documents already. My one hope here is that the UK Government hasn't got a taste for having October budgets at Westminster, which I think would make all of our lives much more complicated if it continues on that basis.

Indeed, and that was going to be my next question, actually, because you have indicated previously that there would be challenges to providing a comprehensive forecast analysis, given the timing of the Welsh draft budget and the UK budget. So, what impact could that potentially have, then, on the level of detail, for example, that you'd be able to provide to us?

It would make the task of producing dedicated documentation that little bit more complicated. I hope that—. This was, for a variety of reasons, quite a difficult forecasting process at the UK-wide level that we've just been through in terms of compressed timetable. I hope that the document that we provided on the devolved tax forecasts has not suffered too much from that. I can see some bits of cutting and pasting from paragraphs that we would have squeezed out if we'd had another week. But it clearly causes an issue in terms of your policy response over a relatively short timetable if there are small gaps here, and it's just going to make life that much more difficult. But my working assumption, touch wood, or Formica, is that this is related particularly to the Brexit timetable this year, and that we're likely to move back to something that's more in the usual late-November timetable, but I think we have to wait and see.

So, when you told us last year, or this committee last year, that it would be a challenge, you've reassessed somewhat, actually, the circumstances that might have transpired at that time, and that it isn't as bad as you thought it might be. Sorry—that's my reading of what you're saying, effectively.

Well, I think, clearly, if you're having the draft budget and the UK budget very close together, that poses issues for both the Government's response, how the sequencing of that works, and our ability to be able to, for example, understand clearly what the assumptions have been in the draft budget. This is in the framework when we're not doing the forecast, but I think we've still managed to look at the numbers that the Welsh Government have produced, and I think satisfies us that, in fact, the forecast differences, as you'll have seen, are not that large. But the process that we go through of talking at the moment at challenge meetings that involve the Welsh Government, but also the Scottish Fiscal Commission, the Scottish Government and the devolved tax experts at Her Majesty's Revenue and Customs, et cetera—that process works well, and I think we have a good engagement and a good exchange of views there. So, hopefully that wouldn't be compromised by the difficulties of timing, but, of course, it would mean that the Welsh Government officials with whom we'd be interacting, and the Welsh Revenue Authority, perhaps would have other things on their plates at the same time.

Sure. Okay, thank you for that. David.

Can I just take that back a little bit, just for my clarification? I appreciate that the UK budget came forward this year, and there's a tight timescale, but last year there was a very tight timescale between the UK budget and the final budget of the Welsh Government, so I'm assuming that you'd have similar—. Although the bulk of the work will be done, and the forecasts are draft, there's still quite a lot of consideration to be done in that very tight timescale between a UK budget in late November and the final budget in early December.

That would be true, and the difficulties and the challenges that creates obviously depend on how much substantive policy change there is in the draft budget that we would need to take account of in the final forecast. So, as you say, you're kind of constrained at both ends and, in an ideal world, you'd have the maximum gap between the two. But whether you end up in that world or not is out of our hands. But, yes, it's—.

Okay. That's interesting. You've had a memorandum of understanding with the Welsh Government, and I understand you're updating that memorandum of understanding. Is it ready to be updated? I know it's not going to be ready, or published, until April of next year, but where are we with it?

Well, we had hoped—. In the original expectation of a rather quieter October than we've all had, the hope was to have made a bit further progress on that, but I think, as I said, the working relationship works well, so I don't see that any updating of the MOU is in the light of a problem that needs fixing. So, I think we just need to make sure that there's enough time on both sides to ensure—. The other point with these MOUs is you have to see how they're operating in practice, and therefore the notion that these things are written in stone the first time you do them would be foolish. 

I would expect them to be reviewed after one year, anyway. 

Yes. The key thing with MOUs and with these sorts of working relationships more broadly is to have a shared agreement and understanding about when information needs to be exchanged in order for everybody to be able to do the jobs that they have to do under the framework to their own satisfaction. The challenge that we had with the UK budget that's just passed is that, notwithstanding the fact that we had well-written-down deadlines for when things need to be notified to us, a lot of them weren't notified to us in that timetable, and that makes life more difficult, and you come back and revisit that with the knowledge that that's there. I think that the scale of the packages that we're likely to be talking about in our engagement with the Welsh Government is not going to be over 80 measures, so you don't have that sort of challenge. And, in the UK context, there are particular issues around sets of policy measures that require the Treasury to play nicely with the Department for Work and Pensions, which sometimes works very well and sometimes merely works well. So, that, again, I hope, would not be a situation that arises with quite the same level of difficulty here. 

Well, you never know. [Laughter.]

Let's put it this way, I wouldn't take it as a model to emulate. 

You've also said, I think, if I heard you right, that you've already had discussions with Bangor University, who obviously were doing the forecasts previously. Have you learned anything from those discussions that might actually clarify the situation of the memorandum of understanding much better?

Not so much with the memorandum of understanding. We've been involved; we've participated in the challenge meetings with the Welsh Government and Bangor, and the way in which the Government officials have conducted those reflects the way in which the challenge meetings that we've been operating on the devolved side already have worked well. So, my colleagues who have attended those say that they work as you would expect, and as well as you would expect, and I think that that level of external scrutiny and challenge has been useful. I think one thing in terms of moving to the new arrangement, where more of the forecasting responsibility falls upon us, is that you don't want to lose that input of external challenge, and this is, as I said, one of the reasons why I think we need to think about the level of engagement we're having with people you think we ought to be having it with outside the core period of the forecasts as well so that we're, as I say, sighted on data or information or analysis that you think would be relevant to us that we might not otherwise—or that people at Bangor, for example, might think is relevant but we wouldn't pick up. 

Because, obviously, Bangor last time were very concerned about the Wales-specific data under their work, so I'm assuming that that's something you'll be looking very carefully at as well. 

Yes, and of course we're now seeing the Welsh Revenue Authority specific data. If you can distinguish here between the data and forecast inputs at the broader economic level versus the data on how the tax streams are performing in real time, and there we do have that and we're about to draw on that. So, for example, I think on the land transaction tax, we're obviously able to base the last forecasts that we've done on five months of outturn data for 2018-19, and we can look at it in that context. The more challenging one has been on the landfill disposals, where we've basically got one point of data that has come in quite differently from what was anticipated. We've placed quite a lot of weight on that, on the grounds that the previous set of forecasts had a lot of assumptions, necessarily laid upon assumptions, and so we've put quite a lot of weight on that single data point. Therefore, we have—and I would highlight the uncertainty around that. 

On income tax, obviously, we are some way away from the point at which we have the 'C' flag that will, obviously, change the nature of the forecast and the evidence base for that. I suspect we may come on later to what we've seen in Scotland in terms of when you start that flagging process and you end up with a very different picture from what you've anticipated beforehand. There's a job of work to be done understanding why that has happened and, fortunately, they've gone first and acted as guinea pigs as to whether there's anything there that we need to think about in the context of when the same comparison of the survey of personal incomes and the 'C' flag is there as well as the SPI and an 'S' flag in Scotland.   

I appreciate that, and I'm sure my colleagues will ask some questions on that. But I suppose, in one sense, since you talked about income tax, this is the first budget that we've had with income tax as a factor. The UK budget has now come forward with some proposals. Have you had a chance to look at the forecasts as a consequence of what happened with the UK budget? 

Yes. So, the forecasts that we've produced for the devolved taxes incorporate the policy measures that were announced in the UK budget. So, if you look at our forecasts for Welsh rates of income tax, they are reflecting the fact that you've had, in particular, the personal allowance and higher rate threshold changes. So, that basically means—. To take one step back, the outturn data coming in so far this year has been stronger than anticipated, and that is true across a range of the major taxes at the UK level, not just income tax. You then have a set of policy decisions. The decision on the personal allowance and the higher rate threshold obviously reduces the amount of revenue you get, because you're bringing forward those changes but has less impact in future years. There's also then some offset from the IR35 changes—people working through service companies, et cetera. But we are able to take those into account, and they are broken down in explaining why the WRIT forecast has changed since the last time that we did it. And, obviously, if you're making the comparison between our forecasts and the Welsh Government ones from early October, they don't obviously, include those policy changes. But, as I say, the forecasts are pretty close together. There are moving parts in that, but there's nothing that would cause me concern about the level of difference between those two, even if you tried to adjust without the policy changes. 

Nick wants to come in if, he may, on this.

Just on that issue of the greater outturn than expected, the finance Secretary has told us that, in terms of landfill disposals tax within Wales, there's been a much higher outturn for that tax in the first quarter than was anticipated. Now, okay, it's only the first quarter, but it's still a lot more than was expected. From what you just said, is that fitting into a pattern with a lot of taxes across the UK at the moment, with more of a gain than was expected?  

Well, it's clearly in the same direction, but I think it's a different category. The buoyancy of tax receipts across the UK as a whole, I think, is reflecting a combination of relatively stronger than expected employment growth over the summer. It also may reflect the fact that growth in the cash size of the economy—so, the total amount of spending and income in the economy—may be more rapid than the official statistics are currently suggesting. The Office for National Statistics measures activity in the economy in three ways: you add up all the output, you add up all the income, and you add up all the spending. When you do that, you don't get the same answer from all three of them, and the ONS over the two years after the first data comes in has to squish all these together in a consistent fashion, and one of the things that it draws upon to do that is what you've learnt from tax receipts. And it may well be that, for example, at the moment, if you add up all the incomes, you get a smaller number than if you add up all the spending and you add up all the outputs. So, maybe those things will move in a way that suggests that, not necessarily real activity in the economy, but nominal activity in the economy has been more rapid. 

I think, in the context of the disposals tax, it's simply the basis that we did not have a decent evidence basis before, and you were having to do a variety of linked suck-your-finger-and-hold-it-up-in-the-air exercises to the best you could, and this is the first time we've actually got some proper data. One data point is better than no data point, but you'd certainly be happier with four or eight or 12, and then you can start to build up—

So, it will be interesting to see in the longer run what the other taxes actually deliver. 

Indeed. The other thing on the disposals tax is that one of the issues that seem to be here is that I think the Welsh Government's original expectation, which we regarded as being entirely reasonable, was that a smaller proportion of disposals would be at the relatively highly taxed standard rate than was the case for the UK as a whole, whereas, judging from this one data point, it looks as though it's pretty close to the UK as a whole. Now, the question there is obviously whether that is just a one-off, or does it suggest that there's—. But, for the time being, we'll assume that this is the best we've got to go on, and we can push towards—

Or has the collection been more thorough over a Welsh basis?

Well, yes, that's another very good point, because with the setting up of this, and starting the process, the revenue authority, I think, have managed to talk to the 20 or so people you need to talk to about this, which HMRC is not in a capacity to do. So, you could have seen the whole process of moving to this as something that's effectively reduced the tax gap—the gap between what you are getting and what you should be getting in Wales, relative to the rest of the country. So, that's quite plausible. Now, again, one of the challenges there is: would you expect that level of engagement with the taxpayers or rather with the people who have the information here to be sustained over the longer term, or is that just a fortuitous consequence of the set-up phase?

Okay. There's a lot of ground we need to cover still, so we'll make progress. Nick, your questions are next. 

Fine. Okay. You've mentioned already the lack of Welsh data in your last answer. Bangor University have mentioned that the Welsh Government are looking at the potential for developing a Welsh-specific gross domestic product. Is that a good idea, and would that be helpful in modelling tax forecasts?

Well, I think I would come back to what I said to the Chair at the outset. Clearly, more data at a Wales-specific level is valuable. I think there's a distinction though between what is useful from the perspective of you wanting to have a good oversight of what's going on in the Welsh economy, and that being a guide to broader policy considerations, versus what you need specifically for the forecasting of these taxes. So, obviously, for income tax, employment and earnings growth patterns are very important, projections on population, et cetera; and for the disposals tax, activity in the residential and commercial markets—commercial being a particular challenge because it's lumpy and can be distorted by one-off changes there, and then, the issue of what is it that you want to draw upon to try to reach conclusions about where disposals are going, which is a tougher proposition. 

Now, an overall Welsh GDP measure might be desirable for broader policy-making purposes. How much—. It could be a large statistical hammer to crack a few relatively small nuts. 

So, it could, potentially, be a bit of a red herring, in that, okay, it might be interesting to have, but it wouldn't be directly helpful in the taxes that we're getting. 

Well, yes. To what degree it would be helpful beyond—. In producing that, you would want to be drawing on what's been going on in the Welsh labour market in terms of earnings growth and employment growth, but it's those elements that are the key here. What view you take of the size of Welsh business investment, or the net change in stocks and inventories at a Welsh level, is not going to be terribly helpful. This, as I say, is an issue with which the Scottish Fiscal Commission, in particular, has grappled. And I think they have taken the view that it is a good thing for the broader purposes of their mandate to be producing an economic forecast, and I suspect that that goes beyond what they would think would be particularly necessary for the specific revenue forecast they have. There's also the way in which the fiscal framework operates—there are trigger points, depending on whether you identify a Scottish-specific economic shock. But, again, coming up with a whole set of GDP figures just for that is another, potentially, sledgehammer-and-nut operation. So, it's not to say that I don't think it's valuable, that you wouldn't find it valuable, but, in terms of thinking about who ought to do this and how much effort ought to be expended on it, I would think about it more broadly than the particular requirements of what we have to do here. Obviously, if it's produced, we'll look at it, and draw what useful information we can from it.

You've answered my second question in what you've just said about the other data. What support are you giving the Welsh Government in progressing its overall macro-economic data collection agenda, or what support are they looking to gain from you for that?

Well, we're obviously very happy to talk on the basis of how we use the economic data in the forecasting operations that we do at the UK level and more specifically. There's also a flow going in the other direction that's important—the information that you get out of administrative data on the operation of the tax system can feed into your view of what's going on in the economy; it's not just the other way around. As I said, the fact that we've got buoyant tax receipts at the UK level may be telling you something about the inadequacies of the measurement at the macro-economic level, not that, if we had the macro-economic level, it's telling you that the revenue numbers are wrong—you know, revenue coming in is revenue coming in, so you obviously place relatively heavy weight on that. We obviously are very happy to offer advice and suggestions there. I think, again, this is an area where talking to—for both the Welsh Government and for us—. To talk to the Scottish Fiscal Commission about what they found helpful, and less helpful, in terms of the expansion in their macro-economic work, might be fruitful in this area. But I think that I would focus in particular—. We're going to learn more from drilling into and assessing the robustness and identifying the seasonal and other patterns in the administrative data that we're starting to get in from the WRA. That's a more fruitful seam to mine in the early period than going at the broader macro-economic picture, I suspect.

Interesting.

Yes, very. Mike.

Which brings us on to the Welsh Revenue Authority. You're getting data; we've talked about that fairly substantially up until now, and you use that for forecasting, but, if you just take landfill disposals tax—tell me if I'm wrong, but I would expect that to be relatively flat, quarter by quarter. Whereas, with land transaction tax, sell one £400 million shopping centre and it will create a quarterly distortion. So, there are two views on this, that landfill disposals tax, if it's coming in at—and your prediction was it would come in at £44 million—. If it shows it's coming in at £15 million the first quarter, would you up its expectation to be £60 million for the year, because people tend to get rid of landfill disposals in a fairly flat manner? Whereas, if, the first quarter of land transaction tax, we were expecting about £240 million, and it came in at £80 million, that could well be distorted by one or more very large sales. And, if you take one year, and you expect £240 million, if you sell two shopping centres, all of a sudden—assuming you're not going to sell two shopping centres a year—that will have a level of distortion of £40 million, £50 million. So, I'm really trying to work out how you take their data and then analyse it afterwards.

Well, you've highlighted the very difficult points there, and that there are two sets of timing issues here. There's the question of: do you have a regular, seasonal pattern to the numbers? So, for example, do you expect there to be more residential house sales in the spring than at other times of the year, for example? Then there is the question of volatility, where you can have a very uneven pattern through the year, but not regularly so. And, as you rightly say, commercial land transaction tax, as at the UK level and the Scottish level as well, is a very good example of that. For example, I think, on the commercial LTT side, it now looks, with the benefit of hindsight, as though 2017-18 may have been a bit of an outlier in terms of the amount of revenue that was coming in, and that having assumed that you'd moved to a stronger path might not be a sensible thing to assume going forward. So, you do try to make the best of that. 

One of the challenges with, as you say, the selling of big shopping centres distortion is that you don't know when it's going to happen, but you do know that people are going to sell big shopping centres from time to time, and therefore trying to take into account in a forecast—you know, you have to basically say, 'Well, look, we don't know exactly when this is going to happen, but let's assume that one of these big things happens most years', and you might therefore incorporate that. It doesn't help you very much with your monitoring on a quarter-by-quarter basis, because if somebody sold a big one in Q1 2000 or Q2 2018, will the next one come in just before the end of the current fiscal year or into the next one? But, on the other hand, what you don't want to do is say, 'This is a one-off, therefore we'll ignore it', because you know that there will be these one-offs over the time period coming forward.

This is a classic case, with the data that we're starting to get in with the WRA, of that you need some time to see where the regular seasonal patterns are, and they may not be the same in Wales as they are in other parts of the country. There's also the challenge that, when you bring in a new tax or have changes in policy, or you have the expectation of changes in policy, that, in itself, can change the time profile. So, I think, at the moment, residential LTT is coming in over the first five months of the year a little weaker than you would expect if you simply drew a straight line between the beginning of the year and the end-of-the-year forecast. But, in the expectation there may have been some forestalling ahead of that, you might expect that to be relatively depressed. So, that means that the seasonal pattern that we identify when we get to the end of this year may not necessarily be a very good guide to the seasonal pattern that you would expect in future years. So, you're absolutely right to identify that this is a very difficult challenge, and it's an area, I think, as well, where you rely, in part, on the experts at the revenue authority to give you some judgment, as we do at the UK level, on whether there are particular transactions that they are aware of that would lead to those sorts of distortions.

Now, you're in slightly tricky territory here because of your access to taxpayer confidential information, which we rightly don't have. We can talk directly to HMRC analysts, in a sense, which gives us a slightly better position than authorities that have to rely upon the public-use versions of these data. So, again, it's one of those areas where you're looking to draw on the intelligence of the people who are monitoring this from the implementation and monitoring process as well as at a policy level.

Just finally on that, I think that the position, of course, is, when you're dealing with the whole of Britain, you can confidently predict a certain number of shopping centres will be sold every year because you've got that level of churn, whereas you've got a lot fewer in somewhere like Wales, so then it'll happen infrequently and then you'll have jumps that will be outside your predictions.

It's more difficult, but it's not straightforward at the UK level either. Commercial stamp duty land tax is not a straightforward thing to do at the UK level. You're right; there are just more transactions going on. The classic problem with the land transaction tax and its equivalent in the UK is that it is naturally going to be more volatile and variable than, say, a tax on income or consumer spending will be, because it will be affected both by the price at which the transactions take place and the volume of transactions. I think, again, if you're thinking about, 'Well, let's benchmark our forecasts against other ones', it's very difficult to draw a conclusion about whether your particular forecasting model is performing better or worse when, actually, the relative merits of the forecasting approaches may be swamped by the fact that it’s just very hard to predict what transactions levels are going to be, because they're not hugely closely linked simply to the amount of wage growth that's going on in the economy, for example.

And don't forget at some stage we'll have a recession and land transaction tax will drop substantially accordingly. I know you never predict a recession, but, if you believe in the cyclical way of the British economy, one is inevitable.

That's also true, and you also have the challenge of, you know, if you have a depressed level of transactions, what is the normal to which it is returning? Because you have to make some assumption about that, but it's pinning the tail on the donkey.

I had another question, on the Scottish Fiscal Commission, but I think you answered that earlier. 

Okay. Thank you, Mike. Rhun. 

Looking at WRIT forecasts—and this is all a learning curve for all of us, of course—how do you use HMRC income tax data now for modelling WRIT forecasts? And how do you expect the use of that data to be refined once WRIT is actually implemented, so that the forecasts become more accurate over time?

The challenge here is that, at the moment, we're reliant on the survey of personal income for deriving, as it were, the Welsh share overall. So, that's a sample of about a little under 750,000 taxpayers. It’s deliberately over-sampled at the top end of the income distribution, because of the relative amount of revenue you get there. So, you sample more people on relatively high incomes to do that. Unfortunately, it doesn’t over-sample on a nation-by-nation basis, so it's not over-sampled for Wales in a way that would be useful.

Does it sample proportionately for Wales?

Proportionately, yes, exactly, but doesn't overdo it. And you then use—. So, basically, at the moment, we're basing forecasts on that survey for 2015-16, which is the most recent one that you have, and you then basically say, 'Well, what’s been going on in the economy since then?' And you're, as it were, forecasting where you are now by getting from that 2015-16 point to today. Now, when we do that, we're able to, we have access to—. We don't have direct access to the individual taxpayer confidential data, but we have access to HMRC analysts who have access to that data, which allows you some richness there.

The challenge for the Welsh Government and the Scottish Government as well is that they need to use the public-use version of the SPI, which composites, which bundles together, observations in order to protect taxpayer confidentiality. We will move to the 2016-17 version of that survey, which, as I say, is conducted by HMRC, in the spring.

You also, in addition, have some new information that is coming through, what's called real-time information, which, again, is HMRC; it's collecting more data on a more timely basis, which is something that we've been looking forward to for ages as something that could give you a better real-time handle on what is going on with wage growth and the performance of that. One of the challenges there is it's only PAYE, so it’s, again, under-sampling from the perspective of some relatively high incomes.

The big change will come with the flagging of individual taxpayers as being Welsh taxpayers or not, as we are going through in Scotland, and that, I understand, will be available from summer 2021. As I say, again, there’s this issue about what we’ve learned from the first time you’ve matched the SPI to the outturns in Scotland, which we may want to come on to separately. There is some other HMRC information that can be useful. For example, I think the last time I was before you, I noted that HMRC was doing some work looking at whether tax motivated incorporation would be showing a different pattern from Wales than from other parts of the country. That work’s been done, and it doesn’t look as though there’s a particularly different pattern there, but that—again, you can take into account some information from corporation tax to draw upon and some tax avoidance data as well.

So, I think the flagging is the thing to look forward to as the change, but, as I say, we've discovered with Scotland that you get some surprises from that, as well as new information you can draw upon. One was struck, I think, when Mr Harra was before you—. I think one thing they learnt from the Scottish experience is that they've actually been more systematic in writing to everybody, explaining how the system works and the flagging system, rather than a relatively small sample, which might limit one potential source of surprise out of the flagging data when it comes in, relative to the survey data.

The Welsh Revenue Authority will be focusing solely on what happens in Wales and will be looking for ways of fine-tuning its assessment of what's happening in Wales. As the body contracted to do the forecasting for Wales, and to assess from an OBR perspective what's happening in Wales, will you be looking for specific new data for Wales or only getting the Welsh version of whatever you're looking at for the rest of the UK, if you understand what I mean?    

Well, I don't think there's any—. There's no new science out there that we can crack that's Wales-specific on that, but that's not to understate the importance of having this flagging data coming in, and that telling us whether the survey data upon which we have been resting—everybody has been resting—so far has been giving an accurate picture or not. 

Sorry to interrupt but, for example, when you said that a higher proportion of higher earners are measured in data collection, maybe the models should be different in Wales and you should be over-sampling in other parts of the Welsh economy—are you going to continuously be looking for the best way of getting data for Wales specifically? 

Well, the key advantage of flagging is that you no longer have to rely on a sample, so you're not concluding from a relatively small sample—I say 'relatively small', but it is a large sample—and you will have that overarching pattern. As I say, the Scottish experience of telling you whether there's a systematic difference, for example, in what people say their place of residence for most of the year is versus the postcodes that happen to be captured in the survey is the same, or whether there is a behavioural response—whether people have actively decided, because they have homes in two places, that they want to declare it in one place rather than the other.

The next stage we will have next year is that the next phase of the survey of personal incomes—. The people who reply to that—you still have their postcodes, which is the basis on which you've identified the Scottish share to date, but you'll also have how they flagged themselves. So, we'll be able to see, clearly, at least one potential source of the difference, which is that maybe a whole x proportion of the population say, 'My postcode is in Scotland, but I'm not a Scottish taxpayer', or vice versa. Again, I think, given residential distribution across the border, that could be an issue here, presumably, as it is there.      

Was that the main factor in the overestimation of the SRIT—the Scottish rate of income tax—estimates for 2016-17, and have we learned enough from what happened there? Because it was a bit worrying for Scotland. Have we learned enough so that we don't end up in the same situation here in Wales?

Well, we haven't learnt enough yet. We will learn considerably more when we have the first survey of personal income that includes the flagging data as well. So, the possibilities are that there are simply sampling problems in the SPI: the SPI is, for some reason or another, not giving you an accurate representation of the underlying taxpayer population. So, as far as you can tell in the outturn data, it suggested that there were fewer overall taxpayers in Scotland than you would have suspected from the survey, and that a smaller proportion of those were on relatively high incomes who would be paying relatively high amounts of tax, hence the overall receipts. So, there's an overall number change and a compositional change.

As I say, one possibility is that the postcodes that people have on the SPI just do not reflect where they live most of the time in the year, and therefore it's not that people have chosen to game the system, but it's just that the SPI does not record accurately, because it was never designed in that sense to record accurately whether a person is a Scottish taxpayer or not.

Another possibility is that, again, so far we're relying on the survey of personal incomes for 2015-16. So, you're comparing the outturn data for 2016-17 versus what the survey would have told you for 2015-16, which you've had to, sort of, mini-forecast up to 2016-17, and maybe that process of getting from your 2015-16 survey to what you think it would look like in 2016-17 was awry.

Another possibility is just that HMRC had administrative problems implementing this, and therefore maybe the flagging eventually will look more like the survey, but, to begin with, the flagging hasn't captured the right people. So, it's not necessarily the survey that's 'wrong'; it could be that it's early days. As with any big administrative change of this sort it takes time to bed in.

And then there's the possibility that you've had a behavioural response, which would be either people could have genuinely migrated—they will have decided that they no longer wish to spend the majority of the year in Scotland and will want to spend it somewhere else in the UK—or that people are basically in a position to choose between the two and that they've chosen not to be Scottish taxpayers for the purpose of this. Now, clearly, at that level, you might think that this will be more of a problem in Scotland than it would be in Wales, because this change has happened in Scotland at a time in which there's obviously been very open talk about the possibility of higher tax rates on relatively high-earning people. So, you're giving them something for their behaviour to respond to, which has clearly not been the same in Wales. The conversation hasn't been the same way, so there's been less in the political debate, perhaps, to prompt people to behave in that way than there has been in Scotland. So, I think you're first guess would be that you would assume that there would be less of an issue for Wales than there would be for Scotland, but I think we need to understand more about the Scottish position before you can be too clear about it. 

And it might change. 

And it may change. 

And very briefly—I think you've answered this—on the difference in the Welsh Government's forecast for 2019-20 from the OBR's—there's £40 million difference there, with the Welsh Government's forecast being £40 million higher. Is that entirely attributable to the tax rate changes that you alluded to earlier or are there other factors at play there too? 

For 2019-20? 

Yes. 

Very largely. I don't know whether you have the document—at table 2.8, the personal allowance and higher rate threshold on its own is making £44 million of difference in that year. It's less in later years, and it is, as I've said, partly offset by the off-payroll working, and obviously that wasn't there for the Welsh Government to incorporate, so that's basically where it's come from. 

Thanks. 

Thank you. Jane. 

Obviously, it's early days in terms of forecasting our devolved taxes and challenges, as has already been identified, but how do you think actual land transaction tax data from the WRA will improve your forecasting? And how long will it take, do you think, to get this to be a fit-for-purpose situation in terms of data collection? 

On the transaction tax I think, again, I return to the point about it's always going to be much more difficult on the commercial side than it is on the residential side, because of the potential large changes that Mr Hedges was talking about. I think on the residential side, it's obviously—. The fact that we are getting months of data out of the WRA is great. The question here is how long it takes you to identify whether there is a systematic pattern to the receipts through the year. So, as I say, at the moment, if you were to draw a straight line between nil at the beginning of the year and where you're forecast to be at the end, so far the receipts are coming in a little bit weaker than that line would suggest. But that's kind of what you would expect given the potential for forestalling as the system has come in. So, you're therefore likely to need some years before you're clear as to whether there's a systematic pattern. Now, of course, if, at any point in that, you have a change in the tax regime, then you're not quite back to square one again, but you're then having to disentangle not only the steady state impact of any change in terms of when it's implemented and how it feeds through to that pattern, but also whether it results in any forestalling behaviour. So, if you announce you're going to do something and that results in transactions either being brought forward or pushed back, you then have the separate challenge of saying, 'Well, how much has that distorted whatever the underlying pattern would be?' So, summarising, it's great that you have that specific data, that's it's coming in and we learn what we can from it, but it will take time to distil from it the regular patterns from anything else. 

Because I think at the outset you talked about the fact that you have to actually zoom in to great detail in Wales. It's a much bigger forecasting responsibility that you have. But that zooming in is about Welsh circumstances as well as the wider changes that could happen that would affect Wales.

So, you've mentioned that your forecasts for LTT are produced on your behalf by Welsh Government analysts. How do you see this changing once you're responsible for providing independent forecasts to ensure that we get this independence in terms of the process?

Well, at the end of the day, this is an analogue to the relationship that we have with HMRC at the moment at the UK-wide level. There is a model of fiscal watchdogs that have all the forecasting capacity in-house. If you do that, we would be like the Congressional Budget Office in the US, or we'd have a staff of 200 to 300, which is not a delight that I would want to have.

So, the model that we have used, and I think it has worked well at the UK level, is drawing upon the day-to-day administration expertise that comes from the people who are operating the system on the ground. As I say, I'm perfectly happy in terms of the consequences of independence of that sort of model, where, basically, we think together about the machine that we need to do the job and we say, 'I think you ought to assume this about house price growth, this about transactions', and somebody can then turn the handle on that.

Interestingly, the precise—. All of these transaction tax approaches are: you basically come up with a picture or a shape of the distribution of transactions that takes place in any given year, and then you take that picture of a relatively large amount of them at the bottom and a smaller amount of them at the top, and then you have to project that forward through time. And there are different ways of doing that, and at the moment we're actually doing it in different ways in Wales, Scotland and the rest of the UK, in the sense of whether you have a very detailed—what's called a micro-simulation approach, where, basically, you derive a picture from all the actual transactions and move that forward through future history. Or you take some information about what the average size of the transaction is and what the most likely size of the transaction is, and you draw a shape around that in a more stylised way, and you move that forward through history. Or you take the approach we have in Wales, where we've adopted the approach that the Welsh Government was using, about not looking at every transaction but putting them into what are called 'price bins'. So, you take groups and then move that forward through history.

So, we are road-testing a number of approaches here across the different sets of forecasts, and that may be an useful source of learning over time, and it depends a bit on—. At the moment, these seem to be the appropriate approaches in each of the places that we're doing this, but the fact that you have those things to compare will, over time, I think, be informative for itself. But I wouldn't understate, as I say—. Judging from five or 10 years of performance whether price bins work better than a log-normal distribution or works better than a full-blown micro-simulation is a tough ask when, frankly, the biggest problem is that you don't know whether transactions in total are going to rise or fall by 10 per cent next year. So, distinguishing between bad luck and bad judgment is never as easy as you would like it to be.

Thank you.

And that's probably what it's all about, really, isn't it? Okay, and the final set of questions are from Neil Hamilton.

Following on immediately from that, because you've mentioned the figure of 10 per cent, looking at your forecasts for land transaction tax in March 2018 compared with October, you're forecasting for the next four or five years a 10 per cent reduction in the amount that land transaction tax is likely to raise—about £150 million over those years. And, looking at a deeper explanation of this, it's because you've forecast that non-residential properties are likely to raise less revenue. Can you explain why you think that's likely to happen—more likely to happen than it was a few months ago?

Yes. I think this is very much one of those cases when you're drawing on the information that you've had from the recent outturn data rather than an overarching change in your judgment about the performance here. So, if—. I don't know whether you've actually got the document or not, but—

Well, we've got an extract from it.

If you look at the path—. If you do, it's chart 3.7. If you look at the path of the commercial LTT receipts over the first five months of the year, they're clearly some way south of the straight line that you would have drawn based on the March numbers. And I think what's happened there is that 2017-18 was a significantly stronger year in outturn from the previous two, and that back in March we were assuming that that would stick, and that therefore you would push that through into future years. The data for the first five months of this year suggest that, if anything, 2017-18 looks like an outlier and 2018-19 is therefore likely to be more similar to 2016-17 and 2015-16, and we're working on that presumption at the moment. If you look at the way in which the receipts are coming in over the first five months, it's relatively consistent. As I say, it's slightly below the latest October forecast we put in, because we're assuming that the receipts will be somewhat more depressed in the early part of the year, forestalling. So, I think the key issue there is that the data is suggesting that 2017-18 was an oddity, rather than a new normal, and that's the main reason for the change.

Right. Well, that's very interesting. The other interesting feature of this is that your forecasts differ significantly from the ones produced by the Welsh Government. You forecast that residential LTT revenue will be £63 million higher than the Welsh Government forecast, and non-residential £53 million less.

Yes, and oddly it's washing out overall. No, you're exactly right. I think—. Well, obviously, one reason is that the Welsh Government was having to draw on our March macro-economic forecast, and the macro-economic forecast has moved on there in terms of the expectation of slightly stronger earnings and employment growth than we had then. I think, on the residential—sorry, on the commercial—side, you've got slightly different judgments upon what the profile through the year would be. So, I think, again, particularly on the commercial side, how much weight you place on this judgment about 2017-18—you know, is 2017-18 abnormal or not—is the key to the difference there. But, as you say, if you add the two of them together, the impacts broadly wash out. But that's coincidence rather than a—constraints of the model in designs to achieve that outcome; let's put it that way.

And, lastly, then, again your forecasts for LDT revenue are significantly higher than they were, by a very substantial margin in fact. If you aggregate all the five years from 2018-19 to 2022-23, in March you were forecasting a total revenue of £109 million; in October, that's gone up to £186 million. So, it's an increase of £77 million on £109 million, so it's virtually double the original forecast.

And that is—. We're now in the wonderful position of having one whole data point for the second quarter of 2018, which is better than the none that we had beforehand, and we and the Welsh Government, I think, both reached the same view on that, which is that it is the best we've got to go on, so we place a reasonable weight on that, but strongly signal the uncertainty around this, because we don't know whether this will stick or not, rather than to say, 'We don't believe this in the light of the forecast that we had done previously', because of the lack of data upon which we had to base the previous forecast. So, to that extent, it is something of a gamble that this one data point is full of news and not noise, but I think it's—. Given it's what we've got, we both reach the view, I think reasonably, that you put your weight on that, but at the same time you say, 'Look, we just don't know yet how this is coming out.'

Now, as I say, one reason you would be anxious is not merely that you've lifted the level up, but that it implies a big increase from last year to this year, and, in forecasting terms, you're always slightly suspicious about something that shows a big jump from one year to the next. And that is one reason why we're highlighting the uncertainty, but I think the challenge here is that we don't know whether the history is correct, as well as the forecast. So, our suspicion is that, actually, the past years have been higher, and that, therefore, if this new number from Q2 2018 sticks, what it's telling us is that it was understated in the previous years, rather than there's been some astonishing jump in the amount of disposal that's been taking place in 2018-19. But only time will tell whether 2018-19 sticks or not, and, of course, at the end of the day, we'll never know what the true picture was in the past, but, at this stage, we haven't completely rewritten history on that basis because that doesn't seem appropriate at this stage. But I think that's—. If the new, higher number sticks, that, I think, is something you'd want to think about—as to whether, actually, this number should have been higher all the time in the past. 

So, for quite a long time yet, we're likely to be flying blind, in effect—

Yes, and I think, with this one, you know—

—and the forecasts are hardly worth the paper they're printed on, if I can be blunt. 

As with any forecast, you base it on the data that you have, and, you know—. I would say, if a forecaster is complaining about data, it's like sailors complaining about the sea: it's what we have to sail. [Laughter.]

Exactly. 

Well, thank you, and on that note, then, we'll draw that session to a close. Thank you very much. You've given us a very rich mine of evidence there that we can draw upon in our deliberations on the draft budget. Thank you for your attendance. You will be sent a transcript, a copy of the transcript, to check for accuracy. So, thank you very much. 

Thank you very much, indeed. A pleasure to be back. 

And, if we may, Members, I think we'll move now into private session and take item 9 to reflect on that evidence before we ask the Cabinet Secretary to join us for the next item. So, we'll move into private session. Thank you. 

Welcome back to the Finance Committee. We move to our seventh item on the agenda this morning, which is to receive further evidence on the Autism (Wales) Bill, and we welcome Vaughan Gething, the Cabinet Secretary for Health and Social Services, who's joining us this morning. I'd ask you to kindly introduce your officials—or they can introduce themselves. 

Morning. I'm Matthew Jenkins, I'm a deputy director in the social services and integration directorate. 

Gareth Haven, from health and social services finance.

There we are. Welcome to the three of you. We'll go straight into questions, if that's okay. 

Of course. 

And I'll kick off, if I may, by asking you to explain why you didn't provide costs for the current provision of autism spectrum disorder services to help inform the regulatory impact assessment of the autism Bill. 

The costs, such as they are—that we're investing in the service—are available. In terms of the costs for the Bill, that's not a matter for the Government. There's been—. Through the Bill process, there's been an interesting conversation about how we divide up costs and how we provide those costs. Because we're currently delivering a system based on needs, and we're investing money in the integrated autism service to underpin the strategy we have, to then undertake an exercise to try and portion out direct costs for autism would have required us to run through a not uncomplicated process, and, really, that's a matter for the Member in charge, as opposed to the Government undertaking work with a real cost to it. So, it's for the Member in charge to set out from his own perspective how he assess the costs as they are and the costs of the Bill and the legislation that he seeks to introduce. 

But your answer suggests, therefore, that you're not clear yourself how much the Government is spending on autistic spectrum disorder services, because, if it's such an onerous exercise, the suggestion is that you haven't done that. Because, if you had, then you'd be able to provide those figures.

We can provide figures on what we're investing in improving services, but part of the challenge is if you want to take a—. This is a challenge about a diagnosis-led approach or a needs-led approach, and, if you want to diagnose, and that's the point, and that's what the Bill seeks to do, then you've got to spend time, effort and money on doing that. And, if you're looking at the needs, then the needs are in more than one area, and so—. And that's about seeing that whole person for all their needs. Do you really want to go and undertake an exercise that says, 'Well, how much of different areas of activity are undertaken on support?' I don't think that in itself is necessarily very helpful. We're investing money around a strategy to try and improve outcomes and the experience and so on. Those costs are, of course, completely transparent and available. And, if you want to go into the sort of technical exercise you need to go through to get to that level of detail, then Gareth can provide you with information on how that would need to be done and what that would mean practically. 

But do you understand the frustration, therefore, from the Member in charge, that he felt that, with those figures not forthcoming, it was making it very difficult to fulfil his ambition in terms of the Bill?

To be fair, when I've been in roughly this position, but in this position as a backbencher, working together with Mick Antoniw on the Recovery of Medical Costs for Asbestos Diseases (Wales) Bill, we didn't have the Government saying, 'Here is the information we have.' We looked at an area that we wanted to change, we looked at what we thought the costs were, and we then engaged a health economist to come up with a robust exercise in understanding the level of costs, how the Bill could work, and the mechanisms. And that's what you have to do as a Member in charge of the Bill. I don't think you can say, 'Well, I don't know what the costs are and the Government should tell me, and I'm not prepared to do any work around that.' I'm genuinely surprised that the Member in charge hasn't looked to engage a health economist to have a better understanding of costs now and the cost of running the Bill as he proposes it. But that is the job of the Member in charge, and that really isn't the job of the Government. I'm trying to be as polite as possible—

Yes, sure; I hear what you're saying. 

—but also as clear as possible about the role and responsibility that the Government has and whichever Member of whichever party in being able to present a proposal to the Assembly that isn't just a good idea, but then practically what that means and how to get answers and the clarity that you would expect from any Bill, whether it's a Government Bill, and the standards you would expect and set, or indeed an individual Member Bill.

So, in the absence of those figures being provided for the Member in charge, the Member has clearly included in the regulatory impact assessment calculations for costs he believes are valid for the current delivery of services. How fair a reflection of the actual costs are those, in your view?

Matt, do you want to go through the current envelope, because I think it'd be helpful to look at that, and then to get into what the Bill is actually proposing, because what the Bill proposes is essentially diagnosis to get to services, and lots of the costs that are run through are about the costs introduced by the Bill, and that's essentially a waiting-list initiative? But I'm sure we'll get into that in some more detail. 

Indeed. Thank you. So, there are two options in the RIA. Clearly, option 1 is the status quo option, as if the Bill doesn't exist. We're reasonably happy with the methodology that the Member has taken. If we were to approach that as a civil service, it's roughly the approach we would have taken. There are a range of detailed comments I could make about the prices and years upon which some of the estimates are based, and whether they're completely consistent throughout the regulatory impact assessment. They would be points of detail rather than fundamental flaws in the way that he's set out the first option. Option 2 is clearly about the Bill itself. 

Okay, fine. David. 

Well, let's have the detail, then. We are here to look at the financial aspects of the Bill, not the policy aspects, and if there are concerns—you used the word 'flaws'—what are they? 

Do you or Gareth want to answer about the financial aspects where there is disagreement? 

I can pick out some examples. So, on page 57, the RIA summary indicates that the price base is 2017. Table 5 suggests the costs are in 2013-14 prices. Have the costs in table 15 been uprated to the 2017 price levels or have they not? It's not particularly clear from the narrative. That's one example of a detailed point.

More broadly, looking into option 2, the methodology that's been taken is to look at the current population of the waiting list, and to gross that up using a figure of £2,500, which would be the private sector rate for an assessment process for diagnosis. The Member has costed up the waiting list. The RIA is silent on what would happen beyond that waiting list being eradicated. So, it says nothing about future demand, nor does it say anything about the increased demand that may well come through the system through an approach such as this, which is diagnosis-based. Our experience here in Wales, through the first autism strategy, is that increased awareness results in increased pressure for diagnosis. That in and of itself is not a bad thing—that's something that we've sought to build—but we have evidence here in Wales that that increase in demand is likely to have come through the system. 

Similarly, based on discussions at civil service level with colleagues in Northern Ireland and their approach to having legislation in this area, a significant demand has come through the system for diagnosis as a result. There are different ways of rationalising that. Equally, the conversion rate from an individual being put through to assessment through having a diagnosis confirmed that that has increased as well over the last three years. So, we're seeing that significant drive coming through the system based on experience in Northern Ireland. The RIA is silent on that; it assumes the current level in the system would remain static. So, the costs are a relatively small part of what is quite a broad picture.   

Let me clarify, then. You have concerns over the consistency of the data in option 1, and the different years being used in the data. On option 2, you talk about the fact that it's actually just focusing on the diagnosis and not necessarily what's beyond that. But, surely, this Bill is about, as the Cabinet Secretary has highlighted, diagnosis. Irrespective of whether an individual is diagnosed, if he needs those services, he needs those services. So, the service cost will be there anyway, whether this Bill is produced or not, because the autism strategy should identify the services for people who need those services.   

It isn't quite as simple as that. If you have a diagnosis-led approach, then you'll have pressure on diagnostic services. Don't take my word for of it—it's something that the Children's Commissioner for Wales refers to in her evidence as well. So, you'll drive pressure there, so you'll drive resources there as well. That either means that you're saying that a diagnosis-led approach, not a needs-led approach, is the right thing. That's a significant departure from the way we provide services now across health and social care, and certainly the scheme set out in the social services and well-being Act, but it also then means that you're taking resource away from delivering against those needs. The RIA sets that out itself, actually; it recognises that that's a risk. And if you're doing that, then, actually, if you spend your money on a waiting list initiative and you haven't then quantified how much you want to spend on delivering the service, or the support that's needed after that, then you can't understand the full cost of the Bill. Not only that, though, but you have this challenge of you either say, 'We're prepared to pass this Bill and work out the costs later', or we pass this Bill and accept that we'll just have less resource to go into delivering against people's real needs. And that's difficult, because you're looking at needs in more than one area; it isn't just health spend, of course. The Assembly passed a needs-based piece of legislation on additional learning needs recently. Now, that is about looking at the person's need and making sure that that is met, and there are times you need a diagnosis to be able to understand that need, but the challenge if you just have a diagnosis-led approach is that you force people through a gate and you'll drive more demand that way. And that gets you away from a needs-led approach.

So, this would be a significant departure from the recent Additional Learning Needs and Education Tribunal (Wales) Act 2018 and the Social Services and Well-being (Wales) Act 2014 and it would drive money in different parts of the system. And, of course, the Bill doesn't identify how or where additional resource would be used; it doesn't actually identify what additional resource would be required. And that's part of what I'm genuinely a bit surprised about in the proposal that you've got in front of you—that it doesn't try to understand that, because there is evidence in Northern Ireland about increasing the level of demand that comes into diagnostic services and what that means. If this was a Government Bill, I think we'd have understandably searching questions if that was the approach we'd taken without then explaining what we do later on.

Obviously, you've raised those concerns. I just want, therefore, to know: have you estimated the costs that this Bill might bring in, as a consequence? Because if you're talking about resources being diverted, being redirected, which leaves a gap elsewhere, have you worked out what that cost would've been if you had to deliver this Bill—what that cost would be to you?

No, this is the Member's Bill, it's not the Government's Bill, and it's for the Member to set that out. If I direct officials in this team to do that, they're not doing the job that I want them to do, which is to do with improvement in these services. So, there is a real cost to do it, it's not just an opportunity; there is a real cost to doing that. If the Assembly decides to pass the Bill, then we'll have to, of course, understand what the Assembly requires us to do and that's beyond Stage 1 and into Stage 2, Stage 3 and then passing the Bill. But at this point, I think it is the wrong use of limited Government resources, and certainly resources within the health service, to do that. I think it might be helpful if Gareth explains the sort of exercise we think you need to undergo to understand the cost of that, because it isn't a simple exercise of putting data into a formula and coming up with a cost. 

I accept that, and I will listen to that, but one of the things I'm thinking about is, if you're saying to us that it's not your job, which I understand—it's not your Bill—and that you're then saying that there are costs that are not considered, I would want to know what those costs are, so that we could work out, actually, 'Oh, this Bill is actually going to cost a lot more than was anticipated' and this could actually have been worked out because of these figures. What you're saying at the moment to us is, 'We think these costs are not appropriate, not relevant. We think that there's a need for far more deep thinking behind those costs, but we haven't got the figures.'

Perhaps Gareth can run through that, and, I'll happily send a written note for you before—. When do you finish Stage 1 consideration on this committee?

We report by 7 December, so before then—

Well, before then, we'll make sure—and not just a day before; at least a week or more before—that you get a note setting out some of what we think are challenges in the way the RIA is constructed. So, you get something written, again, from us setting out some of the detail that we think is missing for you to consider in—two weeks, three weeks?

Yes.

We'll make sure that you get a written note well in advance of the session, because, of course, you've got to lay that, you've got to go through that to understand—

Indeed.

—so, we'll send you a written note setting out some more of the detail. But I do think that an understanding of the sort of process that's needed to go through would be helpful as well.

Yes, thank you.

I'm genuinely trying to be helpful, I'm not disagreeing—

I appreciate that, thank you.

I think as the Cabinet Secretary has said, the information doesn't really exist for us to get easily. It'd be quite a substantial piece of work to pull that together. The NHS programme budgeting information simply doesn't go low enough in its granularity down to the specific condition of autism and separate it from other neurodevelopmental conditions, really. So, as I've said, I think they've worked with the best evidence that's available at a UK level and approximated it for Wales, but we just don't have that information readily available to us without committing significant time and resources to try and pull some of it together.

Do you have figures for neurological conditions? Do you have that, as a totality?

Well, it's—. I did bring—. There are—. We can provide it in the written document; we can just explain what is available. It sort of breaks it down under different mental health conditions—child and adolescent mental health, learning disabilities—but it kind of stops there. So—

But it doesn't go down to neurodevelopment.

No, it doesn't go further down. 

We didn't get an evidence paper from you for this session, did we, and on reflection, maybe some of that information would have been useful before today, as opposed to retrospectively? But there we are. We'll appreciate receiving that as soon as possible. 

I think as well, if you wanted to run through a project with a health board, you'd require the information, you'd be required to understand the data that they have, how far they can go, and that is a significant piece of work that not only the Government would have to do, but the service would have to do as well. So, you'd be taking people in the service to try and understand those costs. Now, I really don't think that when we're trying to improve real services that that's what we should do. If the Member wants to introduce the Bill, then it's the Member's responsibility to set out how and why they want to do it and the real costs of doing it, rather than saying, 'I have an idea, now the Government needs to go and cost it for me.' That just isn't the way it works. 

Can I come back on that point?

Yes, okay, very briefly, then, please. 

Very briefly. I understand that totally, and at the end of Stage 1, when it comes to a debate in Plenary, we'll have a debate on the general principles, followed by a debate on the financial aspects of it. Now, if we haven't got an understanding of the adverse—I'll use the word 'adverse'—financial aspects of it, that debate's going to be made very difficult for us sometimes. 

We'll provide you with a note setting out some of the—not just the logic, but the reality that if you drive cost at diagnosis, you're taking cost away from support, and what that really practically means. So, I'll happily set out a note for that, but, like I said, the person who promotes the legislation that they want to persuade the Assembly to pass needs to set this out. 

But we need to also have the whole argument, the whole set of evidence. 

Okay. Nick, yes. 

There's something about this that's all coming across as a little bit strange. Your officials have said earlier that you don't rule out that there are some merits in having a diagnostic-led approach. 

Okay. With apologies if it sounded like that—I was simply saying that the process in the RIA that had costed option 1 looked methodologically sound to me. I wasn't commenting on the option as to whether that was a good idea or not. I was commenting on the approach that had been taken to developing the RIA. 

Okay. Well, then, later on, the health Secretary said that you wouldn't—okay, I'm paraphrasing you now, so you'll probably come back and say I've misunderstood what you said, but you said that, at this point in time, you didn't see that as a good way of proceeding, because of the cost issue, primarily, I imagine. 

No, there's a straightforward principal disagreement about whether you should have a diagnosis-led approach to legislation, and to the way in which you direct resources around the service in the way the Bill tries to set out, or a needs-led approach. And that would be a significant departure. The social services and well-being Act—needs-led. The additional learning needs and education tribunal Act—needs-led. This would be diagnosis-led. You know, you get into the difficulty of other groups are understandably watching to see, if this is the approach that's taken, whether they want their own specific needs-led or specific condition-led legislation. Now, that's a real issue, a real factor for what happens in future legislation. But for this Bill, whether it's the right way to try and direct the service. And, again, this is not just what I am saying. It's not the Government view. There are a range of people. The Royal College of General Practitioners don't think this is a great idea. The children's commissioner has concerns about it as well. So, it isn't just that the Government are trying to say, 'Let's stop this because we don't like it.' It's a real practical challenge about what would be the impact of the legislation if you have a specific diagnosis-led piece of legislation to direct resources around the service, or if you try and actually address people's needs.

Now, like I say, we've done for colleague committees a note trying to set out all those other—and Rhun would have been on the health committee at the time, I think, or just come off it, I think—groups that agree with the approach that we're taking and actually have concerns about how you would direct resource around the system and whether it's the right thing to do. So, there's a principal difference between that sort of movement, but I think we'd all agree that we want to see improvement in not just the delivery of services but real lived experience. 

I'm reminded of Al Gore's book An Inconvenient Truth, which deals with some issues like this. Because you're not necessarily saying that a diagnostic approach wouldn't benefit the people out there who have autism, or are caring for people with spectrum disorders, but you're saying it doesn't fit in to the wider way of approaching health issues at the moment. So, that doesn't necessarily rule out the benefits of a diagnostic approach.

Well, diagnosing a condition, in some cases, is a part of understanding and then properly meeting someone's need. Having legislation that says that you can only access services if you get through the gate, well I say that's part of the challenge then, because you're directing people to a diagnostic service, and you're saying that's the ticket to get something else, as opposed to how you support people to meet their needs in any event. There are people with undiagnosed conditions who have needs that are still met, and support needs that are still met. So, it is quite a different approach to take. So, it is far from being as simple as saying, 'Diagnosis is just a good thing in itself; that's all the Bill is trying to do, so let's get on with it.' I just don't think that is a full view or an honest view on what this piece of legislation would do in practice. And we're getting away a bit from the money, I know, but—.

Yes, okay.

Can I—

Can David just come in on one point? Very briefly, if you may, David.

It's a very brief one, but I wanted to make sure that we focus on the financial aspects, rather than the policy aspects, which we are starting to talk about a bit. If you have a briefing paper that discusses the difference between the diagnostic approach and a needs approach, have you looked at the implications of costing the financial aspects, as a consequence of that? In other words, is it more costly to do a diagnostic approach than it is to do a needs approach, so that we understand the implications of finance and costs as a consequence of that?

There's a principal point, and then, in the note that we'll send you, we'll set out what we think some of the cost challenges are, and understanding what those cost challenges are—

That would be helpful.

—as well as being able to set out for you the wider concerns that exist, not just from within the Government, about directing resources at what is a fairly large waiting-list initiative, without then understanding the ongoing cost for future diagnosis and support needs as well.

That would be helpful.

I think we can share the papers at the health committee as well, which I think will help us on that, and we'll do the stuff about the RIA as well.

Thank you.

Can I just ask finally, before other people come in? You mentioned earlier—well, you've said a few times now—about your concerns about costs further down, treatment further down the line suffering because of the costs being borne earlier on. But doesn't the RIA provide additional costs to ensure resources aren't diverted?

No, because in terms of what the RIA is saying, it isn't identifying, 'Here is how money can and will be found', it's directing the service to go at diagnosis—it's directing the service to have one particular track. That will inevitably move resources over, and that's very clear. What it doesn't then do is identify (a) what you then expect to do in terms of meeting people's needs, and the cost of that as well. And, again, that is for the Member in charge to do.

And just very finally, because in terms of that—the point David Rees made—we'll all have to consider the general principles of this at a point in Plenary. And I am struggling at the moment to think, okay, I understand what you're saying about the Member in charge has to provide certain information, but surely there must be an onus on the Welsh Government, if you're going to refute that information, to have your own information. You can't simply put the onus—if we're going to have to decide whether this is a good thing or not, general principle wise, then you can't just put the entire onus on the Member, surely.

Well, we've been very clear in the general principle of scrutiny, in front of the Health, Social Care and Sport Committee, about the difference of approach, and other people who agree with our broad view on whether this is the right thing to do or not. We're happy to provide you with more information on finances, but the finance is only part of the argument. And, actually, when the Member brought forward the proposal, it was about improving services and outcomes, and on that we agree. The Member agrees on lots of the things we are trying to do, and on the aim and objective of the services that we are rolling out to make a real difference to people's lived experience. But I don't think that shared agreement about what we want to do in reality is going to be advanced by the Bill that we actually have in front of us. And there's a difference between what the Bill says and what the perception is about what the Bill would actually do.

Your current code is costing money though, isn't it? So you must have costs for that.

Yes, and that's being delivered within the current resource we have to actually improve autistic services. So, we've budgeted for that, we know what that is, and we think that will make a difference. It will be a code issued under the Social Services and Well-being (Wales) Act 2014 and under the National Health Service (Wales) Act 2006. So it will bind, effectively, health and social care together about what our expectations are. So, we're doing that within a resource base we understand; what we don't get is the proposals that the Member makes. And, like I say, that is a matter for the Member in charge.

Yes, okay. Jane.

I suppose we've moved on to the question I was going to ask, and we've got to focus on the financial aspects, of course, but you have updated the autistic spectrum disorder strategy delivery plan. You say, of course, that the code is within the financial envelope, but has the updating resulted in additional costs being incurred, and what impact would that have, or could it have, on the cost presented in the Bill's RIA?

The Bill's RIA doesn't take account of the code that we're developing. To be fair, in the conversations we've had with the Member in charge, we did offer the opportunity to get involved. He chose—and to be fair, we all have limited time and personal resource ourselves—to give his time to developing the Bill and said he didn't have the time and the capacity to be involved in being more involved in the development of the code. That's an honest choice and I don't criticise him for making that choice, but if he'd been more involved in that, he could have seen some of the costs and how the benefits would have worked.

We've budgeted for the cost of the code. If you want, we can provide you with an approximation of what the development costs are, but it's about how we change a service, and that's the point of delivering a code—to try and underpin some more of the certainty about expectations for people, because lots of the demand around the legislation is because people recognise that real, lived, practical experiences are often very difficult for families and it's about how we make a real difference with and for them. It's across different sectors as well. So, a lot of the demand and the challenge comes around education and then, of course, there are challenges around adults and the support services that they have as well. So, we are trying to take a holistic approach, and again, based on people's needs.

I think it might be useful to have a note—

On the approximation of the cost.

—on the cost in terms of the development of the code.

The development costs of the code—that won't be difficult.

I think that would be helpful.

Thank you. Rhun.

Yes, that's what I wanted to ask. We would appreciate an approximation or a clarification of the costs of the code, whether that's now or in written form, and the implementation cost. Also, this is something that you can answer now certainly: how will the code of practice deliver improvements in support services and outcomes for people with ASD? Because that is the 'meeting people's needs' and there is a price tag to it, and you must have an idea what that price tag is.

This is about how the code will reinforce existing duties and underpin the delivery of the strategy. It's about how we make best use of the £13 million we're investing to try and actually deliver that real lived experience and to improve that. On a positive note, when I was in north Wales last week, I met people who have taken part and who have helped them to design the roll-out of the integrated autism service in north Wales. Autistic people had a direct say in how the service is working and what it's looking like and they were positive about their involvement and about the difference it's making. There was a comment: 'We've been waiting for this and it's a step forward.' That's really encouraging, but also we've got to have the honesty about whether it's still making the difference.

So, we're going to see the roll-out of the rest of that service in this calendar year through the rest of Wales. We're committed to an evaluation, so the money that we are committing to try and deliver that improvement is there, is identified. We'll have an evaluation on the first areas that went and the roll-out of the integrated service in the early new year—that's being undertaken then—and so there'll be honesty in the improvement, or otherwise, that's been made and that will help those areas that have the roll-out later on. So, we are genuine about wanting improvement and also about learning from what's happening as well.

There will be implications to other budgets, though, of the rolling out of a new code. Will the figures that you can provide for us, as a committee, include the approximation of how much the implementation of a code will cost to other public service delivery bodies in Wales?

No, we're talking about how we use the £13 million and how people meet their current and existing responsibilities, so—

So, £13 million is the price of implementing the code then.

That £13 million is the money we're investing in improving services for autistic people. That's the money that underpins the delivery of the autism strategy. That's new money to go to accompany the strategy as opposed to rebalancing and moving money around from somewhere else.

Can I ask one question that arises from, I think, an e-mail that was sent to all of us as members of the committee yesterday by an anonymous individual? Referring to the RIA option 1, saying that it's estimated to cost £5.7 billion over five years to address the needs of children and adults with ASD in Wales, and this individual found that laughable, that £1 billion, which is a significant proportion of the entire Welsh budget, is being spent on ASD services of one kind or another.

Going back through the document, I think that’s a misinterpretation. I think what option 1 does is to describe the costs of the condition, in effect, in terms of the limitation that that means for individuals—so, the cost to the economy and the cost to them personally. So, they’ve used a UK study, in which the costs have been apportioned across the nation, in a sense. So, it is that, rather than—

So, it's not cost that's been met, it's—

It's not what’s being invested in public services, clearly.

Thank you for that clarity. Neil.

I think my questions have been answered, really, but just to be clear, in your written evidence to the health committee, you said in respect of the RIA that there are elements that require further explanation in order to test the robustness of the calculations and areas where additional evidence to support the analysis would be useful. I presume that the note you're going to provide us will fill the gaps there.

Yes, we can happily spell that out to you.

Thank you. 

Okay. Thank you. Finally, Mike Hedges.

You highlighted a number of provisions of the Bill basically urging the Welsh Government to do things that are already being done. Can you net those off for us?

What, you mean in the note we have about—

Yes, in the note you'll provide, can you net them off for me? If you’re saying they’re already being done, you can tell us how much, what we're spending to do them now, and then that can come off the cost of the Bill.

I’d expect to be able to be clear about what we’re already doing and what are additional costs are and what are unexplained costs from the RIA.

But, you can give us the numbers we can net off from what you're currently doing.

Yes.

Which would be, I think, helpful to us.

We've had long discussions, and I'm not sure how far we've actually got on this. But leaving aside whether it's a good idea or a bad idea to have this Bill, on which we as a committee have no view whatsoever, as a committee—we might as individuals, but we have no view as a committee—our view is solely about how much it will cost, and then to report that to a Plenary session where a decision can be made about whether the cost-benefit analysis done by each person in their own mind is a direction in which they wish to go forward. I'm not sure how we're able to do that with the information being provided by both you and the Member in charge. Do you think we can, with the information we've had from both of you, actually come up with our own sort of group cost-benefit costings to actually be able to clarify what it's going to cost? 

But we'll provide clarity about what we think are unexplained costs, or in dispute. But, as I say, whoever wants to persuade you, as Assembly Members, to vote for a piece of legislation needs to be able to meet that test. And if it’s a Government Bill, we expect to have to meet that test and, as we know, this committee has been critical of RIAs in the past and has seen change in RIAs as a result of the criticism and the questions that this committee has asked. So, it’s the same process for Member-led legislation as it is for Government legislation, and I would expect you to apply the same rigour if I had to appear before you with a Government Bill in the future. I would not expect you to go soft on me.

I think you’re absolutely right, because we’ve got it wrong on one occasion when we massively underestimated the costs of a Bill and the local authorities had to pick up the pieces.

Rhun, you wanted to come in.

Yes. The challenge that I would give, though, is that the financial consequences and unexpected consequences of this Bill—and I’m remembering that I’m now on the Finance Committee, having been in the health committee at Stage 1 and looking at the policy side as well. One of your strongest arguments against this is essentially a financial one, in that there will be potentially unexpected consequences in the way money is spent in different parts of the NHS, and money being taken away from other services because of the need to fulfil the requirements of this Bill, should it become law, and so on. That’s why I think having those estimations from you of how much this would cost, even if not in total amounts of money extra that needs to be spent, but the kinds of moneys that need to be moved from one part of the Welsh budget to another in order to meet the needs of this Bill, is really vital for us as the Finance Committee to be able to decide what the financial implications of it are.

I would be confident not just to share the paper on, if you like, the principle concerns, which I guess are really policy imperatives, but to highlight the parts of that evidence that points out stuff that reiterates some of the concerns about use of money, but also, as I've said, to fill in some of the gaps in the commentary around the RIA, and why we don't think costs are either robust, or indeed where costs have not been provided but there is an inevitable cost consequence from the Bill. So, yes, I'm happy to do both those things. 

Okay. David and then—

The Bill obviously highlights, as you say, diagnosis and needs as a consequence of that. I'm conscious that there are many young adults out there—in fairness, you met some when you came down to my constituency—and the implications are there will be support for employment. That's a Department for Work and Pensions area, I would have thought. Have you had discussions with UK Government in the sense of implications across that might be upon them as a consequence of this, because they identify needs that can be served by DWP or other bodies?

I haven't had any conversation with the DWP about the Member's Bill. With the integrated autism service, there is a practical working relationship with some parts of the DWP. For example, when I was in north Wales, the conversation took place in a DWP centre where they actually undertake drop-in advice for people. I met someone who was there not by arrangement to see me but actually because they were taking part in the drop-in service for support and assistance to gain employment. So, there is a practical working relationship with the DWP that is being delivered within that integrated autism service. As for anything additional from this Bill, again, that is—

For the Member.

Well, it is his Bill. [Laughter.]

Can I just finally then, in conclusion, ask, I was, for example, a member of the Children, Young People and Education Committee when we scrutinised the additional learning needs legislation, the Government, of course, expected other public sector organisations to provide it with costings for services as part of the regulatory impact assessment that you produced for them—local authorities, health boards, et cetera—do you not accept that it's reasonable for a Member responsible to expect a greater deal of co-operation from Government in this respect? Albeit maybe the scale of ask is quite considerable, if I'm hearing what you're saying clearly, but do you not accept that it would be fair for a responsible Member to expect greater co-operation from Government, because, otherwise, where does that leave future potential pieces of legislation coming from backbenches in this place?

As I say, going back to my experience of being a backbencher and supporting a Member through the Bill process, we did undertake a range of that information gathering and robustness to underpin what we were doing with an independent source. There are plenty of health economists around with standing who could have helped, who could try and provide some robustness to that, and there would then be a different conversation to have. The costs that are set out in the Bill are looking at the diagnosis and looking at driving more demand through that diagnosis—as I said, it's a waiting-list initiative, essentially. However, if you're looking at the broader costs of the system, I would expect the Member in charge to have more information available about them, then of course there's a different quality of conversation to be had between us.

So, basically, you're saying had the approach been different, your approach would have been different. 

Yes, but then there's a choice made in this. The Member has chosen that this is the scheme of the Bill that he wants to promote, and there is a disagreement on the principle point of view about the Bill. So, in that case, there's a challenge of: does the Government say, 'We'll go along and help you to provide a range of costs for something that we think is a bad idea'? Well, if I said that I directed finite public resources to do that, I expect that you'd be asking me different questions about why on earth I've done that when I think it's not a great idea. 

There we are. Okay, well, that's a clear message at the end of the session, and that's probably it in a nutshell, isn't it, really? Okay, can I thank you and your officials for your time? We appreciate it very much. You will be sent a transcript, of course, to check for accuracy. We, as a committee, will now move into private session to consider the evidence that you've given us. Diolch yn fawr iawn. Thank you.

[Inaudible.]—produce a report in advance of that. 

Yes.

We'll make sure that the clerking team contacts your officials to be quite clear about what we expect and, hopefully, then you can be clear about how quickly you can get that information to us. Thank you very much. Diolch yn fawr. Okay, we'll now move into private session as a committee to conclude the rest of our deliberations.