HSBC and HMRC: what can we learn from Ireland

Last week, the Chairman of the Irish tax authority wrote to its Public Accounts Committee. You can read the letter here.

For students of HMRC’s handling of the Falciani HSBC disclosures, the letter makes for illuminating reading. The similarities between Ireland’s tax system and our own invite comparisons between how the Irish tax authority handled the disclosures – and how we did.

There seem to me to be five potentially interesting points of comparison.

1. Yield. I have written elsewhere about the comparisons to be drawn between us and our continental neighbours on tax yield from HSBC accounts. UK received information on about 6,000 individuals and businesses and recovered tax and penalties of £135m. France and Spain – both with fewer billionaires than the UK – have recovered £188m from 3,000 and £220m from 3,000 respectively.

It is fair to point out that there is no direct read-across from the yield figures for France and Spain as neither has a non-dom rule. It is probably fair to say, as a generality, that money held by UK residents abroad is less likely to give rise to a UK tax liability than money held by Spanish residents a Spanish one. Ireland has a non-dom rule – but sadly the letter does not give a comparable yield figure.

2. Criminal prosecutions. There were 88 individuals with Irish addresses. 20 were considered for criminal prosecution (23%), four were selected for prosecution, three were convicted (3.4%) and 1 remains under investigation. In the UK, there were 3,600 individuals of whom 3,200 were traced, 150 were considered for criminal prosecution (4.2% or 4.7%) and 1 conviction (so far) has been obtained (0.03%).

So the Irish have been spectacularly more successful than us in achieving criminal prosecutions.

3. Does the French Defence stand up? The absence of criminal prosecutions has been blamed by HMRC and the Treasury Minister on constraints imposed by the French on the use to which we could put the Falciani disclosures. Initially both were rather tight-lipped on what those constraints were. However, we know now that the information was disclosed under the terms of the UK France Double Tax Convention Article 27 of which limits its use to tax matters (including tax evasion offences). However, the information was also disclosed to the Irish under the terms of their Double Tax Convention with France (and the EC Mutual Assistance Directive) which contain the very same limitations. It hasn’t stopped the Irish obtaining convictions.

So this evidence suggests we were not justified in blaming the French.

4. What effect did the Falciani disclosures have on the use of amnesties? The point of an amnesty is to encourage people you might not otherwise find out about to come forward voluntarily and put their tax affairs in order. Why would you offer an amnesty to those you already knew about?

I have written here about our misuse of amnesties. But, in fact, the situation is even worse than I then appreciated.

Lin Homer said this to the Public Accounts Committee:

I think that, on previous occasions, we have told you that we thought that there were possibly about 15 or so that we hoped to get through to criminal prosecution. As I have told you on a previous occasion, a number of those moved into the disclosure facility and took themselves out of prosecution that way. A number were discussed with the CPS, and, in the end, of the three that we and they felt most likely to be able to prosecute, they felt that only one had reached the test. That is the top tier.

Quite why those actively being investigated with a view to criminal prosecution should have been permitted to ‘take themselves out of prosecution’ through use of the amnesty I do not know. Certainly this was appreciated by the Chair of the Irish tax authority who said:

The commencement of a Revenue investigation means that the taxpayer is precluded from availing of a qualifying voluntary disclosure.

5. Different approaches to transparency. But to me the most striking thing of all about the letter to the Irish Public Accounts Committee is its tone. The letter demonstrates that the Chairman of the Irish tax authority understands that transparency is important. Our own HMRC, I am afraid, does not. In its dealings with the Press, Parliamentarians, and the public at large it routinely seeks to deflect scrutiny and to discourage accountability.

There is not a day when we, the public, do not read stories of legal avoidance and criminal evasion by powerful corporates and wealthy individuals. Those who have suffered in consequence of the squeeze in public finances are entitled to ask of HMRC, are you policing the line? HMRC’s disinclination fairly to answer that question is highly corrosive of public faith not just in HMRC itself but in democracy and society at large.

Follow me @jolyonmaugham

29 thoughts on “HSBC and HMRC: what can we learn from Ireland

  1. Incidentally, there is a very well-researched paper on this very issue by a whistleblower, featuring the PAC and HMRC and entitled “Transparency and Accountability of Tax Administration in the UK: The Nature and Scope of Taxpayer Confidentiality” here
    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2056297

  2. Thanks. I wasn’t aware that Mr Mba had written a paper on the subject. I will have a look.

  3. You must have touched a raw nerve with the excellent point about Lin Homer and David Gauke using the purported French “restriction” to mislead Parliament and the public continuously. It’s interesting that your blogs have been much more effective in illuminating this issue than two Parliamentary sessions and all the media coverage. Gauke first suggested on Today and then in the UQ in Parliament that the purported restriction precluded any action against HSBC and Homer repeated the same thing before the PAC. However, she was clever enough to say that the data was received for “tax administration” or “tax purposes” but not for “money laundering”. Clearly, the former includes prosecuting HSBC and their staff for conspiring to cheat the public revenue by aiding tax evasion but Homer implied that the only option open to HMRC was to pass the data to other agencies like the SFO or FCA for money laundering. Sadly, she and Gauke continue to get away with this obfuscation. Interestingly, John Mann asked her for a record of the discussions they purportedly had with the French on the matter but she quickly turned to the more assured Edward Troup who provided an excellent Sir Humphrey and in the event the obliging Andrew Tyrie opened the door for HMRC to wriggle out of the issue citing “confidentiality”.

  4. Jolyon,
    Thanks for raising these important issues, especially on the process by which HMRC undertakes criminal investigations with a view to getting the CPS to take up and prosecute.
    A brief comment on the Mba analysis. HMRC did not accept this when the PAC advanced it, and the PAC, who received the whistleblower’s disclosures, did not get any independent legal opinion to support that analysis and so obtain further information. The Irish letter also states that taxpayer confidentiality is “fundamental”, with breaches a criminal offence (as in the UK).
    On Jolyon’s main points;
    2. The vagaries of statistics are such that I don’t think you can really draw much from the different rates in the UK and Ireland. It would be as illuminating as saying that other countries with no convictions are even worse than the UK and Ireland. It’s not as if Country A is convicting dozens and other doing zero. Without access to the specific cases I think we can’t really draw many conclusions on the numbers we have available.
    3. I believe the issue on the terms of any disclosure by the French were not whether the tax authority could prosecute; it was whether they could share data with other law-enforcement agencies, either for possible prosecutions to do with things like money laundering, or to corroborate the HSBC data. The Irish letter states they regarded the information as allowing for a tax criminal prosecution; they do not say if it was shared or checked with other law enforcement agencies (like the police). They also noted the Swiss refused to help in their enquiry, regarded the data as stolen, and would not co-operate in any investigation.
    As I understand it the corroboration point was important given the origin of the French data as “stolen/whistleblown” data.
    4. I agree it would seem odd that a taxpayer could take themselves out of prosecution. But I wonder if this was a misphrasing in oral testimony? HMRC’s written statement was that “In 150 of these cases, we sought to collect evidence for criminal prosecution. To do this successfully, we needed to demonstrate criminal intent (rather than error, for example). In addition, because stolen data is considered ‘dirty’ it needs additional corroborating evidence.
    With these tests of evidence, and with the exemptions arising from the Liechtenstein Disclosure Facility, we could only prepare three cases for submission to the Crown Prosecution Service. Having examined the evidence, the CPS considered only one case to be strong enough to take forward, and that was successfully prosecuted in 2012.”
    (https://www.gov.uk/government/news/statement-by-hmrc-on-tax-evasion-and-the-hsbc-suisse-data-leak)
    So the questions go back to whether the LDF was needed. Pinsent Masons have a good summary of what it means in practice, http://www.out-law.com/en/topics/tax/tax-litigation–disputes-/the-liechtenstein-disclosure-facility-ldf/
    It may be that it too was prompted by worries over the nature of “stolen” information that HMRC had paid for, evidential issues of corroboration, plus its use as part of a wider amnesty to encourage disclosure where there might be no evidence of wrongdoing.
    5. I agree the tone and content of the Irish letter is different from the oral report to the PAC. The Irish letter seems to be a response to some specific queries from their Committee. As far as we know our PAC have not asked similar questions (or if they have it has not been published). I’d like to think HMRC would respond in a similar way to the Irish, but I suspect there is now a great deal of mistrust between both sides and that normal relations have broken down. The tone and exchanges at the PAC are surely proof of that, so that attempts to answer (e.g. on the non-dom position) were almost talked over.
    HMRC offers not just the LSS it also offers Alternative Dispute Resolution. Perhaps both parties (the two H’s) should be encouraged to take advantage of it, so that we return to a system of Committees genuinely holding HMRC to account and HMRC being more forthcoming and pro-active in its reports and explanations?

  5. It’s interesting that the fines levied on the three successful Irish prosecutions are only €4k, €10k and €25k. This seems to make those cases relatively small beer, whereas one might have expected the prosecutions to occur in the bigger cases. The costs of the prosecutions would seem to outweigh the fines, for one thing.

    The UK Litigation & Settlement Strategy would suggest not prosecuting in small cases where it’s not cost effective. Do you know what the Irish equivalent (if there is one) says?

    And do we know what the amount at stake in the successful UK prosecution was?

    Whether the LSS approach is a good one or not is, of course, a wider question 🙂

  6. Andrew, The Irish convictions were at Summary level for incorrect returns where the courts normally rarely give custodial sentences (not just for tax, the lower courts have tended to make up their own law involving poor boxes and specific community services e.g. deliver three loads of turf to the local SvP).

    Revenue (DPP) tend to proceed on indictment only where there is a significant history e.g. Michael Lowry TD at the moment.

    Because it is summary there is little issue with cost, but it allows publication on the defaulters list under a separate heading allowing the press to report on it (which from this debacle is actually the key benefit).

    The interest and penalties significantly outweigh the tax in almost all of the cases, the interest alone being the biggest component in many cases at over 8% per annum.

    Andrew Gallagher as president of the Irish Tax Institute called for a reduction in the penal interest rate on underpaid tax on Friday night at the annual dinner. I am not sure that his timing was ideal when the interest rate is the major punishment here (along with the potential co-operation only penalties).

  7. HMRC did not accept Mr Mba’s analysis when the PAC tried to use it to compel them to account for the sweetheart tax deals but HMRC themselves used that very analysis to fight off Patrick McKenna’s JR two years later in Ingenious Media Holdings Plc & Anor, R (on the application of) v HM Revenue & Customs. http://www.bailii.org/ew/cases/EWHC/Admin/2013/3258.html

    So, the PAC may not have obtained any independent legal opinion but we now have a case law and HMRC’s admission on the matter. It’s this sort of hypocrisy that undermines HMRC’s credibility. Hartnett claimed he would go to prison if he discussed Vodafone or Goldman Sachs with the PAC but was quite happy to say all sorts of things about McKenna and Ingenious to journalists from The Times.

  8. Thanks Tyler. If anyone’s interested (and you should be) in the Ingenious judgment – and what it says about HMRC’s attitude to transparency – you can read my post on the subject here http://wp.me/p3D1eN-gg

    I will address Iain’s points when I find a moment (hopefully later today)

  9. Thanks for the link to the blog (where the Mba paper is referenced) and for any follow up on the other comments
    But to pick up on sweetheart deals and Ingenious, I don’t think HMRC is necessarily being hypocritical or contradictory in this. The Mba disclosures were all to do with taxpayer case specific information. They involved the basis for tax, possible sources of profits, their amount, the level of tax payable, possible international issues and double taxation. HMRC’s position was they could not disclose to the PAC that level of detail without breaching CRCA. An independent judicial review took place, with access to the facts, and that did not find any evidence of the alleged sweetheart deals. It found weaknesses in governance (and the failure to claim interest on a portion on one settlement). HMRC accepted the criticisms and changed the way it controls such cases.

    The Ingenious decision did not support the disclosure of any information about the taxpayer’s tax affairs and was to do with policy views. As the judge said it was a very limited disclosure, and justified. “In my view, the attitude of disapproval and scepticism which HMRC adopted in relation to use of film investment schemes, and the decision to seek to inform the public about that attitude in order to influence their behaviour, fell well within the wide margin of appreciation applicable to decisions of state authorities in the field of tax policy.” (Para 75)

    But he specifically drew a difference between this and the disclosure of facts about personal tax affairs. “Thus, for example, it would have been a matter for grave concern and close scrutiny by the court if Mr Hartnett had given the journalists (even in an “off the record” briefing) access to the full tax files of Mr McKenna and Ingenious Media, with all the private information which they had supplied to HMRC about themselves, even if he did think that such a course might help in some way with the collection of tax.” (Para 65).

  10. Thanks Aisling, that’s interesting. Am I right in thinking that one could get civil penalties for incorrect returns in the absence of summary criminal judgment? If so, why do some cases go the criminal route and some the civil?

    There doesn’t seem to be a UK tax equivalent to the summary criminal route.

    Perhaps a close UK equivalent would be the recent naming and shaming in connection with National Minimum Wage, though that was just publicity for civil offences.

  11. Yes, civil penalties are under s1077E but If a taxpayer disagrees with e.g. a “co-operation only, material tax loss” penalty you have to challenge it directly to the High Court (not the Appeal Commissioners).

    s1078 is the offence, and the penalties under s1078 (other than on indictment) are small potatoes relative to the normal tax geared penalties and interest under s1077E so Revenue don’t often prosecute under it.

    http://www.revenue.ie/en/press/prosecutions/index.html

    We live in an odd little world in Ireland where Revenue know that a significant number of their powers are on weak enough constitutional grounds were they to use them in other than the most egregious cases. By deferring to the courts Revenue hope to mitigate this and pre-empt a constitutional challenge.

    You can end up on the defaulters list here for a careless error on the return as distinct from deliberate (if material & co-operation only once settlement exceeds €33k), but you can make the same list if a court imposes a €5k penalty.

    If one compares the number of s1078 prosecutions on the defaulters list against the s1077E penalties defaulters list where the level of penalties suggest deliberate behavior, the obvious conclusion is that Revenue are not prosecuting all or anything like it.

  12. Iain,

    For some reason, you’re speculating on the type of information HMRC withhold from Parliamentary Committees. HMRC’s letter to the PAC dated 19 Oct 2011 “setting out HMRC’s legal view in relation to the disclosure of taxpayer confidential information to a Parliamentary Committee” contains a blanket rejection and is very different from the argument you’re now advancing on their behalf …
    http://www.publications.parliament.uk/pa/cm201012/cmselect/cmpubacc/1531/1531we02.htm

  13. And here’s HMRC’s written statement to the TSC showing how the blanket policy works in practice … refusing to answer “whether the internal procedures were followed”
    http://www.publications.parliament.uk/pa/cm201012/cmselect/cmtreasy/731/731we14.htm

    6. Q413: Mr Umunna: Now, in relation to Goldmans, there are serious allegations which have been made in the media in relation to HMRC settling this case and also in relation to Mr Hartnett in particular. Would you consider as an organisation publishing or providing to us information about that case so that the public can be assured that the proper procedures have been followed?
    7. Q415: Mr Umunna: Could you also tell us in relation to the particular case that I have raised whether the internal procedures were met in relation to the Goldman’s settlement?

    HMRC has carefully considered the extent to which they can answer the questions asked and have concluded that they cannot give any information, for reasons of taxpayer confidentiality.

  14. I’m not trying to defend HMRC just to be contrary, only pointing out that I think it is very important that the UK’s tax authority is not willing to hand over what I regard as deeply personal data (which could be mine) without legal authority. Not only that, but I would not want to see its staff potentially exposed to sanctions for following/not following instructions to disclose, including possible prosecution.
    In reply to Tyler’s response I hate to get into “I said/you said” but I feel I’ve got no alternative here.
    Far from advancing arguments HMRC did not, I have actually paraphrased the arguments they have publicly put forward. The material in your first link to the PAC is explicitly and consistently about whether or not HMRC can disclose taxpayer confidnetial material to the Committee. There is no “blanket rejection”. Instead, CRCA is set out in detail, and confidentiality is a key theme in the opening analysis. The very first consideration was “the legislation provides a strong regime for ensuring the confidentiality of information in HMRC’s hands—particularly where the information in question relates to an identifiable person’s tax affairs. Section 18(1) prohibits the disclosure of information.” That analysis concludes “There is no express power in the legislation to make disclosures to Parliament.”
    So, the next stage of the analysis considers “whether disclosure would fall within the exception in section 18(2)(a), which permits disclosure for the purpose of HMRC’s functions.”
    There then follows a discussion of what is permitted, refers to advice from Counsel, and a discussion at Excom. They decided “to maintain the confidentiality of information which identifies and relates to specific taxpayers in responding to questions from the Committee.” It also referred to obligations to Ministers, to OGDs, to public reputation for impartiality, that it would be “inappropriate and unfair to create any real risk of exposing officials to criminal sanction in the course of their duties,” and that information given to the NAO was held under legislative control over confidentiality.

    As for the TSC, Q413 is clearly asking for information on the facts of the Goldman Sachs case, which is presumably covered by the opinion given to the PAC (above). As for Q415, I do not know the basis on which HMRC argued a refusal was covered by the non-disclosure analysis. I could construct a line that says a simple reply, assuring the TSC that internal procedures were met, would probbaly not be received well by the Committee who might feel they were being fobbed off, without the “information” requested in Q413. Or HMRC might argue that Q415 was contained within Q413, so there was no need to answer it separately.
    As far as I know the TSC accepted this answer and did not return to it. Their Report said they did not challenge the need for taxpayer’s affairs to be kept confidential (para 160). But they said that settlement of large cases needed to be reviewed and they would take further evidence on that.

    Click to access 731.pdf

  15. By the way all, the Ingenious case has been appealed and judgment comes out tomorrow morning, 3 March 2015, from the Court of Appeal

  16. Of course, it could be that the reason there are more Irish prosecutions is that the British are more honest than the Irish.

    It’s just a thought.

    🙂

  17. HMRC have got £135m that says otherwise!

  18. Honest mistakes? Tax is just SO complicated these days.

    I am joking, of course.

  19. I know 🙂

  20. Playing with numbers, I notice that if you take the amount recovered by the UK and Ireland and divide by the numbers of taxpayers on the list, the amount per head comes out roughly the same.

    UK: £135m from 3,600 heads is £37,500 each.
    Ireland: €4.5m from 88 heads is €51,000 each, or £37,250 at €1.37 to the pound

  21. only if you assume the 135m and 4.5m relates only to individuals…

  22. Has anyone seen the Ingenious judgment?

  23. Thanks. It was on the cause list for today.

  24. I can’t find an appropriate place for this comment, but in the context of a discussion about fairness and enforcement action, have you seen http://www.bailii.org/uk/cases/UKFTT/TC/2015/TC04287.html

    HMRC is shamelessly trying to assess poor Mr Fessal for the same profits twice (issuing a discovery assessment for an earlier period, when he was out of time to make a repayment claim for a later period, when tax has alreayd been paid on the same income).

  25. Thanks. Look forward to having a proper read.

  26. They’re both totals which seem to cover all account holders. Perhaps “per head” is the wrong phrase, but the amount “per taxpayer” is still comparable.

    And surprisingly low, to be honest. There must be a lot of smallish accounts there – or maybe it’s just that the bigger ones are taking longer to finalise (or are untraceable).

Comments are closed.