Public Accounts Committee

Oral evidence: Digital Services Tax, HC 732

Thursday 8 December 2022 

Ordered by the House of Commons to be published on 8 December 2022. 

Watch the meeting 

Members present: Dame Meg Hillier (Chair); Sir Geoffrey Clifton-Brown; Mrs Flick Drummond; Sarah Olney; Nick Smith. 

Gareth Davies, Comptroller and Auditor General, Adrian Jenner, Director of Parliamentary Relations, National Audit Office, and David Fairbrother, Treasury Officer of Accounts, HM Treasury, were in attendance. 

Questions 1 - 105 

Witnesses

I: Jim Harra, Chief Executive and First Permanent Secretary, HMRC, Jon Sherman, Director of Business, Assets and International directorate, HMRC, and Mike

Williams, Director Business and International Tax, HMT.              

Report by the Comptroller and Auditor General

Investigation into the Digital Services Tax (HC 905, Session

2022-23) Examination of Witnesses

Witnesses: Jim Harra, Jon Sherman and Mike Williams. 

Q1 Dame Meg Hillier: Welcome to the Public Accounts Committee on Thursday 8 December 2022. Today, we are looking at the digital services tax, which is a temporary tax meant to serve until new international frameworks are adopted. It is for businesses that work online. A lot of it stems from the work of this Committee under my predecessor, Dame Margaret Hodge, who really pushed on why we were not able to tax properly some of those big internet businesses that we all know about. 

Last year, the digital services tax raised about £358 million, but about 90% of that came from just five businesses. Today, we are discussing how it is working, what lessons HMRC is learning, what gaming might be going on by businesses, how it is going to work while we wait for the full international agreement to kick in, and what progress there is on that. So, lots to cover, and it is all critical, because it brings money into the Exchequer, which is critical at the moment. 

I welcome our witnesses: Jim Harra, chief executive and first permanent secretary at HMRC; Jon Sherman, director of the business, assets and international directorate; and, from the Treasury, Mike Williams, director of business and international tax. Welcome, Mr Williams. I don’t think you have been in front of the Committee before. 

Mike Williams: I have been virtually only. 

Chair: Virtually—forgive me, I didn’t recognise you in real life. 

Before we go into the main session, may I get an update from you, Mr Harra, on the move from CHIEF to CDS? To be clear, the original deadline for transition to CDS was going to be 20 September. It was extended to 31 October, which was the hard deadline, so it is now in place. In the runup to that, however, the trader dress rehearsal system caused quite a lot of challenge. One of the issues there was that HMRC was changing some of the tariff codes. 

Just to explain to people who might not have been following this, businesses were able to do a dummy run, enter a code and get the paperwork, but the code was a temporary one, so sometimes a week or so later they would find that the code was no longer valid, and they had to redo the whole process. It has been quite a torrid time for businesses going through that. I wonder whether you can give us an update on where we are now and, in particular, whether you think there are any lessons from that dress rehearsal system. 

Jim Harra: The migration of import declarants from the old CHIEF system on to the new customs declaration service has gone well. You are right that we had a trader dress rehearsal service, which enables declarants to practise using CDS, to test their own business plans and to make sure that they are ready to migrate. That was used extensively. In the week commencing 21 November, 91% of all import declarations were made on CDS. That is the extent of the migration that there has been. 

I think what you are referring to is that there are a small number of declarants who, because of the way we hold their data, are not able to migrate to CDS, even if they try to do so. We know who they are, they know who they are, and we have been in communication with them to ensure that they are comfortable that they can continue to use CHIEF while we resolve that. It is, for example, businesses that had an EORI number, but then registered for VAT and therefore get a different EORI number. The two systems do not match, so CDS does not recognise the number they try to use. The vast majority of declarants have not had that problem. 

At the end of November, about 248 declarants accounted for 87% of all the declarations—240 had migrated successfully to CDS by then. About 3,000 declarants who account for the other 13% of the declarations, although some of them are regular declarants at low volume—some might only declare once a year, if they just import seasonally. About 2,500 of them have migrated from CHIEF to CDS. Some of them will not have migrated because they are not ready to make declarations yet. 

In addition, there are about 820-odd declarants who are new on CDS and who were not previous users of CHIEF, and they are also successfully using the system.  

switch off CHIEF. Do you have any estimate for when that is likely to be? Or is there a long tail, related to some of the functions of some of the businesses in the Northern Ireland trade. 

Jim Harra: We want to stop using CHIEF as soon as possible and get everyone on to CDS. It is expensive and complex for us to run two services in parallel. It is also expensive for the declarants because they have got to keep two sets of software running in parallel. We want to do that as quickly as possible, but we need to make sure that we give everyone who needs it the opportunity to migrate. Our experience with import declarations has been that while there is a small tail, every single one of them who has asked for approval to use CHIEF beyond 31 October has got a plan and a date for when they believe they will successfully migrate. They are now small enough numbers that we can help them one by one. I do not think that is an issue. 

There are one or two areas of functionality—sort of edge cases—that we need to deal with. For example, one of the reasons why some people could not migrate from CHIEF to CDS is if they were an Isle of Man declarant some of the functionality was not quite there for them. We have got a little more of that to do but it is really small numbers. 

pleased to get more than expected. 

Jon Sherman: On that, the impact of the pandemic would be different in different sorts of areas. For example, if the services you are providing are to do with travel, then during the pandemic those were obviously hit in the other direction. Some companies will have benefited quite significantly, but for others the impact will have been different. 

Q10 Chair: Is the Treasury building in an expectation that you will get a certain amount of income from the digital services tax? How bold are your assumptions on that? 

Jon Sherman: The forecasts will get updated. I think they will have been updated just now, in the latest OBR EFO and— 

Q11 Chair: To be clear, you are relying on the OBR forecast for what you are going to get? You do not have a target internally—Treasury or HMRC. 

Jim Harra: HMRC does not have a target for what it achieves from a tax. We don’t forecast either; it is the Office for Budget Responsibility that forecasts, although of course we feed into its forecasts, giving it data and assistance with that. Its forecasts are central forecasts, so they are not bold forecasts, I think it would say. 

In the case of the digital services tax, we have to be alert to, first, digital services being a fast-moving market, so the players change all the time, and social media platforms grow and decline—so that is a slightly moving target. Added to which, as Jon said, the pandemic meant that some platforms that were potentially within the scope of DST would have been badly impacted, if they were offering travel services, for example, but others really benefited from the pandemic, so there were shifts in the population. But 90% of it is paid by 5 digital giants, and we pretty much all know who they are. They have been a more stable base for this throughout, from the start of the design. 

While we will keep monitoring that, obviously, the intention is that it will be replaced by something different by 2024, and that it is an interim tax. 

Chair: Yes, well, we will move on to whether that 2024 date is realistic. I will hand over to Sir Geoffrey Clifton-Brown.  

Q12 Sir Geoffrey Clifton-Brown: Good morning, Mr Williams. Can I start by examining some of the upper and lower thresholds, and why you have fixed on those particular thresholds in designing the tax? In terms of the actual rate of tax, which, as Mr Harra has made clear, is on turnover, how did you fix on that 2%?  

Mike Williams: Obviously, it was the then Chancellor of the Exchequer, now Lord Hammond, who fixed the rate, because he was Chancellor at the time of its introduction. At that time, some countries had gone for really quite high rates—over 5%—which I think you would have difficulties with, with a turnover tax, because if the business to which you apply the tax does not have a very high margin, you risk taking away more than their profit. Other countries, some in the EU, had gone for a 3% tax. If I recall correctly, the EU had had a proposed directive suggesting 3%. On the other hand, there were other countries—Germany is probably the best example within Europe—that had decided not to do this at all, so they effectively have a 0% rate. I think that the Chancellor thought that, against that background, 2% was reasonable for something that was meant as an interim.   

Q13 Sir Geoffrey Clifton-Brown: Fair enough. On the upper thresholds, groups are liable to pay DST if their worldwide revenues from in-scope activities are more than £500 million, and more than £25 million had arrived from UK users. The group’s first £25 million of UK revenues are also exempt from DST. That, presumably, is why we are only really getting most of the tax from those five big digital players. How did you fix on those thresholds?  

Mike Williams: Well, again, those high thresholds tend to be a feature of digital services taxes that have been introduced. I think that is for two reasons. First, it is a sector that tends to have a few dominant players—if you read the literature, many would argue that that is a problem or issue with the sector—so you can, fairly readily, get most of the revenue by focusing on those players.  

Equally, doing this sort of revenue-based tax, where you are looking at revenues not directly derived from the users—you know, you and me using the search engine—but from the people who pay for the advertising, it is quite a complicated and complex tax. There is not really a need to apply that to the people at the bottom end, who would then have to put the systems in place to track it, because there just is not much money in it at that end. In a sense, the disadvantage of digitalisation is that you tend to get these dominant players, but the upside is that you can more readily focus a tax.  

Q14 Sir Geoffrey Clifton-Brown: Given that this is essentially a displacement tax on where businesses are domiciled and where they are making their profits, you must have had—we know, from the Report, that you did— some fairly sensitive discussions with the US tax authorities. Was that an influence in where we set our various thresholds for this tax?  

Mike Williams: I think that all Governments, not just the UK Government, introducing a digital services tax, in the face of what was known to be US opposition, would have considered that US opposition in framing the tax and setting the rate.  

Q15 Sir Geoffrey Clifton-Brown: So you framed it. You obviously discussed with them the thresholds and so on, because that would have been critical to their businesses—which would be affected and which would not. Was there a discussion with the US tax authorities around these thresholds, or was it just something that we fixed?  

Mike Williams: No, no. This is a UK domestic tax; we decided the rates and the thresholds. If you then abstract to pillar 1, then, of course, with that requiring international agreement, all the thresholds, all the allowances and everything within that, have to be agreed by all of the participating countries.  

Chair: We will come on to where we are at with international agreements later.  

Q16 Mr Djanogly: On the international front, obviously the OECD must be looking over your shoulders on this, because they are looking at implementing it for about 140 countries themselves. Have they been in touch with you? Have they given you their views on what they think we are doing right or wrong?  

Mike Williams: I don’t think they would ever be quite so direct, if I can put it that way. But yes, we have been heavily engaged with the OECD on this from the start of the base erosion and profit shifting—BEPS—work that started in 2012. One of the actions in that was to consider how best to tax the digital economy. You could reasonably argue that that was the unfinished business from the BEPS project, which reported in 2015. The OECD returned to it in 2016, I think—probably sooner than people at the OECD originally expected, because there was more pressure and more concern about the fact that the existing rules do not tax big digital businesses terribly effectively. 

Mr Djanogly: Right.  

Jon Sherman: I think the OECD also produced some guidance a few years ago on interim measures. The DST is broadly in line with the principles they set out.  

Mike Williams indicated assent. 

Q17 Mr Djanogly: Broadly in line? 

Jon Sherman: Yes. 

Mike Williams: At one point there was the prospect that there would be a template for a digital services tax, as you know. The OECD would have said, “If countries are going to introduce this interim measure, you should broadly do it this way.” It was not possible to get international consensus on that, so countries were, in effect, left to themselves to craft their own DSTs.  

Q18 Chair: Which begs the question of whether we will ever get international consensus from the 140 countries.  

Mike Williams: Which is probably another reason for introducing something on an interim basis in the meantime. 

Q19 Mr Djanogly: Are you getting any feedback or making any studies into the effectivity of our rates compared with international rates? You said that some started off a bit higher and we started off a bit lower, but is there a review going on as to which are more effective? 

Mike Williams: Like all taxes, we keep the DST under review, and we probably look more closely at newly introduced taxes. Given that the DST is, by its nature, imperfect, the main aim is to move as quickly as we can to pillar 1 and tax profits, rather than tax turnover.  

Q20 Mr Djanogly: In the meantime, is the main purpose of this to have people paying their fair share, or to raise revenue? What is the motivation? 

Mike Williams: That is a very good question. There have been articles in the US tax press—they tend to always focus on the UK and France—that say, “The DSTs in those countries were introduced, and they can’t afford to give up the revenue.” But you are talking, in both cases, about half a million pounds of revenue a year in economies that are over £2 trillion a year. The revenue is clearly important, but I would say that getting to a fairer system is more important and more the driver.  

Q21 Mr Djanogly: So you are saying that, on the whole, the US do not want to see this as a fairness issue.  

Mike Williams: I am not sure that they would put it in those terms, and I would not want to put words into their mouth.  

Mr Djanogly: Fine. Thank you.  

Q22 Sir Geoffrey Clifton-Brown: Can I come back to the timing? This is a sunsetted tax; you are not intending to continue it after the financial year ending in April ’24. Is that correct? 

Jon Sherman: The commitment is that it will be retired once pillar 1 takes effect, but there is also a provision that it should be reviewed. If that has not happened by the end of 2025, I think there is a commitment to do a

review.  

Q23 Sir Geoffrey Clifton-Brown: That is the next question, Mr Sherman. If, for some reason, pillar 1 is delayed, will this tax continue? 

Mike Williams: I think it would be the subject of this review. As part of that review, the question would be, “Should you continue with the tax?” Again, if we were to get to that point—obviously, it is hypothetical at this stage— the Government would want to consider what the alternatives are. You would hope that the international solution would come very quickly: either keep on with what we have got or, in effect, do nothing. But again, I would have thought that you would want to look at what other similarly placed countries were doing at the same time.  

Q24 Sir Geoffrey Clifton-Brown: Okay. In designing the thresholds that I was asking about earlier, was one of the considerations that there might be double taxation involved here, given that this DST cannot be offset against corporation tax?  

Mike Williams: That is certainly something that we took into account— the possibility that some businesses will pay corporation tax or corporate income tax in some other country and in the UK as well. That is one of the factors, and one of the constraints.  

Q25 Sir Geoffrey Clifton-Brown: So a situation does exist, at the moment, where people will be paying double tax on the same activity?  

Jim Harra: By its nature, if they are profitable, the turnover that is being subjected to digital services tax is contributing to their profits, which are then attributed to whichever country they say has taxing rights over those profits. The reason for the digital services tax is because of the sense of unfairness that those profits are not being attributed back to the UK, hence the OECD work to improve the way profits are attributed. So, in that sense, yes, in theory, a profitable business must be subject to tax somewhere on the profits that are generated from that turnover.  

On the other hand, if they are not profitable, there is a provision in the digital services tax that they can apply to sort of disapply the DST. If they demonstrate that there is no profitability for them from those activities, then we won’t tax them on their turnover. 

Q26 Sir Geoffrey Clifton-Brown: Where does the alternative methodology come within that, and has anybody used the alternative methodology?  

Jon Sherman: That is the alternative methodology. It kicks in, basically, if the margin on the services that are in scope is less than 2.5%—then it is advantageous for you to use that basis. It is a different calculation. It is basically 0.8% times the margin.  

Q27 Sir Geoffrey Clifton-Brown: Has anybody used it, Mr Sherman? 

Jon Sherman: Yes—I don’t have the actual figures on that, but I would assume that is the case. We could find out for you. Of the companies that have made returns under the DST, a number of them have paid and a number of them have not paid. The ones that haven’t paid will be ones where they are in a loss situation—where, as a result of the alternative methodology, there is no DST payable. In that sense, those companies will certainly have used it.  

Q28 Sir Geoffrey Clifton-Brown: Given that basically the rules for pillar 1 should be known from the middle of next year onwards, have you started to do any thinking and modelling on the effect on UK tax revenues of a change from DST to pillar 1 corporation tax changes?  

Mike Williams: We haven’t yet modelled them. To lift the bonnet on pillar 1 a bit—this is not surprising—countries found it easier to agree where profits should move to, but, obviously, for any particular group, there is only one lot of profits, so in deciding that profits of £x should move to another country, which country those profits are going to come from is the largest unfinished piece.  

That is quite important in terms of determining the revenues that any particular country would get from pillar 1. It is reasonably clear what you will get on pillar 1. The missing ingredient to start doing a calculation is, what you are going to give, because you have activities that are generating profits that you are taxing that pillar 1 may move somewhere else. That is the essential missing piece in the jigsaw.  

Q29 Sir Geoffrey Clifton-Brown: Within that question, financial services are currently exempt from DST. Will they be exempt under pillar 1?  

Mike Williams: There are two main exclusions from pillar 1: one for extractives—oil and gas basically and some of the natural resources—and the second for financial services.  

Q30 Mr Djanogly: Have you seen anything that would show that companies affected by DST are going to be passing it on to their customers? Are they doing so?  

Jim Harra: I believe at least one large company has publicly declared that that is what they are doing. Obviously, their pricing is a matter for them and if they are in a competitive market, they will make their choice. There is nothing in the rules that says what they have to do, but I know at least one of them has said that they are quite explicitly doing that.  

Q31 Mr Djanogly: Have you any feedback or comment that that might reduce online business? Have you seen any implications of this tax reducing online business in any way?  

Jim Harra: There is no evidence from the company involved that it is having any adverse effect on their business.  

Q32 Chair: Covid has to have had a big impact on whether you can make that judgment, I guess.  

Jim Harra: In strictness, any tax is going to have some distortive effect. That is just, in principle, inevitable. There is no tax that doesn’t, but this is relatively small, I think, in the scale of what they do. It is on their services. It is not just added on to the price of goods that they sell. You cannot see any indication in the way that they are performing that it is having a significant impact. 

Q33 Mr Djanogly: Because of the very large threshold—£500 million worldwide revenues—could it have the impact of pushing consumers to deal with smaller online retailers who are unaffected by the tax? 

Jim Harra: My view would be that this is so small, and probably not terribly transparent in terms of how it is impacting the prices that they charge their customers versus the salaries or dividends that they pay their executives, it is unlikely to have that impact. However, we track very carefully what is happening, particularly in consumer-facing digital businesses, because it affects the effectiveness of the design of our different services. For example, most online sales of goods are through online marketplaces or a few giants. From a tax administration point of view, that assists us in what to design. If in 10 or 20 years’ time you had a very different make-up of the digital market, where people were buying direct from lots of smaller businesses, that would definitely affect the design of a number of our taxes, but this is a temporary tax. It has only been in since 2020. It is at a fairly low margin. I doubt it is having that kind of impact in reality. 

Q34 Mr Djanogly: Are you modelling the threshold level to see whether making the net wider would have an impact? 

Mike Williams: If we were to do that, it would bring in more businesses and potentially, going back to an earlier question, bring in more businesses that maybe were also paying corporate tax—who knows? I don’t think that you would get significantly more revenue, so you would be increasing quite significantly the volume of businesses that were faced with having to comply with the tax, but the amount of additional revenue, because of the way the businesses are concentrated, would be really quite small. 

Q35 Mr Djanogly: The vast majority of businesses impacted at the moment are US businesses, I understand. At what level, from £500 million down, does it start impacting on UK companies? 

Mike Williams: We have never modelled that. 

Mr Djanogly: You haven’t modelled that. Thank you. 

Q36 Sir Geoffrey Clifton-Brown: This is probably to Mr Williams, or perhaps Mr Harra. This is no way intended to be a critical question; it is a question to seek information. You originally expected DST receipts to be £275 million. In fact, they have turned out to be £358 million, which is good news as far as the UK taxpayer is concerned. Can you give us an idea of why there is that quite significant difference between what you expected and what you actually achieved? 

Mike Williams: Let me have a go, and then I may ask Mr Sherman to come in as well. There are two things. First, you are introducing this new, quite different tax. In those circumstances, it is intrinsically quite difficult to estimate what you will get. In the nature of things, part of the OBR forecasting process is looking at what revenue you are getting and forecasting the revenue that you will get from that tax for later years, which is obviously very sensible. You do not have that sort of information, so that would be my first point. The second point goes back to what the Chair said. The tax, quite fortuitously, was introduced just as the pandemic came. I think that then changed habits. There is press comment that it led to digital businesses in particular doing quite well from the fact that people were stuck at home and doing more of the things that generated profits for them. 

Jim Harra: A couple of other points on that. First, strictly speaking, we do not forecast tax revenues; that is the Office for Budget Responsibility. Although HMRC published a forecast in the impact note, it was lifted from an OBR forecast. In its forecast, it built in a 20% discount because it assumed that we would see some tax planning behaviours from the payers of this tax that would reduce the take.  

In fact, our experience is that we have not seen that. I would hope it is down to very good design of the tax, but I suspect it is also down to the fact that it is an interim tax, and they may have concluded that it is really not worth their time to try and do that. But that does mean that the receipts are higher than the forecast, because with hindsight, that discount was not required. 

Q37 Sir Geoffrey Clifton-Brown: It is a great credit to you and your organisation, Mr Harra: you introduced it with a very low level of resources and a low number of people, and it seems to be a very precise and welldesigned tax. Are you able fairly straightforwardly to estimate the tax gap—what you should have collected and what you have actually been able to collect? 

Jim Harra: Our compliance work is ongoing. We have been concentrating on making sure that we have got within the tax all of the companies that are potentially within its scope, and then on supporting them to help them to self-assess their liabilities, more so than investigating potential evasion or avoidance. As I say, we have not to date seen evidence that they have tried to avoid it. 

We have identified 101 businesses that are potentially within the scope, and we are working through all of them. A small number of them are still in the risk assessment phase, but by and large, we are finding that the people who should have declared this tax have done so. Obviously, 90% of the liability is concentrated on the five giants, so I think we are pretty confident that that gap is very low. But for our own internal purposes, yes, we will want to make sure that we understand the scale of the risk here, because while it is a relatively small team, I do have some specialist resource deployed on it, and I need to make sure that that is the most efficient use of those resources going forward. 

Q38 Mrs Drummond: The NAO thinks that cryptocurrency is within scope, so can you just let me know whether you identify the sales—commission and things—and how you identify capital gains tax and other forms of tax from cryptocurrencies? It is a massive market. 

Jon Sherman: Cryptocurrency would be in scope, in the sense that if you had a marketplace where the service being provided was crypto, that is potentially within the scope of the DST. 

Mrs Drummond: The marketplace for cryptocurrency is about £1.6 billion, I think. 

Jon Sherman: If they are providing an in-scope service where they are active as a marketplace, they will be within the scope of the tax. For the end buyers of the cryptocurrency, we have produced a lot of guidance on the tax treatment—when capital gains tax is due and so on. Obviously, it is a rapidly evolving technology, and we are developing our compliance approaches to that, but we are one of the first tax authorities to get pretty comprehensive guidance out around how crypto is taxed, to try and make sure we are informing that. 

Jim Harra: Just to be clear—because I think there are two things potentially conflated in your question—you are right that a digital company providing services in relation to cryptoassets could be within the scope of the digital services tax. That is not relevant to the capital gains tax, which is— 

Q39 Mrs Drummond: I understand. I was just wondering about the compliance with that, really. 

Jim Harra: We do actively monitor that as a risk. Ownership of cryptoassets is growing, and if you sell them or swap them for another cryptoasset, then if you have made a gain on that, that is liable for capital gains tax.  

There are two key areas of risk there. One is whether we can manage the risk from the really big gains and get transparency, but the other is that there are probably a very large number of fairly unsophisticated people who do not normally make self-assessment returns who might make a moderate gain if they own cryptoassets, but above the annual exempt amount. We need to educate them and raise awareness, so that they are aware that they will have an obligation if they do that. We have a forum, which includes the Low Incomes Tax Reform Group, that tries to make sure that we reach not just the big, sophisticated end of the market, but also that broader end. 

Mike Williams: There are also the separate work, separate to pillars 1 and 2 at the OECD on exchange of information between countries on cryptotransactions. If successful and if put in place, that will reduce the temptation, say, for someone in the UK to transact on a crypto market outside the UK in the expectation or hope that HMRC would not find out. If we put this in place, HMRC would get data from outside the UK that would help with that. 

Q40 Mrs Drummond: Do you know how much income you get at the moment on commission in the marketplace for cryptocurrencies?  

Jon Sherman: Do you mean as part of DST revenues?  

Mrs Drummond: Yes.  

Jon Sherman: No, we don’t have a breakdown for that.  

Q41 Mrs Drummond: What sort of category has it been put into, then? There are a few categories here. 

Jon Sherman: I assume an online marketplace is what we would be talking about.  

Q42 Sarah Olney: When we talk about transitioning from DST to OECD pillar 1, how much certainty do we have around timelines? Are we still expecting that to have happened before DST is reviewed in 2025?  

Mike Williams: Yes, that would be the expectation at the moment. The OECD timetable has the multilateral convention that you need to implement pillar 1 ready for signing in mid-2023. That is quite a challenging timetable, but, equally, there have been challenging timetables before on pillars 1 and 2.  

Q43 Sarah Olney: So when we talk about DST being a temporary tax, it is still very much your expectation that it will end by 2025. 

Mike Williams: I think that is the expectation. Equally, as we discussed earlier, if it does not end by 2025 because pillar 1 has not been implemented by then, the review would consider what should be done.  

Q44 Sarah Olney: Thinking about the review, what factors do you think you will be using to evaluate how successful DST has been? Even between now and 2025, what do you think the success factors are for DST?  

Mike Williams: I think a key success factor, looking at it in isolation, is that we are bringing in revenues that would not otherwise have been brought in, in circumstances where I think many people would say that was fairer—that these businesses are paying more of a fair share.  

As I said earlier, it is hypothetical to look forward to 2025—we do not know, as it is three years on. I think a key question will be, where are we placed in 2025? Has the project, if you like, fallen by the wayside because some countries, despite having signed up in July and October 2021, have decided that they won’t or can’t implement, for some reason? It seems to me that is one background, but another possibility in 2025 is that it has just taken slightly longer than we all thought, but it is basically going well and there is momentum. It seems to me that they are not similar scenarios, so it is quite difficult.  

Q45 Sarah Olney: What I am getting at is more, should DST be with us for longer than we expect, how can we be sure it is operating in the best possible way? We have a review planned for 2025, but there may well be interim reviews, for whatever reason. You said “fairer”. Exactly how will you be judging that it is suddenly fairer? Is it just that these companies, which earn a lot of global revenue, are now paying more tax in the United Kingdom? Is that your benchmark for fairness?  

Mike Williams: I am glad you raised that, because I think one person’s definition of fairness will be different from another’s. The businesses paying this tax would say— 

Sarah Olney: They will certainly have a different view.  

Chair: But there is a definite public expectation.  

Mike Williams: Yes, and there is a difficulty with this, obviously. I doubt that the businesses paying this tax that were not paying it before see it as fair, as is the nature of things. But if you accept that the weakness with the existing rules is that there is a lot of user activity taking place, particularly on social media platforms in the UK, and that activity is not reflected in the attribution of profits to the UK generally at the moment, so we don’t get tax on those profits, then that does not seem very fair. We cannot change those rules ourselves, so this is our way of saying, “We will do this second-best thing in the meantime.” 

Q46 Chair: So you would describe it as the second-best thing.  

Mike Williams: Yes.  

Q47 Sarah Olney: So we do not have a robust way of saying, “DST is working in exactly the way it was intended and it is bringing in the revenues that we think it should be bringing in, based on some fairly fundamental measures of what we expect the tax to achieve.” It is just that, generally, we are getting a bit more tax revenue from people who are earning a lot, and we will accept that slightly woolly measure of success because we don’t expect DST to be in place for very long. 

Mike Williams: I would put it slightly differently. There were those who said, “You shouldn’t introduce a DST because people will avoid it to such an extent, so why would you bother?” I think the evidence is that that is not what has happened. Their challenge really was, “You cannot introduce a robust enough tax, so don’t waste your time dealing with this interim solution.” I think the evidence is, having put in place the DST as an interim solution, it is. 

Q48 Sarah Olney: Can I just quickly ask you about that? We know that 90% of the tax income is coming from five companies—we all know who they are, but we are not going to say. How do you know that they are paying the right amount of DST? 

Mike Williams: I think HMRC colleagues would have to answer that. 

Jon Sherman: We will engage with all those businesses and do risk assessments. Where necessary, if we think there are significant risks, we will challenge the returns—for any company.  

Q49 Sarah Olney: But do you think you have a robust measurement of how much they should be paying, based on their revenues, versus how much they are paying? 

Jon Sherman: With any business that is filing a return for DST, we will do a risk assessment of that return based on the information we have and consider whether there are risks that the return doesn’t actually reflect what they should be paying under the law. We will look at that. 

Jim Harra: Our approach to that is twofold. First of all, we have our general approach to managing the compliance of these large businesses, which has been very effective. We effectively man-mark them. It is quite a resourceintensive method of managing their compliance because of the scale of the tax they pay and the type of risks they pose. That applies to this tax equally as it applies to corporation tax and VAT. All of these businesses have a customer compliance manager who understands their attitude to taking risk with tax compliance, their internal governance, the quality of their tax department and so on. In addition, we have a small team of DST specialists who are also available to work specifically on the methodologies that the businesses have used and how they have complied the data.  

Actually, a lot of these large businesses engage with us well in advance because they have big challenges here: “Okay, you want to attribute this data to activity by users in the UK. How do we determine from our data who our users in the UK are?” They will have engaged with us on the methodology they are using to give us and them confidence that it is going to come up with a result that we think is correct. We have a high level of transparency for those big taxpayers over how they have gone about this. We have worked closely with them to make sure that they have a methodology that is acceptable both from our point of view, in terms of its outcome, and from their point of view, in terms of its feasibility for them to administer. 

We have a quite high level of confidence in both the initial returns and our ability to manage that. Our primary concern at the moment is making sure that we have captured the complete population of businesses, and in particular businesses that might not have to pay it now but might have to pay it in due course, so that we know who they are and they are ready with a methodology if they need to apply.  

Q50 Sarah Olney: Does every business group that is paying DST have that level of relationship that you just talked about? 

Jim Harra: The largest of them certainly do. The smaller ones may be in our mid-size business population, but even there we have a pretty resourceintensive way of managing their compliance; they are logged by the central DST team, who will know if they want the compliance group to deploy some resources on looking at that business. Although there is a small team of specialists, they have behind them the whole apparatus of— 

Q51 Chair: What about the ones that are not UK based? Do you cover those too? 

Jim Harra: First of all, with the way this tax is designed, it can be attached to basically any part of the group. Most of them will have a subsidiary company in the UK, but even if they don’t they are a UK taxpayer for this purpose and they will therefore be on our books. One of the challenges with this tax could be the enforcement of payment if these groups have no real presence in the UK. That has not been an issue so far, but it is one of the risks on our control register. 

Q52 Sarah Olney: At the moment it seems that we think the tax has been successful because we have raised income, but we have also identified the fact that the covid pandemic and change of consumer behaviour has really influenced that. When you are reviewing how the tax has gone so far, to what extent do you think covid has been significant in respect of the levels of tax revenue we have seen? How do you think that might change? 

Jim Harra: Our view is that the disruption will have impacted on revenues in that first year. As Jon mentioned, there will be some groups—for example, in the travel sector—who we might have expected to be DST payers who will have had a very bad time of it; there will be others whose business model will have benefited from the circumstances of the pandemic. How that is all going to change over time, we cannot predict— 

Q53 Sarah Olney: I understand that, and you said that earlier. What I am trying to get at is to what extent the unusual nature of consumer behaviour during covid has potentially led to higher revenues than you thought it would. To

what extent is that influencing your feeling that this has been a success? Might you feel differently about it in different circumstances? 

Jim Harra: Broadly speaking, despite the impact of the pandemic, once you strip out the 20% discount from that original forecast, there is not that big a difference between what we have collected and what the gross amount was. The main difference between what we are seeing and what was forecast is that we are not seeing the kind of planning and avoidance behaviours that we might have expected to see and that the OBR conservatively took into account. 

In terms of success, as Mike said, our view is that the best thing to tax is profits, not turnover. This has been successful in the sense that the payers who are liable to pay it appear to be paying it. It is addressing what is generally perceived to be an unfairness or flaw in the way the existing international rules for attributing taxing rights over profits work, but the full intention is not to continue with it if we can get those international rules in place. 

Sarah Olney: Thank you. 

Q54 Sir Geoffrey Clifton-Brown: Can I go through one or two details of the tax? Mr Harra, you are to be greatly congratulated that your implementation costs were lower than you originally thought they would be. You implemented it for £6.3 million, which seems to be an extraordinary achievement. You were under budget on the technology side and on the personnel side. What lessons can you learn from the design of this new tax and the resources you devoted to it, in relation to other taxes? 

Jim Harra: In terms of project implementation, it is our standard practice to do a lessons learned at the close-down of the project, and we certainly have with this one. I think that is a bit different from the lessons you can learn from the design of this tax, for example. 

Q55 Sir Geoffrey Clifton-Brown: I was really thinking about the former and what lessons you can use from how you implemented this tax in relation to how you might implement other taxes. I am of course using this as a lead-in to asking in a moment about how you are going to implement pillars 1 and 2. 

Jim Harra: One key lesson from this is that it is a small group of taxpayers with whom you can interact on a one-to-one basis in relation to a new tax. A key lesson from that was about really managing the relationship with the taxpayers and making sure that we were introducing the right level of flexibility in the methodologies that they use so that we could have confidence not only that they gave robust results but that internally they were going to be feasible to implement. That would have been much more challenging if you had had a much larger mass population of taxpayers, but here we were able to have pretty much oneto-one relationships. A key learning was that it helped with the implementation that we really worked hard on those relationships and dealt with them in advance in respect of any methodological challenges that they might face. 

Q56 Sir Geoffrey Clifton-Brown: I understand that, because it is a relatively tight tax involving a relatively small number of taxpayers. I do not know whether you have seen the evidence from the Chartered Institute of Taxation, but it is quite interesting. They say in their evidence to us: “Pillar 1 will add a further significant compliance burden on MNEs, and we  are also concerned about the resourcing burden that will be placed on HMRC as the pillars introduce a whole new level of complication around achieving tax certainty and dispute resolution.” Given that you have not even started modelling it yet, how are you preparing for this? Given that pillar 2 rules are expected to be introduced with the accounting periods beginning 31 December 2023—they are going to start to make this change from the end of next year—what plans are you making for resourcing to introduce the new pillar 1 and pillar 2?  

Jon Sherman: On pillar 2, the plan is for the rules to start from accounting periods—effectively 2024. We already have an active project in place. We are looking at the IT that we will need to build for that in terms of returns and, in the long run, when we have to exchange data with other tax authorities, but we are also taking on board how we can build up our compliance skills to deal with the new rules. We published last summer draft legislation for part of pillar 2, and the plan is to have the final legislation in the Finance Bill.  

Regarding the burden on companies to file and the compliance activity that will go with that, the first filing date will be 18 months after the end of the first accounting period. It will be the summer of 2026 when companies start filing their pillar 2 returns, so there is a fair bit of time ahead there. The OECD is working on a further package of administrative provisions, part of which will focus on how the process can be simplified or made more certain. We are actively supporting that work at the OECD.  

On pillar 1 and the point about compliance, Mike might be able to say more about how that progress is going, but the way the admin framework for that is shaping up, it will be much more focused on international cooperation. Rather than each tax authority doing its own compliance work, the companies in scope will be able to apply for what is called a certainty process, so they will file to their parent jurisdiction and then that jurisdiction will work with the other jurisdictions affected to agree on how the various amounts are moving around, as Mike described before, where some countries are losing and some countries are gaining. There will be a process that will provide some kind of binding certainty for companies. That is all being thought of, but it will be quite different, in a way, from the way we do compliance within HMRC at the moment.  

Mike Williams: A key aspect there is that pillar 1 will apply initially to 100 or so of—by definition—the very largest businesses in the world. They are the ones who, if you like, have the wherewithal to cope with change and are likely, because it is more efficient to do this, to have closer relationships with tax administrations.  

To go back to the start of your question, Sir Geoffrey, I think we have learned lessons. This proposition that if you can get most of the money by focusing on a small number of large players then you should probably do that, if it doesn’t distort competition—you see features of that in the residential property developer tax, which is already in place, where we put in place quite a high threshold. Again, you are taking the tax from the bigger players. You also see it as a feature of the electricity generator levy, which the Chancellor announced in his autumn statement; again, the thresholds are high enough to keep out the smaller players who would struggle but would also probably divert resources.  

So we have learned from that, but you do have to be careful. You could simplify income tax if you said that only people making more than £1 million in profits would pay, but— 

Chair: Is that a new policy? You heard it here first. Williams for President! 

Mike Williams: Where the landscape allows you to do it, because you haven’t got a normal distribution, then I think the evidence is that you are making it easier on both sides. But you are right—you could get cheap popularity through thresholds that are too high.  

Chair: Leave that to us, Mr Williams—we’re the politicians.  

Q57 Sir Geoffrey Clifton-Brown: A more apt example might be that a lot of corporation tax payers would be very pleased to have a £25 million threshold before they started paying corporation tax. That is just by the bye.  

You mentioned 100 of the big players. Figure 5 in the Report shows that only 18 business groups are paying DST. This is a whole different ball game, is it not, Mr Harra, for the work that your organisation is going to have to undertake?  

Jim Harra: First of all, obviously the DST is just a UK tax, but pillar 1 applies globally. In addition, this is about digital companies, whereas pillar 1 will apply to all businesses of the right size, regardless of whether they are digital businesses, so they are different bases.   

Of course, in pillar 1 we are looking at the rules for attributing the taxing rights on profits, which are something that the businesses already calculate. It is building on the existing internal compliance that they already do in their businesses, but also compliance that tax authorities are used to doing. The main change, from our point of view as a compliance organisation, is the commitment in the international framework that any compliance work will be done multilaterally by tax authorities and not unilaterally.  

Q58 Sir Geoffrey Clifton-Brown: The Chartered Institute of Taxation’s evidence says: “in relation to Pillar 1, there is significant ongoing work and the progress reports published this year (in July and October) by the Inclusive Framework identify numerous open issues. Therefore a fully worked through global agreement remains some way off.” Are the current expected timetables realistic, Mr Williams, or is this whole thing slipping?

Mike Williams: There is reason to believe that we can meet that timetable. There is always a danger of slippage. Equally, in these very complicated international negotiations, the nature of things is that it tends to be the more difficult issues that get left to last. Of course, there is a danger that, because they are so difficult, you never reach agreement. 

Equally, if you can solve the most difficult issues, you are then finished and it will be done. In the meantime, it may well look to outside observers as if you are not making much progress, even though you are getting on with it. 

Q59 Sir Geoffrey Clifton-Brown: What are the critical milestones? When will we know whether this thing is on track? 

Mike Williams: The next milestone is the multilateral convention, which is available for signature in the middle of 2023. I do not want to trivialise this, but it is relatively easy for a country to sign a convention; the question then is whether it does the domestic processes required to implement it. The next crucial step will be whether enough countries do that. In the nature of these things, if 100 countries sign, you don’t need all 100 to implement it for it to be effective, but it is a big problem if the most significant players are lagging behind.  

Q60 Sir Geoffrey Clifton-Brown: Again, in their evidence, they say: “We— and other commentators—are doubtful that Pillar 2 will raise over £2 billion annually, as the government is predicting.” You say you haven’t done any modelling on this yet, but these are big figures.  

Mike Williams: That is pillar 2, though, Sir Geoffrey. The Government announced that we will implement pillar 2, and we have OBR-certified numbers in the autumn statement Green Book. I would say that they are consistent with the pillar 2 numbers that the OECD has put out for the global take.  

Q61 Sir Geoffrey Clifton-Brown: It is OECD figures, not Government figures.  

Mike Williams: Well, £2 billion is the figure shown in the Green Book in the line for UK implementation of pillar 2. What we haven’t done is pillar 1 figures at this stage.   

Q62 Sir Geoffrey Clifton-Brown: Right, but again, the Chartered Institute of Taxation says: “We—and other commentators—are doubtful that Pillar 2 will raise over £2 billion”. You say those figures are in the Green Book, but the Chartered Institute of Taxation is doubting whether you are going to raise that amount of money. Can you comment on that? 

Mike Williams: If you have an independent forecaster, you need to go by their forecasts. On £2 billion as the UK share of pillar 2, you can’t directly tie that number to the numbers that the OECD is producing, but there is nothing in tension between them. If the OECD was saying that the total is £5 billion, the proposition that the UK’s share is 40% would look unreal, but there isn’t that sort of tension in the numbers.  

Q63 Sir Geoffrey Clifton-Brown: The issue here is collectability, isn’t it? You can design a system and say that it is going to be a groundbreaking treaty or OECD agreement—you can predict what you like—but actually collecting that is a totally different thing. Again, I want to press you on this: is that number in the Green Book likely to materialise? 

Mike Williams: I think that the OBR, as usual, will have gone for a central estimate that factors in the fact that, first, people will change behaviour at the margins to try to get round the tax and, secondly, that at the margins some won’t comply. Pillar 2 targets businesses with above €750 million turnover. That is smaller than the target population of pillar 1, but they are not small businesses. Inevitably, much of the pillar 2 revenue will come from the very largest multinationals, who tend to engage with the system, I think.  

Q64 Sir Geoffrey Clifton-Brown: When you are looking at both DST and the new proposals on pillar 1 and 2, how concerned are you that these costs are going to be passed on to the consumer and about the effect that will have? There has been a certain amount of publicity around these costs being passed on to the consumer.  

Mike Williams: Yes, at least one business has said that it passes on the costs to the consumer. It is difficult to tell, because what is the counterfactual? The counterfactual is that the DST doesn’t exist. We cannot look in and see what their pricing structures will be in those circumstances, but I think all businesses seek to pass on costs anyway as best they can. That said, going back to the start, it is less good in tax policy terms to tax turnover than profits. Pillar 1, by focusing on addressing the problem more directly and on profits and putting them into the right place, will be a better solution in that sense.  

Jim Harra: Broadly speaking, if the outcome of the OECD work on pillars 1 and 2 is that multinationals pay more tax than they otherwise would, the burden of that has to fall on either their customers or labour or shareholders. That is inevitable. It is quite a challenge, and academics spend a lot of time trying to figure out where the burden of tax economically lands.  

The DST is a bit different, because that is actually a tax on turnover, and it is perhaps a bit easier for a business to say what choice they have made about whether they are passing it on to the customers who pay for their services or not. With taxes on profits, there has been a lot of debate over the years about where those tend to fall economically.  

Sir Geoffrey Clifton-Brown: Absolutely.  

Q65 Mr Djanogly: Have you assessed how easy it would be for business to avoid DST?  

Jon Sherman: We do have an anti-avoidance rule that is built into the legislation, which will deal with what you would call artificial attempts, if you like, to change behaviour to get activities to fall outside the tax. We haven’t seen any evidence yet.  

Q66 Mr Djanogly: You haven’t seen any evidence yet. Okay.  

Jim Harra: First of all, if we saw avoidance behaviours, we would try to tackle them. There are, however, legitimate changes in business models that businesses could choose to make, and the tax could be a factor in their decision. As I said earlier, there was a discounting built into the yield forecast. That is something we have not seen, possibly because the level of the tax does not make it worthwhile for them to change their business model in a way that is otherwise not optimal for them.  

Mike Williams: This was mentioned in the NAO Report. One of the things we were interested in —I am not sure I would say “concerned”—in relation to online marketplaces is whether you would see a behavioural change. The tax is on services, not the sale of goods.  

If I sell Jon goods over an online marketplace, the marketplace pays tax on the commission that they will charge me as a seller for using their marketplace, and that is within the DST. In contrast, if the marketplace itself buys the goods from me and sells them to Jon, that is not within the DST, because they are selling Jon goods. We did wonder whether businesses would do that, but of course, the person that you buy goods from is the person who is also liable if those goods are defective or dangerous. It’s called taking flash title—you own the goods very temporarily and then they move—but the question is, commercially, do you want to do that? 

Q67 Chair: It is a big risk to take on. It is a balance of risks. 

Mike Williams: Yes. Maybe for a 20% or 30% tax, which this could never have been, businesses might have contemplated that, but for a 2% tax, it is quite a big behavioural change. 

Q68 Mr Djanogly: Is it also a product of the fact that you’ve got only a very small number of very large companies? If you moved the bar down so that more companies were included, do you think you would start seeing more avoidance? 

Jim Harra: First of all, it is a small number of companies, but, like us, they expect this to be a temporary tax. Would you completely turn your business model on its head to avoid a 2% tax on commission that will hopefully in their view only last for a short time before it is replaced with something else? If you pushed down that turnover level, you are likely to get more and more businesses that were not profitable and for whom this was going to be more distortive. There is obviously the let-out in the design of the tax for that, but they have to go through quite a bit of compliance costs to get through that hoop. That is what Mike said earlier. Building those costs into the businesses and on to HMRC when the tax at stake is relatively small is just not worth it.  

Q69 Mr Djanogly: If someone was dishonest, what are the penalties that you can apply? 

Jim Harra: There are the general penalties right up to prosecution and conviction. Given the nature of these businesses, that is not the type of behaviour that you see. What you see is that they are bigger than national businesses. They use the international tax rules to their best possible effect. They have very clever lawyers and tax professionals in their organisations. That type of outright evasion is a risk that we watch out for, but it is not what we typically would see in this population. 

Jon Sherman: If we did have a case where there was a business model that had purported to change, if we had concerns that the reality did not match what we were being told, and it was really on the marketplace side of the line, for example, we would want to look at that and see what the facts said compared to the arrangements we were being told about to see if that stacks up. 

Mike Williams: In the nature of the businesses there tends to be data that provides an audit trail. Most of the money in this area is from the sale of advertising, and it is the sale of advertising targeted on us. But of course the payer needs to know if it is paying for the advertising—say, 100 people have received its advertising online—so I assume that the large buyers of this sort of advertising have some means of saying, “Prove that our advert got 100 hits”. The need to be able to prove that is then built into the systems of the people that we are taxing, which is helpful when it comes to seeing whether they paid the right amount. 

Q70 Chair: So you look at those figures when you are doing— 

Jim Harra: Yes. As I said earlier, with these large businesses we went through with them what is the methodology that they are going to use, what kinds of data do they hold, what decisions do they have to make about what that data tells them about whether users are in the UK or not. We have good transparency on what the big payers have done in terms of coming up with their internal methodologies for capturing the base of this tax. 

Q71 Mr Djanogly: And you are not doing it with many people. 

Jim Harra:  No, there are five big payers and 18 overall. But, as I mentioned earlier, primarily at this point our focus has been on making sure we have identified the potential payers, and making sure that we work with the larger ones in advance so that they know they have a methodology they can apply and that they can be confident we are happy with, and then supporting them. But there is still more work to be done to investigate in depth. 

Q72 Chair: Mr Williams, you were talking earlier about why people would not change a business model to temporarily hold goods and would keep the digital transaction, but that is when you are talking about physical sales and advertising to buy physical things. I can see that, but what about online transactions like online gambling? Have you looked at how the business model could be skewed to try and avoid the tax? 

Mike Williams: We did look at that. I think it is the same challenge. In effect, the behavioural change requires the operator of the market to assume more risk, and the question for the operator of the market is: do they want to assume the risk? 

Q73 Chair: Are they a gambling company or an advertising company? 

Mike Williams: Yes, or if the person on the losing side of the transaction does not pay, who faces the liability? Of course, if you change the nature of the transaction, it could become you and you have taken a big hit in return for— 

Q74 Chair: So you are really assuming that because of the legal challenges and the risk involved—moving into a slightly different market, even though you have a foot in that market—that would be enough of a deterrent for a business not to game it in that way. I am not suggesting you are wrong, but can you give us some evidence? Have you talked to companies about this? 

Mike Williams: We did talk to businesses about this in the consultation phase, but the 20% discounted by the OBR reflects the uncertainty over that. You can see that there is the scope to change your behaviour like this, but equally, as we have just discussed, there are risks and difficulties in doing that. Working out quite where the balance is there and in what circumstances it is worth changing behaviour is very difficult before the event, so you then see there is a discounting of the numbers and, as you move, it becomes clear that actually it is not worth it. 

Jon Sherman: When we are looking at doing our risk assessing and we are engaging with companies, for example where we have a case where the claim is this isn’t within the online marketplace category but the other side of the line, then if we had concerns that that was not right, we would want to look into that, get the facts and find out whether we agreed or whether we thought that company should be in DST. 

Q75 Chair: Ms Olney talked about evaluation, but I guess that because this intended to be a short-term situation, there is a limit to what you will do about it. If it went on longer, would you consider doing a deeper dive into whether there had been an overall change of the sort we have been discussing? 

Jon Sherman: As Jim was saying before, we will look at the compliance risks as we go along from our engagement with companies and from the work we do. Obviously, if the tax disappears in two years, then what would we do is. 

Jim Harra: What we would do generally, in this tax as in others, is if we see behaviours that are affecting the level of tax, then that is something we will look into. If it is non-complaint, HMRC will take the action. If it is complaint but we think it is of interest to the policy makers, then we will pass it on. 

Q76 Chair: Okay. It is all a bit odd discussing something that is supposed to have an end to it. By the time we realise all of this, it might well not be around, if 140 jurisdictions get their act together, which I am still a little sceptical about. 

We often talk about the need for transparency. You obviously have your legal duty, Mr Harra, to make sure you manage taxpayer confidentiality, but what was interesting when we were looking at this was that the NAO has been able to look at and report on all the taxes paid by digital businesses. You cannot identify them, but if we sat down, we could probably work out quite easily who are the top five.  

Have you learned any lessons about how to report on a tax like this effectively and manage that? You have a level of transparency that is useful to us, but you also have taxpayer confidentiality. How have you learned from this so it will help us in future? 

Jim Harra: My default is that I want to be as transparent as I possibly can be, but I have to manage the risk of identification. It is interesting that you say that, with a bit of digging, you think you could work this out, because that would be undesirable from the point of view of my requirement to keep things confidential. 

Q77 Chair: There are only five companies that we could probably identify. If we all came up with a list, we would have quite a lot of overlap. 

Jim Harra: While there is some risk involved, I think we have satisfied ourselves with the NAO that, in relation to the anonymised data that they have put in about the DST liabilities and corporation tax liabilities, people would not be able to identify, with research, precisely who these people were. Obviously, a couple of the biggest players have disclosed what they have paid in DST. 

Q78 Chair: I think that they probably talk about it themselves anyway, yes. 

Jim Harra: Of those who have stayed silent, it is possible, using commercial data, to work out that they must be payers of DST, but that is a different matter. My default is that I want to be as transparent as I possibly can be, but I cannot publish even anonymised data if people can then figure out whose taxpayer data this is. That applies, for example, in proactive publication, freedom of information and NAO Reports as well. There is a constant judgment on it. 

Q79 Chair: This is an area that we are all interested in, and it is possibly a shortterm area. This is a very small number; in normal times you would be hard pressed to talk about such a small group of taxpayers and maintain that confidentiality, but you have in this. We all love transparency, but I am a little puzzled as to how you think that you are managing to maintain confidentiality in this case. Could you apply these lessons to other sectors? 

Jim Harra: It is a judgment call. The most conservative thing would have been to say, “We have collected £358 million, but I cannot tell you how many people have paid this”. That is not really a feasible position for us to take. We have worked with the NAO and agreed to the disclosure of as much information as we can to give you the transparency that you need without actually being disclosive in an identifiable way. That is our aim all the time. 

Q80 Chair: Okay. On this occasion, we welcome it. Being the Public Accounts Committee, we always push for more. I am just going to go back to Sir Geoffrey Clifton-Brown briefly. 

Q81 Sir Geoffrey Clifton-Brown: Can I just get clarity on one or two questions I asked previously? Mr Harra, can I take you to figure 12 on  page 31? On the right-hand side there, on the fourth point, it says: “HMRC policy teams were assigned to manage any risk emerging from this. As part of an international agreement, the UK government subsequently agreed that business groups could offset DST paid in 2022 and 2023 against their future Corporation Tax”. I was under the impression that it could not be offset against corporation tax.  

Jim Harra: The UK has reached an agreement with the USA that in the event that pillar 1 comes in, if businesses paid more DST in these years than they would have paid under pillar 1, they can offset that against their pillar 1 liability.  

Q82 Sir Geoffrey Clifton-Brown: Okay. If I could take you, Mr Williams, to paragraph 1.14 on this agreement with the United States: “As part of the agreement, the United States agreed not to impose tariffs on the UK and other countries which had developed their own DSTs. The United States had previously announced that these taxes discriminated against USbased business groups.” This is one and the same issue, is it?  

Mike Williams: Yes, it is. There was an investigation under section 301 in the United States by the US trade representative. This reflects the cessation of that inquiry in the United States.  

Why did we do it this way? If you go back to July 2021, we had reached some sort of tentative agreement between us. It is fair to say that in the United States’ ideal world, they would have seen that that was it—we had agreed pillar 1, so why did we need DSTs, therefore we should stop them at that point. In our ideal world, you would not remove the DST until pillar 1 is in place. The question is—this is what the agreement deals with—what you do about the period in the middle. This offset is a compromise between the UK and the US and the other countries that also have DSTs.  

Q83 Sir Geoffrey Clifton-Brown: To come back to one of my earlier questions, there won’t be a gap, because this is the mechanism you are going to use to solve the gap. There will not be a gap where these types of service companies will not be taxed at all between the ending of DST and the implementation of pillar 1.  

Mike Williams: Yes, that is a fair description of what that agreement is ensuring.  

Jim Harra: It does create a contingent liability, if you like, that we will be collecting DST, which, when we have finally been able to model pillar 1, might actually be available for offset against a future liability, but at this stage you do not have the certainty of that.  

Q84 Sir Geoffrey Clifton-Brown: You don’t know. Okay.  

In a general sense of inquiry, can I take you to paragraphs 3.14 and 3.15 on page 35? This is about international compliance. There it says that if people in other tax jurisdictions do not pay, either under DST or under pillar 1, you can go and talk to those other tax jurisdictions. That begs a number of questions. One is whether those jurisdictions will actually deal with you or want to deal with you. Could you say a little bit about that? It might not even be that they want to; it might be whether they are even capable of dealing with you, if they are not up to speed in this area.  

Jim Harra: We have bilateral and multilateral agreements that include cooperation in assisting each other to collect debts. If there is a UK tax debt, which we cannot collect, but the taxpayer has assets in another country with which we have this agreement, we have the right to ask that country to cooperate with us to assist us to collect that debt. That is reciprocal, and that is something that we are very used to using.  

There is a question about this being a new tax. It is not part of the general run of corporation tax and income tax, so there must be a question about whether all our partners will stand ready to assist us to recover this tax. That is not a problem that we have faced in practice at this point. But the NAO Report rightly points out that while we have a mechanism for doing that, there is a degree of uncertainty—because this is an entirely new tax and a relatively controversial one in the international community—as to whether we will get the same level of co-operation and support as we would normally get. 

Q85 Sir Geoffrey Clifton-Brown: To go back to my earlier question, these types of taxes are displacement taxes. You might tax a company where it is domiciled, in the United States, when actually it is operating in the United Kingdom. But it now might operate in a less sophisticated tax domain. Mr Williams, in relation to the OECD negotiations, bringing those 100 jurisdictions—if that is what it is going to be—into the scope of pillar 1 and pillar 2, how will some of them, particularly with less sophisticated tax regimes, stand to benefit from this? How will they actually operate this relatively sophisticated tax regime? 

Mike Williams: That is clearly a key challenge. Let me go into two aspects of it. We touched earlier on the fact that there is putting money into the pot, for redistribution, and there is getting money out. One way of dealing with this for relatively small countries is to say that there is a de minimis level that means that if, basically, your contribution to the pot will be very small, you don’t have to bother, because it makes no difference in the grand scheme of things. Equally, the proposition is that the tax administration for the group as a whole—if it’s a US-based group, it will generally be the IRS; if it’s a UK-based group, it will be HMRC—would take the lead in doing the work to work out how much money country 96 should get from being involved in the deal. Obviously, country 96 will have its own audit office, which will then want to check itself that it is getting the right share. And of course we have to build in things that cope with that. But that is broadly how it will happen. 

Jon Sherman: May I add to that, Sir Geoffrey? The OECD’s Forum on Tax Administration has a capacity-building network. We chair that, and we also have our own UK capacity-building programme. One of the priorities for the focus of that work is going to be how we can best support countries that have less developed tax authorities to implement and benefit from—

Q86 Sir Geoffrey Clifton-Brown: Ah, so you are actually giving some of those less sophisticated tax regimes advice on how to build their capacity. 

Jon Sherman: That’s the plan. We work with a number of countries anyway in terms of bilateral support on a number of areas for developing their tax administrations. At the moment, we are looking at how we can provide what is more like one-to-many support, through forums and networks where people can get advice, share problems and so on. Implementation of pillar 1 and pillar 2 is going to be a priority focus for capacity-building work. 

Q87 Chair: Before we go to Mr Djanogly, I am interested, Mr Williams, in your almost throwaway line there: “We have to find ways of making sure that we get the measurement of the tax receipts across the different jurisdictions right and that country 96 has got its fair share.” Can you give us an idea of what those mechanisms will be—Mr Sherman may have just hinted at it—and how you get that international co-operation so that you are not having a company playing one country off against another? In the end, the compliance issues are going to be enormous across all those different jurisdictions, aren’t they? 

Mike Williams: I think that’s true. However, you may say that country 96 will be relatively powerless in relation to country 1, in which the group is based, but countries 2, 3 and 4, in which that company operates, are probably not powerless vis-à-vis country 1. Equally, we have said there has to be sufficient data. We have a National Audit Office, which has to be able to satisfy itself. Of course, some of the businesses who, for efficiency reasons, would rather produce less data look a bit glum at this point, but then we have to say it again. 

Chair: It’s so lovely being a head of HMRC! You just say it again and then everyone has to fall into place. 

Mike Williams: There is a danger that some businesses say, “But we don’t want to provide this level of access. We want to give access only to the tax administration in the country in which we are based.” Of course, the answer to that is, “We do have a National Audit Office. It has obligations”— 

Q88 Chair: From that, you are suggesting that you are going to put some of that burden on to the businesses to do it. 

Mike Williams: What we cannot do is say that there’s this black box—this goes back to Sir Geoffrey’s point—the tax administration of country A works with a business in country A, and the black box just produces numbers, because how do we know that they are the right numbers? There has to be a way, if necessary, of going in and auditing those numbers. I think that is the crux of it.  

Q89 Chair: Would country 96 have access? 

Mike Williams: I don’t know whether this would be true in the case of country 96, but we have to assume that they have a body, equivalent to the NAO, that needs some access. Of course, in the nature of these things—this bedevils all international things—each country will have different arrangements, and we just have to do what we can to reflect that.  

Chair: I can see this being very complex, and we know that the next stage is. What we are talking about today is relatively simple by comparison.  

Q90 Mr Djanogly: I just want to have a look at the effect on business. I am going to refer to some written evidence that we got from Expedia Group, which is a worldwide online travel agency. It is currently paying DST and made three main criticisms, but I will mention one of them, which is that the tax is unsustainable for businesses with low operating margins. They say that “in the online travel sector…companies undoubtedly bring value into the UK but in many cases have operating margins under 10%. The existing ‘Alternative Basis of Charge’ regime provides some protection for the lowest margin companies, but this is not sufficient in its current form when you consider that under the current rules companies making over a 2.5% operating profit margin are expected to pay a 2% tax on their revenue, effectively wiping out their UK earnings. This in turn reduces the economic interest to invest in attracting travellers to the UK as compared to non-DST destinations”. Do we have a problem here in terms of making the UK less attractive? 

Mike Williams: It is back to the DST being the second-best solution. The best solution would be an updated means of dividing up profits between countries, because it is better to tax profits—I think that example tells you why. That then leads to two issues. One is the long-term sustainability of this tax as an interim solution. Candidly, if countries equivalent to the UK also introduce DSTs, as several have, you are reducing the impact of that competitiveness angle. But it also impacts the rate, as we talked about earlier. In your example, Expedia described the difficulties where you have a relatively low margin. It well illustrates that you could not have a very big rate, because you would, in effect, be taking more than there would be profit, but businesses cannot exist on that basis. It also tells you that you cannot go for a very high rate at all in those circumstances. The crux of the opposition is exactly that: that a turnover-based tax is insensitive to the margins of those to whom you apply it. At the margins you can flex, but you cannot turn a turnover tax into a profits tax, if you see what I mean.  

Q91 Mr Djanogly: I think you are saying that this tax is not sustainable as a longterm proposal.  

Mike Williams: The question is how long-term it is. I think that is the very difficult question.  

Q92 Mr Djanogly: Okay. Have you received many indications from traditional retailers of what their views are on this tax? More specifically, is taxing 18 groups of companies going to lead to the survival of the British high street?  

Mike Williams: This is a tax on services, not on goods. Some people running high street businesses have suggested that we ought to rebalance things between them—working from stores on the high street—and online sellers of goods. A key concern of those businesses has been addressed in the business rates revaluation with the new rates that are coming in from April 2023. If you look in detail—you can go on the VOA’s website and see the rates for individual businesses—you see that the rateable value of many high street businesses has gone down very significantly as part of the revaluation.  

Q93 Mr Djanogly: They may be happy with the revaluation, but they are still paying rates, whereas these businesses aren’t.  

Mike Williams: That is true in relation to people selling services, but of course most of the people operating on the high street are selling goods. To sell goods online, you tend to need extensive warehousing, because people are not prepared to wait very long. I do not know about you, but I like the goods to come within a day or so. They cannot come from some offshore island if they are going to do it on that basis. If you look into this open website, you see that these warehouses tend to face higher business rates. 

Q94 Sir Geoffrey Clifton-Brown: Can I come back to the issue of displacement tax? Is there a list of those countries that have introduced a DST? 

Mike Williams: Inevitably, it depends on what your definition of the DST is. The classic examples—(Interruption.)

Chair: Apologies. We are wondering what the bell is. No one has died, I  do not think. That is usually what our worry is. Sorry, Mr Williams. Carry on.  

Mike Williams: The classic examples are the ones, referred to in the report, that reached agreement with the US: the UK, France, Italy, Spain and Austria. The other example that is widely cited is India, which has something called the equalisation levy—it is not called the DST. Turkey also has one that is referred to. 

Q95 Sir Geoffrey Clifton-Brown: This is what I half suspected. Given that this is an emerging area of taxation to deal with a very fast emerging digital scene, is there an issue with the west—broadly speaking, the countries that

you talk about are in the west—doing one thing and Asia doing something different? Do countries such as Japan, South Korea or the Philippines—you could go on—have a DST, and will they be signing up to pillar 1 and pillar 2?  

Mike Williams: Japan does not have a digital services tax. Japan is very actively engaged on the work on pillars 1 and 2 to an equal extent to the UK. 

Q96 Sir Geoffrey Clifton-Brown: I used Japan as an example. Asia as a whole— as a bloc—is where a lot of this digital stuff is emerging very fast, and some of their companies are emerging very fast, and presumably therefore making significant profits. Are they fully engaged with the OECD process on pillar 1 and pillar 2? 

Mike Williams: Yes, I would say they are. You can see from the list of names in that case. You could argue that a difficulty is that the DSTs that countries have introduced do not all have the same design, and that is something that impacts on business compliance costs, which inevitably they complain about. They do not all have the same design, because it was not possible to achieve consensus through the OECD on what a common design would look like—no doubt because some countries thought, “If you put a common design out there, more countries will implement it,” and some countries do not want that. 

Q97 Sir Geoffrey Clifton-Brown: Surely a common design is what we are moving towards with pillar 1 and pillar 2, so that there will not be any sort of displacement because it will be implemented around all the major jurisdictions. 

Mike Williams: That is definitely the idea, yes, which is why there is impetus to reach agreement on pillar 1. That is why—back to the question about timetable—there are reasons for optimism about the timetable. 

Jim Harra: But to be clear, pillar 1 is not moving towards a consensus on the design of a digital services tax. It is a means of attributing profits in a different way to jurisdictions to tax, and that will replace the digital services tax. 

Q98 Sarah Olney: Let me reflect on the wider situation. Obviously, the DST is quite an unusual tax inasmuch as it is a turnover tax. I do not think that we have that; I might even go so far as to say that we have not had that before. What lessons have you learned from the implementation of that? Should there potentially be future demand for other turnover taxes? What have you learned from this that you might be able to apply? 

Mike Williams: You have to factor in, first of all, that you have VAT.  

Sarah Olney: Consumption tax though? 

Mike Williams: But it is based in nearly all circumstances—not all, but in most circumstances—on turnover. With things like the zero rating, we exclude some turnover, but where it applies, it is based on turnover.

Running one turnover tax alongside another, if we are outside niche areas, would be a struggle—how would we make them interact properly? 

Q99 Sarah Olney: Okay. We have already touched a lot on the day, whenever it comes, that the DST is replaced with OECD pillar 1, and we have talked about how that might impact on revenue. In effect, it is very different. How will that transition impact on the businesses affected? 

Mike Williams: The businesses affected by pillar 1 will have to gear up to pillar 1. Sir Geoffrey asked about that earlier. Equally, if they are already affected by the DST, they will not in future also have to deal with the UK DST or the DSTs of all the other countries that have been introduced. 

Jim Harra: In operational terms, for HMRC, obviously there is a commitment from the UK that it will shut down its DST when pillar 1 is implemented, so our costs and the taxpayers’ costs would stop at that point. New compliance costs would come in, in relation to pillar 1, but also companies might well ask to do a pillar 1 computation for the two preceding years, to compare with their DST liability, to see if they have any right to a set-off of excess DST under the agreement that we made with the USA. That would be a matter for them to decide—whether they think that is something that would be relevant to them and they want to bear the costs of doing that. 

Q100 Sarah Olney: Once the OECD has reached final agreement, what is your estimate of how quickly then it will be implemented? 

Mike Williams: That depends on how quickly the most relevant countries move beyond signing the convention to adopting it. The big question, because this requires the US to be one of those countries, is how quickly the US moves. That has always been a key issue. 

Q101 Sarah Olney: So 2025 is when we think we might get the agreement, but in practice it could well be more years before we start to implement pillar 1. 

Mike Williams: If the convention is ready on the OECD’s timetable for signing in mid-2023, then implementation in around 2025 is pretty feasible, but the question inevitably is how quickly the US moves. 

Q102 Sarah Olney: That is a dependency for us, in that we are not going to implement until the US has. 

Mike Williams: Yes, and that, of course—back to Sir Geoffrey’s question about the agreement with the US—is one reason why we did not want to stop our DST at the point that we agreed to pillar 2, because, put crudely, you are reducing the incentives to ratify the pillar 1 agreement. 

Q103 Chair: May I just check, because I cannot remember—I think Mr Sherman may be the one to answer this. There is a hope that the legislation will wither away and we will have the new system, but do we have to pass primary legislation at any point to continue it?  

Sir Geoffrey Clifton-Brown:  It is sunsetted.  

Chair: Exactly; it is sunsetted, but— 

Jon Sherman: It is not sunsetted.  

Chair: That is why—I just want to be clear about this. 

Jon Sherman: The DST will not fall away, but there is a provision in the legislation, I think, that provides that we will have that review.  

Q104 Chair: And that review does not require legislation. The Treasury must produce a report by the end of 2025 and that will be laid before Parliament

Jim Harra: My understanding is that the DST legislation is not sunsetted, so it continues unless and until it is repealed.  

Q105 Chair: To be clear, we would have to repeal it. 

Jim Harra: Yes. Pillar 1 will require legislation, and in that legislation would be the repeal of DST. 

Chair: Okay, so we would have to repeal that, so that will be a discussion for whoever is in government after 2024. Okay, right, so lots still to do.  

It has been a fascinating session because this is an interesting tax, but the bigger picture is this big, challenging, international one, which we have been discussing since I joined the Committee 11 years ago. That perhaps gives an indication of how long these things take to deliver, but we are watching it closely and will continue to do so.  

Thank you very much indeed for your time. The transcript of this session will be put up on our website, uncorrected, in the next couple of days. Thank you to our colleagues at Hansard, and thank you to our witnesses again.