Welcome to the Public Accounts Committee on Thursday 12 June 2025. The wealthiest in this country contribute significant amounts of tax revenue to the Exchequer. In 2023-24, £119 billion in personal tax was collected from around 850,000 wealthy individuals. However, the increased complexity of their tax affairs makes it more difficult to get their taxes right and can present opportunities to deliberately not pay enough. While there have been significant increases in the amount of tax collected from HMRC’s compliance activity in recent years, this suggests that levels of non‑compliance of wealthy individuals could be much higher than originally thought. I am sure that we will come to that later in the hearing. Across the autumn 2024 Budget and the spring statement in 2025, the Government committed to scaling up compliance activities, and granted additional funding to tackle serious offshore non-compliance, including fraud, by wealthy customers. This morning we will be examining HMRC’s knowledge about the tax contribution from wealthy taxpayers, as well as exploring HMRC’s understanding of the risks associated with its approach to tackling non-compliance. We will also be examining HMRC’s performance in increasing the tax collected that is owed from wealthy individuals. To help us with all that quite technical stuff, we are very pleased to welcome the new Permanent Secretary and chief executive, John-Paul Marks. You are very welcome to this Committee. I am sure that you will be making many appearances in the future. You have been in the role only since April 2025, having previously been Permanent Secretary to the Scottish Government, where you were responsible for the development, implementation and communication of Government policies and were the principal accounting officer there. Angela MacDonald is a regular attender at this Committee, deputy chief executive and Second Permanent Secretary. Angela has been in the role, as we all know, since August 2020, having worked in civil service since 2009, moving to HMRC in 2017 and becoming director general for customer services then. Penny Ciniewicz, you have been to the Committee many times and are director general of customer compliance. Penny has been in the role since 2017 and was previously chief executive of the Valuation Office Agency from 2009 to 2017, so you are at the front end of all of this. Jonathan Athow is director general of customer strategy and tax design. Jonathan was the department chief economist at HMRC and has also spent 14 years working in HM Treasury. He joined the ONS in 2015 and was the deputy national statistician and director general of economic statistics before joining HMRC in 2021. Philippa Madelin is director for wealthy and middle-sized businesses compliance. Philippa has been in her current role since December 2022, before which she was in operational and technical teams in counter-avoidance. Philippa has spent her career to date in HMRC, joining as a graduate in 2007, and has worked on a range of tax, operational and policy roles. As I have said, a particularly warm welcome to John-Paul and Philippa. I believe it is your first time, so welcome. Let us go into the main session. John-Paul, thank you very much for your letter of 10 June to the Committee about a phishing incident, which was a serious incident. What is interesting is whether this information was known when your predecessor appeared before this Committee before. The thrust of my interest in this matter is that I discovered it only when I read reports of the Treasury Select Committee having discovered it from the press. In this Committee we really want to encourage openness, candour and transparency from you. If there is a similar incident in the future, I would be most grateful if you could let us know. There is another channel for letting us know. I have been on this Committee for 13 years, on and off. Nothing has ever leaked when it is marked in confidence, so I am always happy to receive something in confidence if necessary. You can remain assured that it will remain in confidence, unless of course we discuss, you and I, as the sender and recipient, that it should not remain in confidence. I am asking for candour and openness from you. Let us know in the future. David, Treasury Officer of Accounts, did you want to say a word on this?

I will just explain, for the Committee and the viewers, the requirements of Managing Public Money in this area. There are three bits but, cutting to the chase, Managing Public Money sets a firm presumption and expectation that serious losses are notified to Parliament. However, Managing Public Money deliberately does not define what a serious loss is, because that will be very fact-specific, based on the nature of the organisation, the activities it undertakes and the volume. For an organisation such as HMRC, which is literally collecting hundreds of billions of pounds’ worth of tax revenue every year and dealing with tens of millions of customers, its threshold for materiality will be substantially different from an organisation such as the Treasury, which has a much smaller budget, does not deal directly with customers and is therefore less susceptible to fraud. It is, as always, for the accounting officers to take that judgment on what is appropriate to notify, and, where necessary, to come and explain that to you.

John-Paul, I am going to come to you in a minute. With this particular incident, a large number of taxpayers were affected in a significant way. That is why, I think, the press were attracted to the incident. If it is likely to attract significant press interest, it means it is of interest. In openness and candour, a simple letter to me, apart from anything else, will cover you. If I have had a letter, I shall not be necessarily needing to raise it in the Committee in this way. You might want to come back.

I have just a few points, Chair. Good morning to the Committee. I look forward to working with you in the years ahead. To reassure you on openness, candour and transparency, 100% I agree with you. We welcome your scrutiny, and that is my total commitment to those principles and how we work together. As you say, I wrote to you earlier this week on this serious incident, which was an unacceptable loss of £49 million to the Exchequer, affecting 100,000 of our customers, which is about 0.2% of the PAYE caseload. As the Treasury Officer of Accounts says, given that we collect over £840 billion a year, the judgment on materiality is different for HMRC, perhaps, than other Government Departments. None the less, I agree with the point with regard to disclosure and I will do that in my annual report, which I will publish next month for the first time, so that is properly done according to the rules under public money. The final thing to say is that I regret if there has been any frustration in terms of our handling of this. That was not our intent at all. My intent was to provide some verbal evidence last week and then respond to questions. I have written to you and the Chair of the Treasury Select Committee. I will respond to the correspondence I have also received this week with more detail. I welcome your point with regards to the opportunity to have in-private briefings. The level of security threat is significant and constant. The team detected and disrupted this one well. There was a criminal investigation. In private hearing, I am happy to bring the head of the fraud and investigation service, my chief security officer, to explain more about some of that detail, but also the threat environment and the way in which we are ensuring HMRC is secure now and for the future as well.

That is a very helpful and candid opening remark. This Committee is acutely interested in cyber-attacks everywhere in the government sector. We had a separate hearing on this. We are concerned about where Departments have legacy systems that are maybe more open to cyberattacks. I think it is a subject we will return to, and no doubt you will want to pay very close attention to that too. We all know—you, Marks & Spencer, the British Library—how very disruptive, to customers particularly, these sorts of attacks can be. Thank you very much for those opening remarks. We will move on now, please. We warned you at the top of the session about your Treasury minute response to our recommendations on call waiting times. You have been warned—good.

Thank you to panel members for joining the session. As some background for context, members of the Committee and panel members will recall that, in the Public Accounts Committee’s January 2025 report on HMRC’s customer service accounts, the Committee concluded that there were some issues with the telephone services and that HMRC did not give enough consideration to the needs of customers. It was a constant feature that Members spoke about. Taxpayers are customers for HMRC, but they have no choice as to where they can take their business, so to speak, so their engagement with HMRC is not a normal customer-provider engagement. There were issues around the number of individuals who could access someone to speak to at the other end of the phone. Only just over 66% of customers spoke to somebody, against a target of 85%. I want to ask why HMRC is still not willing to commit to an average waiting time target for customers’ phone calls.

I agree with your fundamental principles that we are a customer service organisation. People need to be able to talk to us multi-channel, and our telephony service needs to be there for everyone who needs it. We are committed to delivering our service standards. It is important that we do. I am confident that, in the year ahead, we are going to continue to see that improvement compared to last year, and we are focused on that. On your specific point on speed of answer, I had a look at the data for last week and last month. It is much better compared to last year. We were at about 11 minutes last week and 13 minutes last month. We will publish all that data shortly, too. Average speed of answer is a key indicator for us, and I agree with you that it needs to be. We look at it all the time. Ultimately, the percentage of calls answered at around 85% equates to around 10 minutes average speed of answer. Certainly in my time working in DWP, that would be the point at which you would often see people disconnect, give up and phone back again. Obviously you do not want that for their experience, but also in terms of failure demand it is not efficient. We have that lead indicator of an 85% customer service standard for percentage calls answered or adviser attempts handled. That equates to that key indicator on speed of answer. Our objective is to be at 85% or better, and that should be at about 10 minutes speed of answer.

What is the percentage at the moment?

Angela might have the latest. In March 2025 we were at 80.2%. I think that we are going to be better in April again, and I am hoping, for this financial year, across the year, we will hit the standard, but Angela can give you the very latest.

I have the data if that would be useful. We hit 83.6% in April. That compared to 53.5% the prior year. We hit 84.6% in May. That compared to 60.4% the prior year.

John-Paul, you referred to a 10-minute average waiting time. Therefore there has been an improvement, because the average core waiting time when you spoke to the Committee previously was a 23-minute wait.

It was, absolutely. The overall result for our 2024-25 figures, so for the entirety of the year, will be 18 minutes, 38 seconds. Actually, for March it was 14 minutes, 44 seconds. In April, it was 13 minutes, 31 seconds. In May, it was 12 minutes, 45 seconds.

Do you have a target then, but you are just not willing to say it publicly?

We absolutely have an internal expectation of what we are aiming for, and we track it. When we are having our very regular performance reviews, the satisfaction with the calls, how many customers get through to speak to somebody and how long they are waiting are all data that we internally monitor very actively. If you think about these as published targets, there are so many targets we could choose to have—which of the many?” Our average speed to answer is tracked internally. It is externally published and we actively monitor it. We take it very seriously.

The speed at which a phone call is answered is very important. There is a strong sense from this Committee that it was an important target, particularly because it was the case that, after 60 minutes, over 40,000 people had been cut off. I think that it was just in an 11-month period. I am unclear whether that is still the case. I think that that still happens, which is a real concern.

Perhaps I might share. You made a number of excellent recommendations when we were last before you, the majority of which we were very pleased to agree to: accurate call waiting times for customers waiting in the queues, avoiding completely the cutting off, a call-back service, and having the resourcing and funding to be able to hit our targets. Only a few weeks ago, we put out into the market a procurement for a complete replacement to our telephony platform, demonstrating a multi-million-pound commitment to telephony being a key part of our channel mix. That will solve those problems. It will give us the technology to remove completely the 70-minute cut-off, which is part of a system limitation that we have at the moment. That will remove that completely. It will also enable us, in real time, to tell customers where they are in the queue and how long they are waiting. Those recommendations are about ensuring that we have the capability to do that and, as I say, procurement to enable us to sort that out.

I am pleased to hear that, because, of all the things we heard, that was the one I think Members were most exercised about. It is one thing having to wait more than half an hour on the phone, but then being cut off after you have spent an hour is quite poor customer service.

We agree.

I am pleased that there has been some movement and it sounds like there will be some change. Very finally, we heard from HMRC that one of the challenges with speaking to customers is the security of being able to exchange communication and customers being able to send you documents, for example. We did not understand why it is the case that the NHS, for example, can do it and HMRC could not. In the brief time we have, do you have any updates or key milestones you can indicate to the Committee on that?

We absolutely want to be in a situation where we can stop the letter exchange with customers and move into a secure communications environment. The challenge is that the standard email is not a secure environment. I refer to the earlier conversation that we were having with the Chair. A secure environment to do that is absolutely vital. We have, in the last number of months, done some testing. There are many IT platforms out there that do this kind of thing. We have done some testing on what our requirements would be and how it would work exactly. We were very pleased yesterday to receive, in the spending review, the funding that will allow us to procure that platform and roll that out across, starting with compliance. The exchange of some quite large-scale documentation is a very big feature of many compliance activities. We will then be rolling that out across customer services. Now that we, after yesterday, know that we have the funding for that, we will go ahead to procure. We have been working in preparation. The procurement will come out very shortly and then we will roll that out across the IT estate, across the entirety of our services, through the SR period.

I am glad that the spending review was a positive outcome for HMRC. We will be watching tightly the procurement process so that can be swiftly implemented.

I totally agree with the thrust of Nesil’s questions. Let me put this concept to you. Any organisation is better with satisfied customers than dissatisfied customers, and no more so than HMRC. If they are satisfied, they will be much easier to deal with in the future. Finally, Angela, can you give us any idea of timescale? You have to procure it, test it and get it into operation. What is the realistic timetable for getting it into operation?

Being realistic, these are really very large procurements. You understand, Chair, the steps we must go through in government procurement. It is likely to be into 2026-27 when we get our first pieces in. However, as I say, we have been getting on doing some testing, actually in Penny’s area. We are working currently with the Treasury and the Cabinet Office to see, across all our procurement processes, how we can accelerate procurement. Government procurement has many steps. Can we move in a productive and agile way to get through these things a lot more quickly? We are getting fantastic co-operation from our colleagues in other Government Departments. This is key for us and we are pressing very hard. The Minister is very supportive. We are doing everything we can but, being realistic, this is big change. It is going to be next year, but now we have the funding it is full steam ahead.

I understand that and thank you for those very candid answers. We now come to the bulk of the report and the subject that we are here to discuss today, the payment of tax by wealthy individuals.

I want to thank the NAO for the report. Chair, you made the point around contributing to our knowledge of the subject, and the model and the approach. There has been a lot of collaboration with the team. We have accepted the recommendations and look forward to implementing all of them. In terms of the opportunity, it is significant. I would probably summarise it as capacity and capability. In terms of capacity, Philippa runs a team of around 1,000 civil servants and we have secured the funding to increase that by another 400. We have scored at least an additional £500 million worth of yield as a result of that, so there is more capacity for Penny and Philippa to undertake more activity. At the same time, we want to ensure that the team have the enablers that they need and the capabilities to be more effective and productive in their work. Those include data exchange, targeting and risking, and opportunities for us to collaborate globally with jurisdictions to ensure that we are operating offshore to the best effect, but also looking at opportunities around policy change that will help us close that tax gap. The team will say a bit more. We are doing things such as recruiting and bringing in external specialists. We are going to be investing in compliance tools and, on the policy side, looking at things such as crypto-asset reporting frameworks and making sure that we are getting all the information in that we can, so the data-led compliance model that is so important to being effective here is delivering the results that your constituents are looking for.

In terms of a timetable, the international data exchange that JP referred to has been really important in recent years. Compared to 10 years ago, we now receive a lot of information on people’s financial accounts offshore, but also entities that they control through the common reporting standard and other data exchanges. As JP mentioned, we continue to work internationally to get more of that data. That will include crypto-asset reporting, but that does not come until about 2027. More recently, we have had the register of overseas entities, which gives us more information on what trusts people might own offshore. Alongside that data, we always do upfront education to encourage people to either amend their returns or tell us about tax that they should have paid. Once we get that data, we are able to target effectively our compliance efforts and then report back on how well we have used that data in due course. That data and compliance effort and the policy change go hand in hand, and you will see that in our increased results over the last few years.

It is a good question. It is a constant endeavour because, ultimately, those who are deliberately going to try to evade tax will continue to try to do so. Our job is to work internationally to increase that data exchange so that, increasingly, there are fewer places for people to hide their assets. We are able to then find those identifying factors, whether that is a big balance offshore that they might not have told us about or a company that they are not reporting income from, and use our investigative experts to help get that money in. We are transparent about how that reflects in compliance yield results but also through the tax gap that we publish. In terms of targeting that, we have said at high level that we want to bring in an extra £6.5 billion in the scorecard period to close the tax gap overall, but we have committed to bring in at least an extra £500 million from wealthy offshore risks in the next three or four years.

I draw attention to my Register of Members’ Financial Interests of the assets that I own. Your register of overseas entities, as I understand it, comes into operation in April this year and only Gibraltar has so far actually filed such a register. Is that correct?

We started getting register of overseas entities data at the end of January 2023. We get that from a number of jurisdictions, but the reporting we get is highly dependent on what central register each jurisdiction holds. When we do not get that automatically, we are also able to make requests where we might suspect that someone has filed an incorrect return or not told us about something. We also do outbound requests to get data if it does not come to us automatically. It varies per jurisdiction, depending on their own approach, but we have had data from over 40 jurisdictions since 31 January 2023, and we use that in our compliance efforts today.

How many more are you chasing?

Do you mean in terms of jurisdictions?

Yes.

At the moment, in the common reporting standard we have over 100 jurisdictions. In that, we have all the major global financial centres, but there will still be people who are not in there. We have places such as Trinidad and Tobago or the US Virgin Islands, for example, that are not in there, but all the global financial centres are in the common reporting standard. The register of overseas entities is a slightly smaller subset of that, but, again, we have targeted that on the major financial centres.

On data, you mentioned this international collaboration. You have agreements with 93, do you?

It is over 100 now, I think.

You are slightly pre-empting my question. Given that there are probably about 180 countries in the world, what are you doing to increase it still further?

We work with the OECD and look at the geographical reach. As I said, we have all the major global financial centres, but the OECD then has a special forum where it looks at whether there are other jurisdictions that might be of interest that we need to then work with and put on to the list. It is done on an analysis of not just geography, but the scope of the reporting itself. That forum decides whether we need to bring that jurisdiction in now because we are seeing a lot of people setting up accounts there, for example.

That is very helpful to have that all on the record.

We segment our approach and look at wealth, but it is not the only indicator of risk. We also look at complexity and opportunity. Of course, wealthy individuals have greater access to, for example, tax advisers who might help them with their planning. Over the years, historically, we had an ultra-high net worth unit, which focused only on wealth. We found that that actually meant that we were missing other risks that might present themselves, because wealth alone is not the only indicator. Over time, we have then built the wealthy operational unit that sits in my area, and that allows us to look across a number of risks. For example, you could have a billionaire who actually could have quite boring tax affairs and report everything that they owe. Then you could have a millionaire or someone who sits beneath that who has set up really complex offshore trusts and structures. We found that taking that segmented approach is better to help us identify risk. For the most complex and riskiest of our individuals who owe UK tax and our wealthy population, we deploy customer compliance managers, who have a more in-depth understanding of those who we feel pose the greatest risk to the tax system.

Yes. It is both. Wealth will continue to be an indicator, but risk as well.

Absolutely, we are not complacent. Our ambitions are undimmed there. We talked about the enhanced data exchange, which has been transformational in us being able to understand the types of risks we might see offshore, but we continue to work on that. JP already mentioned capacity and capability. Through the spending review investments and other investment that we have talked about, we will have more people, which is important, investing in our existing workforce to give them the skills and capability they need, but also trying to bring in more wealth management experts, for example, to help us understand, beyond tax, how people do their wealth planning. From a system perspective, it is really important for us, with that vast amount of data that we get, to make sure that we are able to link that data effectively, target and pinpoint our compliance activity, and then make sure we are as productive as possible in that work. As you can imagine, with some of the really complex arrangements that we see, we get vast amounts of data or information from customers. Being able to better mine that to find out where we need to focus our efforts will be a key part of this investment.

Good morning, Philippa. I have a quick opening question. Exactly how many billionaires currently pay tax in the UK?

We do not hold a list of billionaires because, as I just explained, we do not look at wealth alone. We publish official statistics on the top income taxpayers and capital gains taxpayers, so you will be able to find that on gov.uk, but we do not segment according to how many billionaires there are who pay UK tax. We just look at the UK taxpaying population overall.

You are saying that the reason why you do not know how many billionaires pay tax in the United Kingdom is either because of a lack of information or because you look at it only on a specific case-by-case basis.

Yes, we look at the amounts of UK tax income and gains that are owed. There is not a current requirement in UK tax law to report your total wealth. It is only when you have UK income that is taxable, and also on chargeable events such as when you sell an asset. We do not collate based on wealth data alone, because that is not a current requirement and so it is not the data that we collect.

Do you think that it is effective to pick and choose which taxpayers you take a closer look at—those super-rich taxpayers—rather than to do it by bulk and look at every single one and all the data that is collected? Why are you doing it in that ad hoc way where you only go at certain cases?

We actually risk assess all of the wealthy population and then segment it according to those who pose the most risk. We have a risk-based approach to the majority of those and then we have our customer compliance managers for those who, through that risk process, we identify to pose the greatest risk in terms of complexity or opportunity. We look at all and then home in and segment. We deploy our most senior tax professionals to those who we believe need that.

To be clear, you said that you risk assess everyone.

Yes.

Going back to my first question, if you risk assess everyone, could you tell us how many billionaires there are paying tax in the United Kingdom?

We do not risk assess according to wealth. We risk assess according to chargeability to UK taxation, which will be on income and gains.

Touching on the points that my colleague Oliver mentioned, do you not think it would be helpful to have some sort of segmentation so we can look at the very wealthiest. There is a broad spectrum of wealthy individuals between those earning several hundred thousand pounds a year and billionaires or multi-billionaires. Do you not accept that, perhaps, if you can risk assess everyone, it would be useful to be able to segment to see exactly how many billionaires are paying, or not paying, tax in the United Kingdom?

I will hand over to Penny in a second, because I think she wants to come in. As I say, we risk assess and segment them all, but not according to wealth at this stage.

Our risk assessment is using multiple data sources. Philippa has described some of the range of data that we have available. It is significant. It is a data-based risk segmentation. Pure wealth is not the only factor in that risk assessment. As we have talked about, it is opportunity. Also, we have to use the data that relates to the responsibilities that we have, which are the taxable assets and taxable events that relate to UK taxation.

Out of interest, both Penny and Philippa, could you, if you wanted to, find out the answer to that question internally of how many billionaires pay tax currently in the United Kingdom, or are you not able to do so?

I do not think that it is a question that we seek to answer. We are looking not at the total wealth of an individual, but at the wealth in relation to the taxable events that relate to the UK tax system.

I understand that, but I would like to know whether HMRC has the resource and ability to find out how many billionaires are or are not paying tax in the United Kingdom. Whether or not you choose or seek to do so, are you able to?

We collect the data that is needed to administer the tax system as is and there is no tax on wealth. There is tax on certain transactions, so when those transactions happen—

I very much do not want to get into tax policy. As the Chair will rightly say, that is not the work of this Committee. It is just whether HMRC has the resource and the tools to do so.

We have only the data that is needed to administer the tax system. As there is no tax on overall wealth, that information will not be routinely collected.

To conclude, you do not really know how many billionaires pay tax in the UK.

Because of the way we operate the tax system, whether you are a billionaire does not determine whether you pay income tax. It is the income tax and chargeable gains that determine whether you pay that tax, and that is what we administer.

To help both of you in this dialogue, I gather that the Sunday Times Rich List 2025 listed 156 billionaires, compared with 165 last year, the steepest drop in the list’s 37-year history. That gives us a clue of roughly how many billionaires there are in this country. Whether they all are qualifying for tax or not, I do not know. It is not a terribly difficult piece of computation to look at that and see who is paying tax. As it says—I was looking for the reference; I cannot find it—somewhere in the NAO Report, even one case of getting the tax right for these billionaires can absolutely change the dial on how much tax the wealthy pay. I wonder whether you could pay a little more attention to this.

As Penny and Philippa have said, we will look at a number of factors in terms of identifying risk. There is a wide variety of data available to us from third parties and elsewhere. The key issue is that, if I was to look at that rich list, somebody will have had to make a judgment about whether these people are tax resident and about the value of their assets.

The Chair makes a really important point. Has HMRC ever tried to link the Sunday Times Rich List with your own tax records to estimate how many billionaires listed are tax resident here in the UK? Has that work ever been completed?

I would not want to get drawn on any specific data that would be used. We will use a variety of third-party data to make certain we are aiming our efforts as effectively as possible on those to make certain we have the compliance.

Is that third-party data collected in bulk? Do you analyse all of it, or do you do it on an ad hoc basis?

We use a variety of data, including open-source data that would be publicly available. We would clash that with third-party data. We have already talked a little bit about what we get from offshore, but of course we also get data from UK financial institutions. We work closely with Companies House, the Insolvency Service and others, including Land Registry information. We get a lot of information and data sent to us through things such as stamp duty land tax returns and our self‑assessment returns. We use a huge combination of data: third party, open source and the information that customers give us.

That is reassuring to hear. My understanding is that, in the United States, they allow academic researchers to make that link between tax records and the Forbes 400 list, which I suppose is their equivalent of the Sunday Times Rich List. Would you allow HMRC to do a similar thing, where it works with academics and others to do that linking-up work?

We run a data lab and allow people access to our data for research purposes. If they have other data that they wish to bring in to match to our data, we would allow that. In theory, if people had other information and wanted to use that to do analysis, it is open to academics to come in and do further research on the data we hold. We obviously would not allow any data out that would be identifiable for any taxpayer, so there are a lot of protections around that, but we allow researchers in to look at a number of different issues. I know, for example, that there has been research reports recently on aspects of non-dom policy and those sorts of things. This is an area where high-income or high-net-worth individuals are a focus of some of our research reports. That is open for researchers to come in and do academic research, using our data in our secure data lab environment.

Before we move on, I have one final question, and I think I probably know the answer. Does HMRC know internally, and is it willing to publish it if it does, how much tax billionaires paid to HMRC in the previous financial year? Do you have that information, or do you not think it is worth having if you do not?

We do not collate it in that way. As I said before, simply being a billionaire does not necessarily mean you are chargeable to UK tax. Particularly before this year, when the non-dom changes came in, a large number of billionaires may have been non-domicile in the UK and had no charge to taxation here.

Now that that change to the non-dom tax regime has been made, do you not think it would be valuable for HMRC to know how many billionaires are paying tax in the UK and how many paid tax in the UK in the last financial year? If you generate your wealth here, do your business here and are tax resident here, we should probably know how much tax you are paying, as one of the wealthiest individuals in the country.

Absolutely, and we can look at the points you have made on research. We are always happy, as Jonathan said, to look at research and compare that with what we need to know. I just would not want to get overly focused on billionaires alone, because we see a lot of risk at lower wealth bands, not just from billionaires.

On the non-dom point though, to reassure you, we will start receiving the data in the beginning of 2027 tax returns with regards to the change to non-dom rules and publish the formal evaluation thereafter. To your point of the extent to which we will be able to evaluate and transparently explain the effects, the OBR has scored its assumptions, but we will, of course, evaluate that carefully and then make it available.

Going on the point that you just made, Philippa, about expanding our view from not just billionaires but to the very wealthy in general, is there anything stopping HMRC from publishing the total amount of tax paid by the very wealthiest in this country? Like you said, you do not segment it, but you must have an idea of what the very wealthy are paying as a whole. Is there anything stopping you from publishing that?

We know how many people there are on PAYE, for instance.

The problem is that our information on wealth is going to be incomplete. If somebody has had a transaction recently, so perhaps they have sold a property recently, we will have that information. If they have not traded anything recently, we will not have an up-to-date valuation for that asset. We cannot have, because we do not have the tax policy structured in that way, a total view of somebody’s wealth. We will have estimates. We will have information on what recent transactions may have arisen, but the idea that we will have a total view of somebody’s wealth is not necessarily true. We will have what we know, what we have as information, and, as Philippa said, what we have as third-party information.

Perhaps, JP, it would be useful for you to come in here, as a new chief executive of HMRC. As my colleague Nesil has just pointed out, we know how much tax is collected through PAYE, which is how the majority of our constituents pay their tax every month. Do you understand that, actually, if we were to publish more information about the total tax paid by the very wealthiest in this country, that would create much more transparency and public confidence in HMRC, so people felt that taxes were being collected fairly and that you were doing your job effectively? That answer, “We can publish it for one but not the other because it is not complete”, is particularly galling for many taxpayers.

This underlying conversation shows that there is more to do. The NAO Report talks about us making sure we are reviewing our population definitions, given how that is changing. There is a recommendation about greater transparency with regard to the amount of tax wealthy taxpayers pay and the contributions of different segments within it. We have committed to looking at that recommendation and coming forward with a more comprehensive response of how we will respond. I am with you totally. We want to improve trust in the UK tax system and HMRC. It has gone down in recent years. We want it to recover. Transparency is one way for us to do that. We will take that recommendation from the report and look at opportunities, whether it is with regards to the data lab, research or providing reassurance that, indeed, we understand the billionaires. They are part of the population that is segmented according to our risk model. Where we judge that there is a risk that needs investigating, of course we would do so, because our objective is to maximise the return to the Exchequer given our duties as commissioners.

Are you able to perhaps come back with some more written information to this Committee to explain how you are going to create greater transparency around how much tax is collected by HMRC from the very wealthiest? That is essential for improving the public’s confidence in your ability at HMRC to do your job of collecting taxes in this country. Can you reassure us that that written information will be forthcoming?

To reassure the Committee, we publish the top income tax payments and capital gains tax payments already, but we have not segmented in the way that you proposed there, so we will take that away.

I have just one final point I would like to go on to. Why has the amount of income tax paid by the very wealthiest in this country gone down, falling from £4.4 billion in 2009-10 to just £3.5 billion in 2014-15?

There will be a number of different factors that would go into the total amount of income tax paid. That will be a mixture of what is happening globally in that period, but also, as we talked about before, where the wealthy are resident and reporting that income. Beyond that, there will be various market factors that influence the total income tax take overall. I do not know whether anyone else on the panel has anything to add.

I am not certain I know exactly the figures you are talking about.

They are from the NAO Report. It has fallen from £4.4 billion in 2009-10 to £3.5 billion in 2014-15.

There will have been a number of economic factors that affected income tax from those individuals.

Could one of the factors be that there are fewer penalties being issued to the super rich?

Overall, though, the receipts from the wealthy have gone up. They were £119 billion, as the report outlined. Although different components of it might change, obviously we get wealthy receipts from income tax, capital gains tax, inheritance tax and stamp duty land tax. Income tax alone is not the only indicator of the overall receipts, which actually increased in 2023-24.

I should correct the record there. Those figures are from a previous Public Accounts Committee report on this subject. That is not from the NAO, but from a previous report. Those are the figures regardless. I just want to make sure we have the correct source on record. Also, is that the most recent data? Is that £3.5 billion figure for the amount of income tax paid by the super-rich from 2014-15 the most recent data that is publicly available or that you are able to share with us?

No. I said earlier that we publish the top 50 income tax payments as official statistics. Also, in our annual report and accounts, we estimate how many overall receipts come from the wealthy, which, as I said, went up in 2023-24 to £119 billion.

Could we get a figure for the financial years since 2014-15? I was still at university at the time of that figure. It would be great to see whether that has gone up, down or stayed the same.

We will take that away and see what we can get to you.

It is a decade old. If there is not anything out there in the public domain, that is quite frustrating for us to understand whether the very wealthiest pay their income tax, just the same as all of my constituents do, who are not super rich.

That is fine. We are happy to write to you on that.

I do not want to dwell overly on billionaires and I absolutely take Philippa’s answers and agree with them that it is not just billionaires where the biggest risk is. There is a whole category below that. John-Paul, a bit of inspiration has come to me. With a caution about what I am about to say—we certainly do not deal with individual tax cases on this Committee, and I hope what I am saying and the inspiration that I have had is not going to identify anybody—we know that one large tax case from 2022 to 2024 yielded over £2.5 billion. If that is correct, it is a huge amount of money. It rather drives home the point that Lloyd is making that we need to pay quite close attention to this class of taxpayers.

Penny and Philippa might say a bit more. As the team has been trying to emphasise, absolutely, we need to pay attention to follow the risks and then investigate them very carefully and thoroughly to maximise the yield for the Exchequer. That is the construct of the system that Philippa set out just now. We would agree that, if there was a significant exposure or liability that was due and known, the team would absolutely prioritise resources and investigate accordingly. In fact, there are legal duties on us to do exactly that. It is not optional. We would agree. I do not recognise exactly the numbers on that case, but the team may be able to help us. You are absolutely right. There are some big cases, some of which have been in the public domain already, and I have run through some of those. They take time. We have the capacity to investigate them and we will. Penny and I were talking this morning around the number of prosecutions, the number of serious cases and our expectation that both of those will rise in the years ahead.

We will come on to prosecutions. Unless any of the team want to tell me that those figures are wrong, we should move on from this subject.

I do not recognise that number relating to a single case, so I just put that on the record, but JP’s points stand.

Two million pounds’ worth of assets does not seem a huge amount of money these days. You could own a three-bedroom house in Fulham and have £2 million worth of assets. Similarly, wage inflation means that there are more people earning more than £200,000 a year. The number of the wealthy has gone up by 150,000 in the last five years. Why do you not have a bracket for the super-wealthy, so people who have £10 million of assets? It seems to me that they should be part of your target audience as well. If the super-wealthy group expands, which it is going to do, not least because people are greedy, how will HMRC respond?

Ultimately, I would agree with your point that the population is increasing. As you mentioned, it has gone from 700,000 to 850,000. It is important within that population that we target risk, and wealth and income is one of those factors. When we look at the cases that our compliance case managers are working on, the vast majority are cases where individuals have income in excess of £10 million. In effect, by using a risk model, we arrive at the segmentation that you are suggesting we should, which I agree with, because we need to focus on yield, opportunity and risk, but we do not start from that assumption. We allow the risk model to take us there. If there was a case that was at £8 million, £7 million or £6 million, we would not disregard it. It might be that non-compliance and opportunity is such that actually the exposure to Exchequer is higher than the billionaire who has conventional, lawful, well‑planned tax affairs. The technical term was “boring”.

You do not think that your current definition is producing too broad and varied a pool of wealthy taxpayers for you to treat them as one single group?

We need to keep it under review. The point the NAO makes about keeping the population in clear view, the segmentation point, is well made. We do not disagree about the importance of focusing on the highest risks to the Exchequer and the taxpayer and bringing in the yield that our public services need. We think that the risk model is the right way to do that. If we need to change or tweak that a bit, partly to give public confidence or in regards to the conversation we just had on transparency, of course we are open to that. There is also the investment of additional capacity. Your point was that the population is getting bigger. Yes, absolutely, so we need to increase the team. We are going to add 400 to the team of 1,000 to bring an extra £500 million in.

If it is any level of reassurance, we have adjusted that definition, in my time as director general, not in Philippa’s time as director. We keep it actively under review. We are interested in results and in managing the population to the best of our ability. We do not let any definitions get in the way of that.

JP, you touched on this a bit earlier in the comments you made. Do you see different risks from the very, very wealthy as opposed to just the wealthy?

Philippa, you already touched on this.

I would not necessarily say that through wealth bands. You see that people’s propensity for risk will vary. For example, you could have a very wealthy person who is paid a lot through pay-as-you-earn who may not have any offshore investments. Therefore all their income is very visible to us. You might have another individual who owns various properties, for example, in the UK, and decides to put those into a trust, then set up an offshore trust and transfer between different companies. They might, to us, be a millionaire, but be very non-compliant. It is that range of risks that is why we find it really important to have a risk model rather than a wealth model, because otherwise we might be missing out on people who are causing more harm to the tax system than just the very wealthy.

I will cover the headlines of the measure and then we will come to the point on the compliance model. Overall, the OBR certified a £34 billion central estimate for additional revenue, so in terms of that scale over the five-year period. The OBR then assumed a behavioural response of 12% to 25% of non-domiciled individuals to leave the UK this year. That assumption will be ultimately understood in the outturn. We will start to get data in January 2027 tax returns, which we will be able to evaluate.

That is within the assumption of the £34 billion.

That is already within that.

The £34 billion over five years assumes that behavioural response. Obviously that 12% to 25% range has a high degree of uncertainty. We will know with more certainty in 2027.

Could you go into a bit more detail about what work HMRC is currently undertaking to ensure that the changes to the non-dom tax regime actually mean that more tax is collected? That is what the majority of our constituents probably are keen to see. This change has been much trumpeted and I think we would all want to see that actually it is going to lead to more tax being collected by HMRC. They will want to have confidence that you are doing everything in your power to make sure that this tax change increases your revenues.

There are a few features of the new regime that seek to smooth that, to reduce that behavioural response, and to encourage new arrivals who have not been UK tax resident in the previous 10 years. This is the four-year foreign income and gains regime. There is a temporary repatriation facility, along with making sure that the way we provide our case management and support for customers is of a high standard as well. The team could say a bit more about the compliance model. In terms of the implementation of the regime, the guidance, working with representative bodies, explaining how the regime will work and making sure that customers who are either coming to the UK or are currently in the UK understand it and get the support they need, that is an important priority for us.

The key part is that we work really closely with policy colleagues in the design of the legislation, so that we are able to put our operational insight as to the risks that we have seen from non‑domicile or offshore interests into the design of the legislation in terms of what information we might ask for, or, indeed, some of the offshore avoidance legislation that we have recently done a call for evidence on. It is that combination of applying more resourcing, but also designing your policies and processes in a way such that people cannot get round them once you put them into place. This is the biggest reform in over 200 years of taking domicile out of the UK tax system and replacing it with a residence test, which, as JP said, is really internationally competitive. Ultimately, it will mean that, for the first time, people who are earning and resident here, and putting their income and gains here, have to report on and pay on their worldwide income and gains.

Without wanting to stray into tax policy, for all the obvious reasons, the United Kingdom is, at the moment, the only country in the G7, aside from Italy, that allows very wealthy individuals to build up capital gains while living here. They do not have any obligation to pay tax on these gains if they then leave before selling. Would it perhaps be a fairer way of ensuring that closing the non-dom tax regime or loophole is effective if we were to look at that? Has HMRC done any work internally to look at what that exit tax might do in terms of making the non-dom tax regime changes truly effective and, going back to my first question, collect the right amount of tax?

The new regime is already in place. It started this April. There are transitional arrangements. Because we are taking out domicile, that is one margin that is no longer there for people to use. In terms of compliance, we are now moving to a strict residency-based system. We have a statutory residency test, so that is now a much clearer basis on which to ensure compliance. In that sense, that should be more straightforward. There are a number of different angles to this, because many of these people will still have complex affairs that Philippa’s team needs to oversee, but the core principle of moving to a residency basis should allow us to make certain that we are more easily identifying and spotting issues.

I appreciate that it is already in place, but there does seem to be some uncertainty around how much tax will be collected through the changes to non-dom. With that in mind, and there being a risk there, has HMRC done the necessary internal work to look at whether an exit tax or something along those lines would make the changes that have already been made to non-dom effective, and collect the amount of tax that it rightly should?

On the specifics, policy will be a matter for Ministers to decide on at the appropriate fiscal events, but, in terms of the changes, we have recently done a call for evidence on our offshore avoidance legislation. That includes transfer of assets abroad, settlements legislation, and various anti-avoidance of capital gains tax measures. The reason for that is exactly to ask, “Are these fit for purpose?” particularly since the non-dom reforms, and we will be reporting on those summaries in due course. That work is in train to look at how we make sure that people pay the tax that they owe and do not use contrived arrangements to get around that.

Are you able to give us a timeframe for when that work began and when we can hope to see it conclude?

The call for evidence was in the spring. I believe it closed in March 2025, and we expect to be able to publish responses shortly.

Is looking at an exit tax part of HMRC’s consideration?

I do not think that we can go there, because that will be a decision for Ministers and that is a policy matter.

Would we have done what, sorry?

It is government policy. Our job is to administer it. In all seriousness, that is the best that we can do. I made the point around the OBR.

That is a fair answer. I do not think that we can go any further than that. We have to be careful about straying from facts into policy. This Committee does not do policy, so we have to be careful about that. I am now going to suggest that the Committee takes a break. Sitting suspended. On resuming—

It just so happens that the first question is down to me, and it is on tax agents. Penny, how will you develop your approach to better target specific risks such as unscrupulous tax agents? In asking that question, before I have every tax agent writing to me, the vast bulk, as I know from my own advisers, are very upstanding and make sure that their clients do pay the right amount of tax, but there are a small number who are quite happy to make sure that their clients do not necessarily comply with the regulations. How are you going to deal with that?

I would have said that as well. Before I start, the vast majority of agents seek to do the right thing, but our work to tackle agents who facilitate non-compliance is extremely important. We have recently combined our agent compliance team with our specialist team who tackle promoters of avoidance, so that we can cross-fertilise the skills between the two. We work collaboratively and use a range of tools at our disposal to tackle those intermediaries who are facilitating non-compliance. We have around 290 people in those combined teams and, over the past four years, they have brought in over £3 billion between them in relation to that work. When we see unacceptable behaviour in the population, teams across my compliance work will escalate those. We focus on the behaviours of agents in general rather than necessarily specifically those with wealthy clients. Agents in general may have a mixed portfolio of clients. We do know those agents who are particularly supporting wealthy clients. We are looking all the time at how we can improve our work to tackle those non-compliant agents. We published a consultation very recently on how we tackle the conduct of agents who are not meeting our standards. We also do a range of things such as referrals to professional bodies. We are also working at the moment to develop a risking model in particular to look at agents and their behaviour.

Do you have any sort of mechanism whereby you have a group of agents who you regularly correspond with, so that, particularly when you make tax changes, if there are consistent problems with that tax change, you can issue the appropriate guidance to all 33,000 tax advisers?

I chair a group called the compliance reform forum. The minutes are published on the HMRC website. We work with the representative bodies for the majority of agents. We work very closely with them on a whole range of issues, but we certainly want to make sure that they are really well informed. Jonathan also chairs the representative bodies steering group, which covers a whole range of changes in HMRC policy and service issues. We seek to have a really fluid dialogue all the time with agents, and we certainly consult with them on a whole range of issues. We bring them in, for instance, to design our “one to many” letters and to help give feedback on the quality of our communications with taxpayers to make sure that they are as clear as possible. They are a really valuable resource for us, but we also welcome any insights that help us to understand if agents are not doing the right thing, and we will pick those up.

If you pick up a particular problem from an individual case, it might not be the taxpayer’s or the tax adviser’s fault, but there is clearly a glitch or an unclarity—if there is such a word—in the system. How do you take that learning on board?

We are constantly looking to understand whether our guidance is clear, particularly in the area, for instance, of legal interpretation, which is an area where we are paying particular attention. It is a part of the tax gap that we are keen to reduce. In those kinds of areas, we put out new guidance. We have this set of guidance notes called guidelines for compliance, and we try to identify areas that are particularly contentious or difficult and set out clearer guidance in those areas. A number of those have been published in recent years. It is important that everybody sees those, so it is not just the production of them. We want to work, again, with the rep bodies to really understand that people are getting to see those and are using them to help them to be compliant.

What about your system of pre-notification prior to submitting a tax return, so that the tax adviser or taxpayer has much more time before a deadline to discuss any areas that might be difficult?

We try to have an open dialogue. Philippa’s customer compliance managers in particular may be in a very open dialogue, particularly with our wealthy customers, to try to understand the risks that they are presenting, and to have a free-flow conversation about what particular events might have happened in a year that would mean that there were taxable events coming in that we should be clear on, before they file. We certainly do keep that dialogue going as well as then looking at the filing and working out whether there are any risks that are presented that we would want to open an inquiry into. It is in our and the taxpayers’ interests that we get as much right up front as we can.

Get it right first time.

Yes.

Very good. Thank you very much.

I am going to ask a bit more about tax agents and the associated risks. Nearly a quarter of wealthy individuals use agents, who are, basically, accountants, for the ordinary person who might not be able to afford an accountant. I want to understand a bit more about non-compliance and your assessment. In your view, is it incompetence or is it deliberate non-compliance?

Is that in relation to the agents?

Yes.

For the record, over 70% of our wealthy population have agents. When you go up into the wealthiest end of that, it is over 90%. It varies according to the complexity that those people might have and whether they need advice. In terms of the range of behaviours, in a way, it maps quite closely to the range of behaviours that we see in customers. That will range from simple errors or mistakes where, as you pointed out, Chair, people might not understand or have got something wrong, through to obstructive behaviours during our inquiries. An agent might, for example, take a long time to reply to information requests, or may look to send us a lot more information that might not be relevant. That then goes right through to criminal behaviour, where we then look to criminally investigate them, if that is the behaviour that is presenting, i.e. false records, misrepresentation of a position, and telling a lie, in effect.

In your analysis of where you might minimise the risk, so setting out the risks that you identify, your interaction with the agents will be a really key component in being able to make sure that you prevent non-compliance in the first place, which, presumably, for HMRC is a much better and more effective approach than retrospectively going after individuals. In terms of your understanding of the agents as a whole, you have given us detail about some of your interaction. Is there a systematic approach from HMRC in interacting with what are a significant number of agents who support the income delivery for HMRC? What is your frequent engagement, understanding and, almost, profile of those agents?

You are right that we need to see it as a whole system, think about agent conduct and promoter powers, and look at the whole suite of opportunity to make improvement. We published four consultations in the spring statement, aimed at raising minimum standards of behaviour, tackling non-compliance, and contributing to closing the tax gap. There are four elements to that: making better use of data, enabling that early identification and getting it right first time; strengthening our ability to take action against tax advisers who facilitate non-compliance; measures to close in on promoters who market tax avoidance; and options to simplify and strengthen penalty powers. We absolutely recognise that, across those four domains, there is opportunity. We have done those consultations and we will set out the next steps. It is a fundamental part of us getting the UK tax system to work really well. When we think about things such as making tax digital, we are trying to nest the UK tax system increasingly in the natural systems of the economy, whether that be through software providers in the accountancy market, or the way in which people regularly record their receipts and expenses, and report their taxes. The more it is the way in which we do business, and the more it is preventive and early, the better, but, as Philippa and Penny say, for those who choose to obstruct or, ultimately, do not follow the rules, we have the powers to go to penalties.

Let me ask you about that. The majority of the population understand that, for wealthier individuals, their tax affairs are complicated. They recognise that they would have an agent or an accountant in the way that the majority of people, and certainly my constituents, are unable to pay for. They recognise that, for HMRC, it is cost-effective to have a risk-based approach so that you ensure that you are supporting individuals to pay the right tax. There will also be wealthy individuals who are simply trying to be non-compliant. In those cases, what is HMRC doing to be as clear as possible and to come down on them, saying, “Non-compliance is you not paying the tax that you should be paying”?

Your summary is spot on. That takes us back to the risk model, and identifying those cases and then taking the appropriate action. Penny can talk a bit about the hierarchy of steps that we would go through there, but, ultimately, we would end up at prosecution if that was the last resort.

If I could just come in there, JP, Philippa started to touch upon this in her response earlier, but the situation is quite a grave one. We know that London is a world-renowned hub for professional services, which means that, perhaps more than any other city on the planet, we have an army of lawyers, accountants, promoters of schemes, and tax experts and advisers. You, as HMRC, have a unique challenge, perhaps more so than any of your equivalent numbers in other countries in Europe and further afield. I appreciate that there is the consultation work that you have already underlined, but what are you doing to take on that unique challenge? Are you able to identify, profile and target new and emerging risks? You have talked about people who take a long time to respond or deliberately provide too much information. How are you looking at the next threat and the next challenge posed by this army of people who go about their day-to-day lives helping the very wealthiest minimise their tax liabilities?

I will let Penny come in and talk about how we try to make sure that we stay current and remain alert to new risks, because I agree that that is a really important part of managing the threat and the exposure. At the same time, to come back to my original answer, it is about making sure that the team have the capacity and the capability.

Do you think that you have? A lot of our constituents would think that HMRC is hopelessly outgunned and that there is, quite frankly, a very well-trained and very knowledgeable part of our professional services sector who go about their life making HMRC’s job pretty impossible. What we would like to see is reassurance that HMRC is up to the job. I do not think that any other body of your equivalent standing in Europe has quite the same challenge as you do here in the UK.

The NAO Report credits the team for the progress that it has made over recent years in more than doubling the yield from our wealthy team and managing to keep the tax gap at around 5%. To your point, we want to go further. We want to reduce that tax gap. My job, working with Ministers, is to make sure that the team has the capacity and capability that it needs. It will have more capacity. It will have more investment and more enablement to do more. Shall I just let Penny say a bit more about the strategy, the plan, and what we are doing?

I would just say that we are far from being outgunned. We have huge expertise in this space, and we constantly seek to build it, because we are not complacent, but you can see in the results that the NAO Report describes that we are improving the return on investment and that, in a limited number of cases that are in the public domain because of criminal prosecution, we can bring the most serious cases to successful conclusions in trials. That range of work that we describe rests on a huge depth and breadth of expertise in HMRC. Philippa has alluded to the fact that we are seeking to expand it by bringing in wealth management expertise to help inform our understanding of how wealthy individuals manage their affairs. We tackle some of the most complex cases in the country successfully, and you can look at our prosecution record to see that.

I appreciate that we have already talked about the consultation work that is under way. HMRC does have existing powers to go after and tackle tax advisers who are facilitating non-compliance, but you do not always use those existing powers. I think I am right in saying that there were no penalties levied on the enablers of offshore tax evasion in the last five years, despite landmark powers introduced by the previous Government in 2017. You have had that power since 2017, and it has not really ever been used to go after those who are making it their day-to-day jobs to enable people to evade tax using offshore tax havens and the like.

We use the powers that we think are most effective to tackle the problem in hand, and we constantly look to use the powers that will disable and bring the best effect to those challenges.

If the one that I just mentioned was used effectively, the penalty at the end of it could include either a £3,000 fine or 100% of the amount of tax that was dodged, whichever is larger. I imagine that it is normally going to be the latter; that is probably a lot bigger figure than £3,000. That is a tool that you have. Why has it never been used?

I cannot speak to that particular power today, but I am happy to give you a bit more information, if that would help. We use the powers at our disposal to apply to the cases that we have.

You do not use this one. That is one at your disposal, and you should be able to outline to this Committee why there have been no penalties levied on tax experts, tax advisers or lawyers, or whatever you want to call them.

Could you just confirm the power? We have levied other penalties.

This is on offshore.

Offshore is sometimes not the lead risk that might present. For example, we have prosecuted 41 enablers in the last few years.

Has it never been the lead risk? Will there be no instances when this power could be used?

There will be when it is the lead risk, but it might not always be a criminal investigation lead risk. We really do have to look at what the behaviours are, what evidence we have, because there is a high bar to use criminal powers, and then whether that is the most relevant tool in our armoury. We do arrest agents if we believe that they should be subject to arrest, and we have a strong record on that, but we can take away that particular power, because it depends on which ones you are referring to.

That is about evasion, so that is about criminal behaviour, not aggressive tax avoidance. You can understand why I and, I am sure, other members of this Committee would want to see those powers used by HMRC.

Lloyd, shall we just ask JP to let us have a note?

Yes, that would be really helpful.

Yes, of course.

We are not 100% clear about which power we are talking about, and a note would clarify that.

I am happy to do that.

Just touching on those who promote offshore tax avoidance schemes, there was a landmark action last month, which I certainly welcome, and you issued stop notices to Paul Baxendale-Walker for covering two offshore arrangements. That action that you took in that particular case was the first time that such action had been taken against an individual rather than just a company. My understanding, although I would appreciate if you could confirm this, is that he could face criminal prosecution if he does not comply. With that in mind, can you reassure us that these powers will not be used so sparingly in the future? This is only the first time that it has happened. Can you reassure us that you will continue to take robust action to stop those who are promoting or enabling aggressive tax avoidance schemes, particularly offshore ones?

Yes, we can. Like you, I welcome the progress that the team made with regards to the stop notice, and we will use them where we think it is the right thing to do and is necessary. To clarify your point, it does require those who are in receipt of the stop notice to stop promoting tax avoidance schemes, or potentially face penalties or criminal prosecution if they fail to comply. We have that power, we will use it, and we have done so recently.

Are you able to reassure us about how we can see it scaled up and used more frequently? This was only the first time that HMRC had ever used it.

It goes back to Philippa’s point about the nature of the risk. We are happy to reassure the Committee that we use the power when we think it is appropriate to do so, and we are happy to stay in touch with the Committee about the range of powers, which is the subject that is coming across here, the ones that we have used, whether property freeze, or disclosure and stop notice, where we are and are not using them, and how they are part of the suite of enablers that the team have. If the team say to us, “There are things that we need that we do not have”, our dialogue with our Minister is constant and regular in asking for the support that we want to give us the enablers that we need to close the tax gap. That is the top priority. We are very determined to do it, and we will use our powers where we think it is appropriate.

Just to put it in context, in terms of marketed avoidance promoters, our impact and use of our powers has had a significant reduction in the number that we see active in the market. We are talking about 20‑ish active promoters. We publish a list in terms of how many active promoters there are, and we also publish how many powers we use, including our stop notices. Over the years, our work on marketed avoidance, and particularly promoters, has had a dramatic effect on that part of the adviser population that you mention. Of course, as JP said, we have done the consultations recently to then think about where we can enhance our sanctions more broadly, particularly in respect of what you are talking about in terms of, say, offshore evasion and other areas.

Do you want to say something about your upstream and downstream activities? In recent years, your upstream activities have produced a significant yield. That includes where you are necessarily obtaining more powers. You are constantly looking at how to make the system work more efficiently. Do you just want to say something about that, JP?

I alluded earlier to the way in which HMRC operates in the economy and how we can increasingly make that a more regular business-as-usual process that nests with the natural systems of our economy. I mentioned making tax digital as an example, and Jonathan and Angela can say a bit more. Rather than reporting in arrears and the complexity of gathering together receipts, expenses and all the rest of it, it has worked well for that. We are mandating it for income tax for those with income above £50,000 from next year, and then we will come down the income threshold. That is about quarterly reporting that is up to date, enabling people to have accurate records, and reporting in a timely way. That will make a big contribution to upstream compliance and support a closure of the tax gap. The team can say a bit more about the small business tax gap, which is a large proportion of the overall. We have talked a lot today about the wealthy tax gap, but, ultimately, we want to enable people to get it right first time. We want them to have clear guidance and good communications, and we want to have good services that are quick and easy to use. The evaluation of MTD for VAT is pretty encouraging. We are hoping to have achieved the same for income tax with roll-out over the next few years as well. We increasingly want to make sure that the data and the systems that we have, our use of APIs, and the way in which we interact and intermediate with the tax system, are as easy as possible, so that people can comply first time. Jonathan, do you want to say anything else on the strategy and opportunities there?

Going back to some of the things that we have already talked about, having this data exchange infrastructure in place is also a strong upstream activity. Individuals know that we receive millions of records about offshore accounts. If we can get the infrastructure right, people will know that we have that information, and that will encourage them to comply. There are lots of ways in which that will play out for different parts of the market. With that information exchange and all the things that are coming down the track, in terms of, for example, crypto-asset reporting, people will start to realise that we will have that information, which will encourage them to make certain that they get their records right and are not careless or otherwise with their record keeping.

There is a tax gap estimate at the moment. In your view, will the steps that you are outlining and the proposals that are being considered at the moment improve that estimate?

Overall, the OBR has forecast a reduction in the current tax gap of around 4.8%, and is forecasting to get that to 4.4% at the end of the forecast period. That equates to an additional annual yield of £7.5 billion. We will publish the latest iteration of the tax gap shortly, and we continue to refine the measurement of that based on the samples that we do and the input data that we can bring to it. Ultimately, our objective is to reduce the tax gap overall, and the wealthy tax gap and offshore tax gap are a part of that. They are currently estimated to be broadly similar, at 5%. With all the work that we have discussed today, one would hope to see that also reduce over the next few years.

Does that include the wealthy figure too?

Yes.

Is there any particular work that focuses on the tax gap for the wealthiest individuals as opposed to the overall figure?

I come back to some of the things that the team have spoken about today. More capacity to do more, more enablement to be more productive and to target according to the better data that we can make available to the team to focus the work, specialist capabilities, and some of the policy changes, also help.

What is your understanding of what that figure is at the moment? How confident are you about that figure? It does seem low.

Overall, the tax gap is something that is very difficult.

Sorry, just to be clear, I should have said that the figure that I am referencing is the £1.9 billion figure. When we consider the numbers from HMRC and we talk about wealthy individuals, a £1.9 billion figure is significant for many people, but, in terms of the overall numbers, seems rather low. How confident are you in that £1.9 billion figure?

I will stand back and then get on to the wealthy. Overall, the tax gap is a difficult thing to estimate. “What are you not collecting that you should?” is a difficult concept to put into practice. We are one of the few countries that publish a comprehensive estimate of the tax gap. Different elements of that have different strengths. Around 75% of that overall figure, not just the wealthy, is based on some very robust statistical methods. As you get into the more niche areas, it becomes difficult, because, as we were talking about with some of the wealthiest, we do not have external data that we can validate our receipts against, so it is harder to do that. That is where we rely on looking at particular case risks. On the wealthy, we work very closely with Philippa’s team on what trends they are seeing and, from that, how we can understand how those risks might manifest themselves and how much might be at stake. Within Philippa’s team, we have deep experts in terms of understanding some of those wealthy risks. Some areas are going to be easier to do than the others, because, as we have talked about before, within the wealthy group, there are a lot of different behaviours and categories of individuals, so that is going to be a more difficult part of the tax gap to estimate.

Having confidence in a figure is important, because it gives the organisation a sense of whether it is successful in using the tools that you have at your disposal and in targeting a group of individuals who will generate significant income for HMRC. I just want to press again on that figure. If you are not confident about it, and if it is simply impossible to ever know what that figure should be, that is really important.

This is our best estimate, but this is a difficult part of the tax gap to estimate.

It might be helpful if you could provide a bit of detail about how HMRC has reached the £1.9 billion figure for the tax gap for the very wealthiest. Does it simply include routine non-compliance, or does it cover deliberate and aggressive tax avoidance? Is it based only on savings and income, or does it include other types of income such as a dividend?

It will aim to be as comprehensive as possible, and to look across all tax types and all sorts of behaviours. One thing that we see in this area is what we call legal interpretation, so people pushing the boundaries of tax or trying to exploit where there is some ambiguity in how tax law should be applied. It is aimed to be comprehensive and to look at all the risks. As I said, this is a whole organisational effort, drawing on the frontline expertise of Philippa’s team.

On a point of clarification, it is personal taxes that this focuses on largely, so income tax and capital gains tax, and components of inheritance tax and trusts that we talked about earlier.

What does it exclude, then?

It would not include, for example, the businesses that wealthy people own, because that is rightly recorded in our business tax gaps.

Would it include dividends?

That would be included in the income tax part.

Around this idea of your estimate of the wealthy tax gap, it does seem that it is not clear. I felt, reading the NAO Report, that it was not clear, so I hope that you can provide some clarity today. The wealthy tax gap has remained under £2 billion for a while now, but the number of wealthy individuals has grown quite significantly, from 700,000 in 2019-20 to 850,000 in 2023-24. At the same time, the increased compliant yield shows that you have been able to find £3 billion more in tax that was not paid by very wealthy individuals. If I understand this correctly, the total estimated tax owed is smaller than the additional tax recovered through compliance, but, at the same time, the net loss in tax revenue has increased by 21%. I would really appreciate some clarity, because it just is not clear. How can the wealthy tax gap remain stable, the compliance yield increase and the net loss also increase at the same time?

There are a few things in there. Let me try to unpack them in order. First, while the number of people captured by the wealthy definition has increased, the increase in the total income within that group has broadly kept pace with the tax gap. The net tax gap is broadly around 5%, and it has been consistently around that level. That is why it goes up to £1.9 billion, but it is broadly the same as a percentage of liabilities. In theory—and I will be very clear about why I am saying “in theory”—what you can do is to add our compliance yield to our net tax gap to get a gross tax gap. That gives you a sense as to, “If we had done nothing, what would the tax gap have been?” That happens in theory, but, in practice, the way that we measure these two things is not always additive. There will sometimes be timing differences around particular payments, which means that, even if we record compliance in one year, it may relate to different years of activity. Broadly speaking, what you are seeing is a picture of, “If we had done nothing, non-compliance would have increased”. What we are doing through our activity is keeping the tax gap at roughly a constant percentage of receipts, and that gives you your £1.9 billion. There are some challenges in there, but that is broadly the picture. We would be a little bit careful, and we do have some guidance on our website to try to discourage people from doing that too much, adding together the compliance yield and the net tax gap, because there are sometimes timing differences in when things are recorded. Some of that compliance yield is not extra money that we have brought in, but is protecting money we already have that we otherwise would not have collected. Does that help in trying to unpack how that works?

It helps a little bit. It would be useful just to move on, if we could—and we are probably going to touch on this at some point later in the session anyway—to look at compliance in a bit more detail.

We have already spoken a bit about wealthy compliance cases, but I would like to just come back to it. The report also talks about the process for being able to deal with those cases, which, by their very nature, are complicated. Cases may take a lot longer because of due process and lengthy correspondence with the taxpayer, agents, and so on. Philippa, you already detailed some of that, as have other panel members. The average time that it took for HMRC to close a case increased each year between 2018 and 2023. It peaked at about 20 months before it then fell again. For investigations yielding more than £100,000, the average time that it took in 2023-24 was 40 months, which is a significant increase from 27 months in 2022. For an ordinary person—a teacher or a nurse—in my constituency who is earning far less than some of the wealthy individuals we have been talking about, and who is taxed before their salary even hits their account, it is very difficult to stomach the fact that HMRC may take 40 months to investigate a compliance case. Even more to the point, that figure has got worse over the last few years. I wondered if you might want to comment on that.

I will bring Penny in. We were having this conversation this morning. The one data point that is important to put alongside the increase in time is the average value that the team is now collecting. Penny will have the exact numbers, but, as I recall, it has broadly doubled, and it is because we are increasingly doing what we described earlier, which is supporting more people to get it right first time upstream, then targeting our resources on higher-risk, higher-yield cases and bringing more in. It does take longer, as you say, and that is always the productivity conundrum. You could do low-value, quick stuff and get a low-value return, or you can take on the more complex cases and recognise that they will take longer, but, when they crystallise, the yield is higher. Penny, can you give us the numbers?

I am sorry to interrupt, because I am really keen to hear your response to that, but does it not concern you that almost half of the compliance cases close with no yield? From what you have said, John-Paul, I would suggest that you are saying, “That does not matter, as long as we are getting high yield”.

Let me talk first to the yield and the return. We have been focusing our teams on more complex casework, and that has been seen in the average return that we are getting from those cases in the wealthy area, which has tripled, broadly speaking, from £34,000 to £94,000 per case on average, with a wide range within that. I should say that we have a constant focus on the timeliness of our cases. It is not something that we just regard as a trade-off for complexity. We are constantly looking at ways in which we can improve the pace of our inquiries. We recognise that there can be delays at the other end of our inquiries, but there is a constant focus in the team on how we reduce the length of time that it takes to conclude, because we want to put our resources into the most productive work and do that in the most effective way. That said, there is sometimes great complexity in some of these investigations. In terms of cases closed for nil, that can be for a number of reasons. It can be that we have opened an investigation and simply find that the affairs were in order. Where we do that, that is closed, we hope, reasonably quickly. We are doing our work in a risk-based way, but we sometimes do not know the facts until we open an inquiry. Philippa, I do not know whether you want to add anything on the nil point.

On nils, we recognise that that is a number that we would like to reduce. That is what data helps us to do, so that we can better target it. We have talked a bit about the investment that we want to make. As Penny says, we will start with a global estimate of what a risk might be, but, when we then get the evidence, or, in fact, someone can supply us with the contracts of share sales or, potentially, the trust documents in terms of what they invested, or can demonstrate that they are not a UK resident, we would look to rightly close that for nil. It will depend on the nature of the risk that we open into those individuals, but all the investment that we have talked about today should help us to continually better target that.

I just want to be clear, and I am not making a value statement here. It does not concern you, as an organisation, that almost half of the compliance investigations close in no yield. That is because, for instance, it might be that you have investigated and there is no reason. I am just trying to get a sense of whether that is okay.

The opportunity is to reduce that. We absolutely would like to reduce that. We do not want to open investigations into cases where, ultimately, it was sunk effort, if you like. At the same time, the team needs to be empowered to investigate where there is a concern. Sometimes, they will conclude that that concern is not real—it might be a whistleblower or whatever—and that it is all lawful and everything should be left alone.

Where I would not want you to end up is not opening cases at all.

Yes, indeed.

I want to be clear about that.

That is right. That is the risk with the data. You want the team to be empowered to operate in a risk-based way, but you also want them to be enabled to risk and target in a better way, so that, when we open a case, it has a higher likelihood that we will get the yield. On the one hand, of course, we would like them to crystallise faster and get the yield in more quickly. That would give us more confidence in terms of reporting performance against our targets. Having said that, coming back to the previous point, some of the most complex cases take a long time. In my first few weeks here, I have stepped through a number that are live now. The teams have been working on some of these cases for months and months, and it is good that they do not give up on that, because, if they do, that deterrent effect, when we do get those cases to court, also really matters.

I want to come back to the length of time, because there are some risks associated with the length of time being so long. If I could quickly talk about the targeting of investigations to cases where non-compliance is happening, has that improved as an organisation? Can you point to one or two things that the organisation has done in recent times, where you can say, “We have got better at targeting our efforts and results are better at the end”?

Yes, absolutely. Our average case holding, as Penny said, has tripled in that time. We have gone from over £30,000 to over £90,000 per case.

Why? What has happened? Is it the resourcing?

It is better data, as well as deployment of our most senior tax professionals to the risk. In terms of better data, it is particularly that that we receive from offshore now. Previously, we did not know about people’s financial accounts offshore, unless we already thought that there was a risk and then requested it. We now get all of that automatically, and are able to then use that in our risking systems to help us pinpoint and target which ones we should open.

That is really helpful. I want to ask about the length of time, but is there something that you wanted to come in with before then, Lloyd?

What is perhaps quite worrying, looking at this NAO Report and other reports, is that the yield is very low. Could we get some clear reasons for why 46% of investigations still close with no yield at all?

Ultimately, how it would work in practice, if I give you a bit of an example, is that we might identify a risk with a specific individual, where we will then go through a sometimes lengthy to and fro to get the information that we require, for them to either explain the position or for us to determine the tax outcome. Our job is to make sure that the right amount of tax is paid. If that particular individual is able to provide evidence that, in fact, they have paid the right amount of tax, we would close that investigation for nil. You then will have other types of projects that we might do, where we see a risk in a particular sector. Particularly having changed our capital gains tax reporting requirements recently, we might, therefore, do a project that looks at whether people have reported on time, in line with the new legislation. That might then open inquiries into 500 wealthy individuals who we believe did not report on time. Of those 500, there will then be a percentage who, in fact, are able to demonstrate that they did, or that there was a reason that meant that they were unable to. You get a real range of what might sit beneath the nil number, ranging from singular inquiries through to project work.

On the point about the yield being low, that is not how I recognise it in my first few months here. The NAO—and I hope that I am saying this accurately, given the report—credits the improvement in the yield from this team. In terms of productivity, we have talked about the tripling per case. If I look at it overall in terms of yield targets that we agree with the Treasury and with Ministers, we are on track again to exceed that. That is what Penny and the team are resourced for, and they are delivering ahead of their targets. Overall, I am content that the way in which compliance is being led and delivered in HMRC is improving.

It is from quite a low starting point, granted.

We accept that we want to go further.

There is a long way to go.

Ministers are clear that we want to close the tax gap. That is the top priority. We are resourced to do so. The OBR is scoring that on the assumption of our plans, and we are looking forward to delivering it, but it does build on some strong foundations. When I look at the progress recently on recruitment, the tax compliance academy recent yield target performance gives me confidence that it is feasible.

The process of investigating has grown in how long it takes, from 27 months to 40 months. You have both said that these take months and months, and that it is a lengthy process of back and forth.

It has for the most wealthy ones. The overall average time has reduced.

Why has it grown? If you were doing investigations and they were taking 27 months, why is it now taking 40 months?

The average can be skewed, depending on which cases close that year. If you get a year, which we did have, where particularly lengthy, long-running inquiries closed, that can sometimes skew it. In general, it will come down to, as I talked about earlier, how willing the customer and their agent are to share the information that we need and whether they obstruct that. Within the due rights that taxpayers have, they are allowed to appeal our information notices. They are allowed to appeal that to tribunal and to say that they are unwilling or do not think that we are targeting that appropriately. Those are all taxpayer rights that are available to everybody if they do not agree with our inquiry notices.

Are some of the individuals being deliberately obstructive, where it is not about appeals or using the proper processes that are available to them? Are some of these individuals under investigation deliberately slowing down the process?

Yes, absolutely.

Do you not worry, then, that, if it takes 40 months on average to complete an investigation, there is a serious risk associated with flight? These people just get out of the country, and it is then impossible for HMRC or, indeed, any other body or agency to track them down.

Yes, there could be a potential flight risk. That is not something that we see a lot in our largest cases.

It does happen. Because you take so long to investigate, people can get out of the country.

I do not have any examples of that from the work that I have seen.

But you do think that it happens.

We have an arrangement called the J5, which is the joint global chiefs of tax enforcement. We work very closely with international partners, so that, if anybody was to try to leave the UK, we would absolutely pursue them, using all the tools that we have, but mainly working with our jurisdictional partners. That is a really important element of the work that we do to make sure that people cannot escape the amounts that they owe. Is it harder to collect debt when someone is overseas? Yes.

If you take the other end, you are assuming that individuals are not trying to just flee the country. We have also been told by professional bodies that, for lots of wealthy individuals and their agents, the length of time that it is taking for their cases to be resolved is highly frustrating for them. It is hindering their ability to get on with their own accounts and their business. There is an impact on them as individuals, and often as part of organisations, of stress, anxiety, and so on. I take the point that it can vary in any particular year, depending on the cases, what that average wait time might be. Does HMRC have a view of what that figure should be? Should it take 27 months for a case to be concluded? Should it be 40 weeks? Is there an aspiration for a period of time that is much shorter? What should the professional body’s expectations be for these complex cases?

Their expectation should be that we work in line with our compliance professional standards. My ambition would always be to reduce that to the appropriate amount of time to get to the right amount of tax being paid. I do not have a specific target, but, quite like what Penny said earlier, we look in the round at a number of those indicators. It is important to me to make sure that we are working in a timely and effective manner, and I review those statistics with my team on a regular basis.

Hypothetically, if the waiting time suddenly shoots up to 100 weeks next year, how, as a Member of Parliament, will I know if I should go to the House of Commons and say to the Minister, “This is unacceptable”?

The first thing to reassure you of is that we would consider it unacceptable as well. To Philippa’s point around timeliness and customer service, and your point around the anxiety and the concern, that matters to us. We want to respond to that and close these cases out as efficiently as we can, but we also have a legal duty to ensure that the right tax is paid. This indicator is an important one. Where it is feasible for us to reduce it, particularly with the improvement of enablers, which might be support for the team in terms of productivity and the way that they are working, we want to give them all the support that they need. At the same time, going back to some of the challenges earlier about the most complex cases, with the best professional advisers and the context that we are operating in, we also want the team to be empowered to take on those cases, recognising that they will take longer. If you are looking for some assurance around us reporting that indicator, so that you, as Parliament and as a Committee, can hold us to account in that regard, we are very happy with that. There is a recommendation from the NAO in its report in terms of transparency of our compliance regime, our strategy, our approach. Is it improving and are we on track? We will, of course, account for that every year in our annual report. I will, no doubt, come to the Committee in the autumn and talk about that. I will bring this crew with me, and we will answer those questions and ensure that that data is transparent for you, because we want, the same as you do, the tax paid as efficiently and in the least disruptive way as possible. It is important that the team takes on those big cases and recognises that they will sometimes take longer.

Is there a relationship between the length of time that it takes to complete these cases and the yield? In your assessment or your understanding, is there a trend that can be identified, or is that simply just not the case?

There are multiple factors at work. One is complexity. We have sought to improve the productivity in this space, which has included taking on more complex cases. Our case mix changes every year, because we are looking at risk and at the most effective way to tackle it. It is only one factor, I would say, and there will be multiple different types of work going on in the wealthy team at any given point in time. Philippa has given you an example of project work involving lengthy one-to-one inquiries that can be international. It varies a great deal, so it would be wrong to try to oversimplify that number.

I have some questions to try to get clarity around this. I have been listening carefully to my colleagues’ questions, and it seems to me that they are on two strands: the number of cases with no result, and the length of time to investigate. Could I just ask some questions around both those aspects? On the number of cases with no result, it does seem to me that one of the things that you need to do is to improve your risk analysis. How do you do that? This is a real question about the way that HMRC works. JP, you have made a great play about employing these extra experienced tax compliance officers, but I wonder whether you ought to be investing more money in digital, so that you can use AI to target this huge amount of information that you have been talking about getting, particularly from foreign people with non-dom status. This is precisely where AI can look at large amounts of data to see where the trend is off track. I wonder whether you are investing enough in AI.

We would agree exactly with what you have just said. It is about more capacity, but also improved targeting and risking, and optimising the use of digital capabilities to support us in doing that. That will lead to improved productivity. One of the indicators that we might, therefore, see shift would be a reduction in those nil cases. However, I do not want the team to overly react to that, because you might then create a risk aversion, meaning they do not take on cases where there is uncertainty.

You were saying that to Nesil.

I want them to be empowered to investigate it, even if it means that it might result in a nil outcome, because that is the right thing to do. Yes, absolutely, it is about the use of AI, predictive analytics, and the way in which we use our risking models and our data. The team can say a bit more about all the work that they have already done, but a big part of our SR settlement, as Angela referenced earlier, is about the productivity of our compliance engine and giving it the investment that it needs. We are standing that up now. There is work building on proof of concepts and work already. It is a big opportunity in the years ahead.

The second question follows the first, and is on the length of time to investigate. I entirely agree with Nesil that the longer it takes to investigate, the greater the risk is that people will disappear one way or another. Again, AI may give you a better steer than your tax compliance officers, because it will tell you the information that you are missing. It may enable you to use your letters or emails of inquiry to better target the information that you want, even to the extent of alternatives—“If this is true, we want this information; if that is true, we want that information”—which would shorten the exchange of information and, therefore, the length of time taken to investigate.

Let me just bring Penny in. We were having exactly this conversation this morning. We agree that the opportunities from AI are significant. The team has been working with risking models, which are augmented with digital capabilities, for a long time, so there is a foundation and a capability to build on.

Over the next few years, we are looking to invest in our risking and our data, because the risking only works off the data. The digitisation of the inheritance tax regime, which was announced in the autumn Budget, is going to be really helpful, because having that data readily and swiftly available to us improves the pace at which we can feed it into our risking and analytics. Investment in our case management systems will enable our wealthy case workers to get access to the data and information in a more timely and effective way. We want to look at the use of what people think of as AI—the generative-type AI—as well in terms of helping us to maximise our impact. We use a lot of the broader spectrum of AI at the moment, such as predictive analytics, machine learning and so on, in our risking functions, but our absolute intention over the next few years is to invest and to see a return from that.

That is very helpful. Thank you. Can I just come back to some of the basic questions? I agree with Nesil Caliskan that, on the face of it, as seen in figure 10, the wealthy tax gap has hardly shifted. In fact, if anything, it has decreased from the £1.9 billion figure that we have been talking about in this hearing to £1.7 billion, yet the compliance yield has gone up from £2.6 billion to £4.4 billion. I very much accept, Jonathan, as you were saying, that, theoretically, one has an action on the other, but I just wonder whether the basic wealthy tax gap is correctly calculated in the first place. This may be too simplistic, I accept, but I would welcome a view on what I am about to say. Paragraph 13 makes it perfectly clear that there was £849 billion in offshore accounts at the end of 2019. I assume that that probably will have gone up by now, but, even if you take a 4% yield on that, that is £33.6 billion. The income tax liabilities for 2023-24 just so happen to be almost identical, at £34 billion. That is without all the capital gains, stamp duty and everything else. On the face of it, that very simplistic analysis would suggest that the wealthy tax gap is far too low.

We will probably come on to this, but, when we have looked at the offshore tax gap, we are able to match individual records that we get in against the information that we hold. That suggested that the level of non-compliance offshore in regard to the information that we get about those accounts is quite low. We were able to match a large proportion of those accounts to correctly filed tax returns. That top-level analysis will potentially be misleading. You need to look at the individual levels. For the wealthy population as a whole, we will then look at individual taxes or behaviours and understand what the risks with those are. That is probably overly simplistic in terms of that, but, as I said, I would also be very clear that this is an area that is very difficult. Tax gaps in general are difficult to measure. Tax gaps among the wealthy are even a bit more difficult than the average, so there is a challenge there. If I am looking at offshore, where we get that information in from overseas and reconcile that against our own records, we reassure ourselves that the vast majority of the income related to that is collected. That is only the case when we get that overseas information on financial assets.

If you are going to defend that £1.7 billion figure for the wealthy tax gap, as opposed to the £4.4 billion figure for compliance, you have to link the two. It does suggest, therefore, as the report suggests, that there is an increased level of non-compliance in the system.

As I said, we need to be a bit careful about adding the yield and the tax gap every year, because there are sometimes timing differences between them. In general, it is indicating that the potential for non-compliance here is higher, but we are more successful at collecting more of that potential non-compliance, hence the yield is going up.

The only thing that I would add comes back to one of Jonathan’s earlier points. In the way in which the tax gap is calculated, some of our work does not count towards it. If somebody makes a claim for repayment, for example, but we reject that because we do not think repayment is due, that will be preventive for future amounts rather than tax gap. I know that that ends up sounding rather hypothetical and analytical, but it just means that it is sometimes not comparative of the compliance yield in that year against the tax gap in that year. However, we continue to want to work with our analysts. As we increase our capacity and find out more about those individuals, that then directly feeds into the level of risk that we ascertain. It is a continuous process in terms of our understanding.

Putting together what both of you have said, there is a theoretical possibility that compliance is getting worse, but you are collecting more tax. Is one of the reasons that the cases are getting more complex, it is taking longer to investigate because they are more complex, and, therefore, you are rolling over money from one tax year to another, leading to a lag in these figures?

As I said, there are some challenges, particularly with very large or complex investigations. They may cover a number of years, but they may conclude in one particular year. We see this also with our large business population, which has some similarities in having a small number of people who pay large amounts of tax. There is a risk here. As I said, it would be more that the potential for non-compliance is higher, but what we are doing through the work across the organisation is maintaining a net tax gap, which is down to better data and also the hard work of our colleagues on the front line.

We had a good conversation with the NAO about this, because we recognise, as Jonathan has said, that the tax gap is an estimate. We are continually iterating the methodology. I have been through it with the team. What the UK does in regard to the tax gap is global best practice, but that does not mean that we should not continue to improve it, and we take that challenge. Our analysts are excellent and are up for that challenge, so we will continue to iterate it. We will publish the latest iteration of the tax gap shortly. It will continue to move around as the baseline moves around. Our objective, of course, is to see it reduce in the long term.

That assurance is very welcome, if I may say so.

I just wanted to make sure that we did not forget to talk about the offshore tax gap, although we have started to touch upon it. We have talked about the overall tax gap, and the tax gap for the wealthiest individuals. When your predecessor came to this Committee not so long ago, we had quite a back‑and‑forth about this. I would appreciate an update from you, JP, just so that we can get an idea about what HMRC’s current thinking is. The current estimate that was shared by HMRC to this Committee at the last meeting was that the offshore tax gap was just £0.3 billion in 2018-19. That was the most up-to-date figure that it could provide. This seems like quite an implausible figure, particularly when we consider that HMRC’s own common reporting standard outlined that UK taxpayers held £850 billion in foreign accounts in that same year, of which £570 billion was held in tax havens. Of course, there are also stories that hit the press. The infamous Roman Abramovich tax scandal that came up not so long ago suggested that he might have avoided, allegedly, around £0.5 billion in tax, which, alone, dwarfs the figure that you have given us. Since our last meeting, could you give us any reassurance that your estimates as to what the offshore tax gap is are in any way accurate?

I am going to ask Jonathan to cover the analytical side, because his team leads this. Having the previous interim national statistician on your team is a helpful thing.

Just to clarify—and apologies—I may have fallen down the same trap. What we did with that particular estimate of £0.3 billion was the non-compliance relating to automatic exchange of information. This is where we receive millions of records. For the year in which we did this evaluation, we received information on 7 million accounts relating to 4 million people who were resident in the UK. What we did there was an exercise of matching what we had been told from those overseas accounts with what we had on our own records. Where there was a discrepancy, we then did a random inquiry process where we took a sample of 400 of those cases, and worked them through to see how much was not compliant. We added that all together and, for those cases where there is automatic exchange of information for that part of assets held offshore, that tax gap was £0.3 billion.

Just to be really clear, do you feel that the £0.3 billion figure that your predecessor, JP, brought to this Committee is, in any way, a robust estimate of the offshore tax gap, particularly considering that HMRC’s own common reporting standard points out that UK taxpayers hold £850 billion in foreign accounts? I am not saying that that is all part of tax avoidance or evasion—I am not making that statement at all—but you would like to think that it is probably more than £0.3 billion.

When we worked that through, we reconciled the information about those offshore accounts with the information that we held. There are some assumptions in there, and we have been very transparent about how we have gone through that method. By the time we did all of that, there was only £0.3 billion left.

It is £570 billion held in tax havens. These jurisdictions exist completely to enable people to aggressively avoid and evade tax. If it is a £570 billion figure of just UK taxpayers, can you really tell us with confidence that that £0.3 billion figure is an accurate estimate of the offshore tax gap?

Yes, of that particular part of the tax gap related to the common reporting standard and the US equivalent. As I said, individuals will know that, if you hold information offshore, we get that information.

Is this a very effective formula? Is it world leading? Is it robust?

We are the only country to have done that reconciliation, as far as I know.

That does not mean, though, that it is a good formula.

You might be the only ones, but it does not mean that it is effective.

We have always been clear that it is a partial estimate. It is not the totality of the offshore tax gap; it is just a partial estimate.

So you are accepting that it is partial.

That is what we said in our publication. It is a partial estimate.

It relates only to the common reporting standard, and this reconciles all of that offshore information with what we hold. The tax havens are telling us this information. They are telling us, account by account, individual by individual, how much is held, and that allows us to reconcile it. As I said, that will have an effect on the behaviour of the individuals. Those individuals will know that simply opening an account in one of those havens is no way of hiding your money. We would expect that sort of behaviour, because these people will know that that is no longer—

Surely, you can understand our concern. Sitting alongside what the common reporting standard sets out, we have seen a litany of offshore tax leaks, such as the Panama papers and the Pandora papers, which expose the scale of tax evasion and aggressive tax avoidance. For instance, the Panama papers pointed out that 100 UK-resident millionaires were identified as tax dodgers. Of course, those are allegations in leaks, but that is what those leaks set out. It made clear they were hiding their wealth in offshore schemes based in tax havens. With that in mind, given those leaks, the Abramovich story, and what the common reporting standard seems to suggest, do you really feel that HMRC has the skills—and this is probably one for you, JP, as incoming chief exec—to have any idea of the scale of the problem and to give us, as a Committee, and the British public an estimate of what the offshore tax gap is?

We have tried to set out today all the things that we are going to do to increase the capacity and capability of the team in order to support them in undertaking compliance activity for wealthy customers, including offshore. That is an important part of increasing yields, which will increase year on year, and closing the tax gap. We do not disagree with your underlying objective, which is, ultimately, to ensure that taxes that are due get paid. That is our objective.

My objective rests on you having an accurate figure for what the offshore tax gap is.

That is understood. I have had that conversation with the team a lot. They have always tried to be transparent that we have been trying to measure a very complex thing, which only a few other jurisdictions around the world do. We work with the OECD to try to make sure that our work is grounded in best practice. We will keep doing that, and being clear that it is a partial estimate based on the data that we do have available. We want to expand the coverage of data available, which will include the confidence of the measure as well. My reassurance to the Committee is that I am very happy to continue to come back here and talk about how we are refining the methodology of the tax gap, assuring it as best practice globally, and then risking and targeting additional capacity to bring that tax in, wherever it is around the world.

Since your predecessor last came here, the Government have agreed with the PAC’s recommendation that HMRC should develop a stand-alone strategy to reduce the offshore tax gap, so government Ministers also understand the scale of the problem here. With that in mind, can you illuminate the Committee today as to when you will develop this stand-alone strategy to reduce the offshore tax gap?

We have accepted that recommendation, and we will set out in writing for the Committee our next steps in regard to that. We have covered quite a lot of the nature of the strategy today in terms of the announcements that were made in the autumn Budget and the spring statements, the consultations that are under way in terms of powers and additional capacity, and the new capabilities announcement yesterday in the spending review to invest in our digital capabilities.

Will there be a publication date for the strategy to be put in one place?

If the Committee would welcome that, we can draw that all together and then make sure that we have played it back to you in writing. We will need to agree with our Ministers what they would like to do in terms of the mode of that publication, whether it is fiscal events, part of the annual report, or a stand-alone thing. In terms of what the strategy is, what the plan is and what it will deliver, we want to be as transparent and as clear as possible on that.

If the Committee would like to see how recently we have set some of this out, we published something at the autumn Budget 2024 that set out our current approach to tackling offshore evasion, which is always really worthy of what we are doing right now.

That would be very helpful. Thank you.

I have constituents in Barking who are living in properties and paying housing charges that go to a company owned by individuals living in offshore tax havens. They are getting ripped off through these charges, so they really want HMRC to go after individuals who should be paying their fair share of tax, which is why it speaks to the fairness that every individual in this country wants to feel, and why that figure is so important, because it gives people confidence. I just want to know that panel members understand.

We feel that keenly. Fairness is right at the heart of our charter. The vast majority of people pay their taxes on time lawfully. We want to make that as easy as possible and to improve our customer service. Angela set out how we will do that. Our compliance team does a great job in bringing in additional yield over and above that, and is exceeding its targets. We will be stretching those targets further and, ultimately, raising an additional £7.5 billion a year by the end of the period because of all of that. At the same time, we want to close the residual tax gap to be as low as possible. We are already forecasting to see it reduce, as is the OBR, based on our announced plans, and we are working flat out with our Ministers on options to go further. Ministers will say more about that at the autumn Budget.

Well done, Nesil, for bringing that to the fore, because it is about fairness. I do not think that anybody could disagree with that.

The final policy decision would always be a matter for Ministers. The self-assessment return is one of the ways in which you can collect that information, but we have talked through a number of other ways in which we collect data that have enhanced our understanding and ability to risk the wealthiest individuals, such as some of the international agreements that we have discussed today. The removal of the non-dom regime and the need to then report foreign income and gains will, for the first time, give us enhanced information that we did not have before on those individuals. As I mentioned, we are getting crypto-asset reporting soon, as well as the register of overseas entities. That combination of data does put us in a significantly enhanced position from 2017, when this was last discussed at this Committee.

In 2027.

We will discuss that policy with Ministers. It would always be part of our thinking in terms of a range of options that we might deploy to tackle this risk. What I would say, though, is that it has significantly shifted in that time. You mentioned trustees. That is now part of the common reporting standard and the register of overseas entities. In fact, we are making improvements to both of those to think about things such as joint accounts and also to get more information on controlling persons. That is already in hand. There is a mixture of ways in which you can look at this. What people tell us is one of them, but what we have found really effective is that automatic exchange of information and getting that right, because, of course, what people tell us is more of a matter of choice, whereas we can rely factually on what we get automatically from the jurisdictions that send it.

I would not use that phrase. What I would say is that we do get a lot of information and data, but the risking and analytics that we were discussing previously are so important, because data alone will not help us pinpoint an individual. If someone is determined to set up complex trusts or structures, we then need to follow that through. We may get, for example, information that there is a limited company based in the UK. We then need to work with Companies House to identify who owns that company, and where the income and gains from that company actually flow through to. That is why investigative work alongside risking and analytics is so important. In terms of gaps, as I mentioned, we are working to improve the information we get on controlling persons, so that we can more easily identify the beneficial owners. That is with the OECD at the moment. We are also looking at bringing in e-money. I mentioned crypto, but it is e‑money in general. Joint accounts has been another area where we did not get as much information as we would like, so we are bringing that in. The OECD is also thinking about real estate data, because property transactions is another area we would like to know more about. Ultimately, it is an evolving picture. We would love to see more transactional data, for example, because we get account balances under the CRS, but not transactions. That is something that we need to agree internationally, because us doing it unilaterally will not get us the information that we need.

There are two angles to this. Certainly within the UK issue, you mentioned bank accounts. We have a consultation outstanding on improving the quality of data we get around UK bank accounts to address that. As you say, we struggle to reconcile one-fifth of the accounts. That consultation is ongoing. In the same consultation we are also talking about dividend income, because, again, we would like this. If we can get these right it helps compliance, but it helps customers as well, because if we can get that information automatically and put it into their tax affairs they do not need to tell us. Actually, there are some real benefits from that domestic exchange. Those will have really clear deadlines. Obviously, we need to consult. The banks and the financial institutions will obviously have a view on how far and how fast we can move in that regard. Philippa also alighted on where we are working internationally to elaborate the common reporting standard. That is great. That has been a real game changer, but where do we make that more sophisticated? We have a deadline for crypto-assets reporting, but some of those other things will require multilateral negotiations. That is not necessarily solely within our control, but the direction of travel and what we are looking to do is fill in exactly those gaps that Philippa was saying. There are a number of aspects to data; some of it is within our own control, and therefore we can set more of the timetable, but I would caution that some of this will require multilateral negotiation. That can take some time.

Just on your point around transparency and accountability, I totally agree, and we will roll that into that previous action. We can share where we have live consultations, commitments and dates, so you are clear and you can hold us to account on all of that. Just to reinforce your point, the appetite is very high, and the encouragement and support for Ministers to pursue this, fill those gaps in and optimise the enabling environment to improve our performance is good. Where there are ideas and opportunities, we want to take them as quickly as possible, because we want to close the tax gap.

I have a few factual mopping-up questions; they are short questions, because we have been going a long time. Paragraph 3.9 outlines your recruitment of extra staff. That paragraph tells us you aim to recruit 5,500 more compliance staff, but HMRC has not yet determined how many of these will be assigned to the wealthy team. Further down, it says, “HMRC plans to increase the number of people tackling wealthy offshore non-compliance by around 400 people over the next five years”. Of that 5,500, it is 400 going to the wealthy team.

Yes. We have recruited over 500 already. We are on track, but it is a mountain to climb. I have done this before in Jobcentre Plus; we brought in 13,500 work coaches in six months. You can do this, but it is a big undertaking. So far, so good.

To deal with these more complex cases, you are going to need more better-trained people.

Yes.

Are you training them up yourself, or are you aiming to recruit more experienced people?

Our plan for recruitment involves taking roughly an additional 1,000 people a year, broadly split between our compliance foundation training and our tax specialist programme. We take around 280 tax specialist graduates every year. We are upping that to around 500. The remainder will be in our foundation-level programmes, but as part of that recruitment we are also looking to bring in more external experienced tax specialists. We brought in over 40 last year. We will continue to search for the people with the right experience and expertise to help teams like Philippa’s, or our large business colleagues. However, we do know that we need experience in the wealthy teams. What we will be doing is rolling experienced people into those teams to support the work that Philippa does, while we are bringing new people in and training them. We bring them over several years because they have an intensive classroom period of learning of around nine months, and then they are out in operational teams being coached and mentored. That is quite an intensive process, so we cannot bring them all in all at once without cannibalising the normal compliance work that we would be doing to an extent that is not helpful. It is a graduated process of recruitment.

Is your recruitment package attractive enough to be able to recruit skilled people from the private sector?

Certainly, we have been impressed. We have started looking for those external experienced people, and so far we have exceeded our expectations. We continue to look at what will attract people, and experiment with new ways of attracting, recruiting and retaining people, but so far it is working.

When will this Committee be able to see the results in terms of increased compliance from those people you have recruited?

For the people we recruit at foundation and tax specialist level, we have an assumption-based model, which is how the OBR scores the additional yield. That will come in over the next few years. The summary numbers on that are in the Red Book. As regards to the people who join who are experts, it is quite difficult to point to particular amounts of yield related to that, but they will be part of that strategy for increasing our focus on the wealthy.

How much are you expecting the wealthy team to contribute towards the Government’s aim of raising £7.5 billion in additional tax revenue by the year 2029-30?

As we said, those additional 400 people who are coming in to tackle particularly the wealthy offshore side are expected to bring in at least £0.5 billion.

That is very helpful. I would just like to move on to my final category of questions, which is on penalties and prosecutions. If I could take you to paragraph 3.17, in 2023-24 you issued 456 penalties, but that was down from 1,747 in 2022-23, and down further from 2,153 in 2018-19. This is a considerable drop. Is there a policy reason for this?

No, it is really dependent on the mixture of the work we do and the particular circumstances of the caseload. Penalties can vary depending on the type of case that we are looking at. Philippa, do you want to add anything?

In essence, we charge the penalties in accordance with the legislation. Behavioural penalties, which is what these were talking about, as opposed to automatic penalties where someone is late with their filing, varies in accordance with both the tax lost and the behaviours presented. If someone makes a mistake despite taking reasonable care then they will not get charged a penalty, whereas if we find they are careless or deliberate in their behaviour then they will. That has really fluctuated over the years. Yes, 2023-24 was 5%, but it has been as high as 22% in other years. It really does vary. Some of the work we do does not attract behavioural penalties, so it is just highly dependent on the nature of the risk, but where we find those behaviours we will charge those penalties.

Just to add, Chair, we expect this to go up. We are expecting about a 20% increase in the cases that lead to prosecution, and therefore we think prosecutions will rise in the years ahead.

You have almost read my mind, because where I was going to next was prosecutions. HMRC’s work to tackle fraud makes use of its civil powers, so I am wondering whether you have sufficient powers in there. This is paragraph 18. The paragraph goes on to say, “HMRC’s policy has been to reduce the overall number of prosecutions for all taxpayers in order to focus on high-value, high-harm fraud, while also facing delays in the criminal justice system”. It then makes a fairly devastating statement that says, “HMRC does not fully understand the deterrent effect of prosecutions but is undertaking analysis to improve its understanding”. That is a fairly damning statement, really. If you do not know what the effect of prosecutions is on deterrent by now, when will you?

Jonathan might just say a bit on the analysis and the evaluation, because what the sentence is alluding to is the quantification with a level of certainty of the deterrent effect. Let us be clear: we think it matters, and we are committed to increasing those cases in terms of expanding counter-fraud capability. Our annual charging decisions on the most harmful fraud will increase by 20%, from 500 to 600 per year, by 2029-30. That will increase prosecutions. The message we have been trying to get over today about capacity, capability, taking on the most complex cases and ensuring that, where they need to lead to prosecution, we will do that—and increasingly so—is part of our response to the fact that we agree the deterrent effect matters. Whether you can quantify with precision the deterrent effect is a difficult analytical question that Jonathan will have a good answer to.

We have looked internationally and within the UK context as to whether we can quantify it. If you read that somebody has been prosecuted for tax evasion, how much more likely are you to be compliant? Can we put a pound and pence figure on that? That is the challenge. What we have been able to do is to look at the ways that deterrence works. That actually reinforces the model we have. We find deterrence works primarily through media and social media. The more severe the penalty, the more newsworthy it is, and therefore the more people are aware of it. Therefore, focusing on the most severe cases, which are therefore likely to attract the most severe penalties, is the most likely way that deterrence will flow through the economy. That is the rationale, but can I give you a pound and pence figure on how much extra compliance we have from that? I am afraid that is where the analysis and the evidence runs out.

The Chair raises a really important point here. Following our session with your predecessor, JP, the Government agreed with this Committee’s recommendation that HMRC research which interventions are most effective in achieving that strong deterrent effect for tax evaders and other organised criminals. What progress has actually been made since we last had a session with HMRC in understanding which intervention, whether that is a civil penalty versus a criminal prosecution, is truly the most effective deterrent? What was the value of the penalties associated with that single case that brought in £2.5 billion of compliance yield? It would be useful for us to know when a penalty or a prosecution is most effective.

As I said, in terms of the deterrent effect, the evidence is that the more severe the penalty and the more severe the sanction, whether that is a custodial sentence or anything else, the more likely it is to drive newsworthiness and therefore have a deterrent effect. What you are hinting at is probably our more overall approach to compliance, which is multifaceted, not just relying on prosecutions. Again, it is about whether we can we stop it arising and then whether we can effectively deal with compliance when we find it, using civil and criminal penalties.

I have one final question, if that is all right, Chair, just to park this issue. In 2023-24, HMRC prosecuted only 25 wealthy taxpayers, as set out in the NAO Report. I am concerned that that does undermine the rule of law if offences were not pursued in the courts. There is a serious risk that some very wealthy individuals in this country will think that you can get away with not paying your taxes. How effective is our tax evasion deterrent, if you prosecuted only 25 wealthy taxpayers in 2023-24?

In terms of the stats I have available, at the end of 2023-24, as you say, HMRC had 399 wealthy individuals under criminal investigation, and secured 47 charging decisions in relation to wealthy individuals. That returned to that pre-pandemic level, so it was an improvement. I have set out what we are doing to increase capacity and capability to go further, including an increase in 20% of—

That would only get you to 30 prosecutions, though, would it not?

That is right; not every case leads to prosecution, but the success rate is quite good. I agree with your point. We want to do more of these cases and we want to see prosecutions rise. That will increase the deterrent effect. That is the plan.

We are going to have to call it to a halt there, because we are in danger of losing a quorum. Can I thank our witnesses very much indeed? We have covered a pretty broad field today. You have answered with candour and a depth of knowledge. JP, I am very grateful to you and your team. An uncorrected transcript of this hearing will be published on the Committee’s website in the coming days. The Committee will consider the evidence that you have provided, and will produce a report with recommendations in due course.