.png)
.png)
Time for greater transparency by multinationals over their use of tax havens
KEY POINTS
- Taxation’s weekly round up of tax stories on our blog.
- Fair Tax mark criticised as flawed, methodology under review.
Each week, normally on a Friday, Taxation’s online editor, Daniel Selwood, uploads up a blog entry titled News briefing. It is a summary of stories about tax in the national press, together with a link to the online versions of them and a short comment from the Taxation editorial team, which tries to get to the reality behind the spin.
The idea is that the press cover a wide range of tax stories (often very poorly) that we do not deal with in the magazine because they are not technical enough, but your clients will expect you to give a coherent answer when they plonk a page they have cut out of the Daily Mail in front of you and say, “What’s all this about, then?”
A useful tip, when put on the spot in this way, is to look at the last few paragraphs of the story. This is where the writer and editors normally hide the inconvenient facts that would undermine the shock and horror expressed in the headline and opening lines. However, an even better tip is to read the blog item each week, so that the answers are at your fingertips.
Most of the blog entries in recent months have contained one or more items about large company tax “avoidance”. Most of our comments have been of the “reporters have no idea what they are talking about” variety.
Although the whole point of the blog entries is to deal with tax stories we would not normally cover, the level of interest in tax is currently so high that it seemed appropriate to summarise some of them here, and then comment on the overall picture.
Fair Tax Mark
Let’s start with the Fair Tax Mark. Most of the papers covering this story seemed to take their basic information from Reuters, where the report began with:
“British retailers have been accused of not paying their way on tax, with just two of 25 well-known store groups awarded a ‘Fair Tax Mark’, in a report published on Thursday by the Fair Tax Campaign.”
Most readers would assume that the other 23 companies had paid less tax than the law provided. In fact, the campaign does not claim that. Since the driving force behind it is the well-known campaigner Richard Murphy, it reflects his concerns about full country-by-country disclosure of tax paid, that companies should broadly suffer tax at the headline rate on their accounting profit when averaged over a number of years, and that they should not have subsidiaries in tax havens unless they disclose their profits in full.
It is, therefore, perfectly possible for a company to have engaged in no aggressive tax planning whatsoever but to get a very low Fair Tax Mark by a combination of failing to make disclosures in excess of what is required by GAAP, and making proper use of deferred tax provisions.
Much of the detailed criticism of the Fair Tax Mark has been about the methodology for the calculations of whether a fair tax rate was paid; specifically concentrating on the effect of not weighting for different levels of profit while giving a very high weighting to later rather than earlier years in the six-year period considered.
These were comprehensively undermined in several posts by Ben Saunders over June and July on his blog. (Ben is a colleague at LexisNexis, but was writing personally).
However, my main objection was that the methodology meant the top mark went to Greggs. Greggs?! A company which, in my view, spent the best part of last year trying to avoid paying the correct amount of VAT on hot take-away food?
The Fair Tax Mark does not look at anything other than corporation tax, and because Greggs is a purely UK-based company it gets full marks for country-by-country reporting (UK, 100%...) and for not using tax havens.
A company like WH Smiths, on the other hand, gets the worst score (2/13) because it reports its income split between high street and travel (airports, etc) rather than by splitting it between countries, and also because it has Jersey and Ireland subsidiaries, which the company explains are to hold its retail operations in those countries.
To give credit where it is due, Fair Tax Mark has now announced that it is reviewing its methodology and has invited those who have criticised it to comment.
I can see how it might be reworked into a tax transparency mark, specifically rewarding companies that offer more information than they are legally obliged to (more on that later); I cannot at the moment see how the calculations could be reworked to provide a meaningful score.
PAC-ked hearings
Elswhere, the Public Accounts Committee (PAC) continued its policy of not listening to witnesses, not trying to understand the complexities of tax policy, and insisting that commonsense was all that was needed in order to decide whether a company was fiddling its taxes.
Google, in particular, came in for a roasting (as opposed to Starbucks) about the structuring of its commercial operations. The PAC issued its report into the company on 10 June, and attracted a great deal of media coverage.
What none of the press pointed out was that the report was really very thin for the amount of time the committee had given to it – a page of conclusions and recommendations (which, entirely unsurprisingly, precedes the evidence on which it is based) and a page each summarising and commenting on the evidence of the auditors Ernst & Young (as it then still was), the company, and HMRC.
Since they are not tax specialists, the committee members were, understandably, confused to find that Google maintained a significant number of staff in the UK to advise and encourage UK customers to spend the $18bn that they did over five years with the group, and yet only £10m of corporation tax was paid.
The sensible approach would have been to ask for some independent advice from tax specialists about how this could happen. Instead, they insisted that this constituted “sales” in the UK and that Google must therefore have a permanent establishment here, and that the only thing happening in Ireland was “billing”.
Admittedly the committee had seen documents from a whistleblower, which it said proved that the statements from the company were untrue. However, the detail of what it says the whistleblowers provided serves only to show the problems of a committee trying to make decisions about tax based purely on common sense:
“The whistleblowers, who included ex-employees of Google Ltd, provided the committee with details demonstrating that Google Ltd’s UK staff carried out the substance of work leading to contracts with major UK clients. They also gave us a range of evidence to support their assertions.This included, for example, pay slips showing sales related bonus payments, and Google documentation covering the entire trading and sales process within the UK. The evidence presented to the Committee included a Google diagram of the sales process interaction with UK clients, all of which were executed by UK-based staff. The whistleblowers told us that UK staff had been set sales targets and paid commissions for the sales achieved. This evidence showed that Google Ltd’s UK staff were carrying the substantive work to generate the revenue from Google’s presence in the UK.”
Permanent establishment
The problem is that most of this does not prove that the Google office in London is a permanent establishment for the purposes of the Ireland/UK double tax agreement and, if there is no permanent establishment, there is no charge to tax on the profits generated by the sales.
Article 5 of the agreement, while not expressed in the terms of the corresponding OECD Model Agreement article, still contains the same basic principles.
A place of business in the UK which is there for the purposes of advertising, supply of information or “similar activities which have a preparatory or auxiliary character” is not a permanent establishment unless an employee there:
“Has, and habitually exercises in that State, an authority to conclude contracts in the name of the enterprise, unless his activities are limited to the purchase of goods or merchandise for the enterprise”.
In more than 90% of sales by number, although only 30-40% by value, that is clearly not the case, because the most important part of the contract – the price – is set by online auction through the software on the Irish servers.
For the others, the question is whether the contracts are concluded in the UK or Ireland with the proviso, as the commentary to the OECD Model Agreement puts it, that:
“A person who is authorised to negotiate all elements and details of a contract in a way binding on the enterprise can be said to exercise this authority ‘in that State’ even if the contract is signed by another person in the State in which the enterprise is situated.”
However, the fact that staff in the UK were part of the sales process, had the word ‘sales’ in their titles, were recruited for their sales skills, and were paid commission on the sales which eventually resulted from the clients with whom they worked will not necessarily mean that this very high bar to create a permanent establishment has been crossed.
I can write in the luxury of knowing absolutely nothing about how Google actually operates, but I can well imagine that UK staff would be allocated customers to work with, and would be given ratecards and guidance on discounts to inform their discussions.
If the process was then that the UK staff negotiated the best deal they thought possible and submitted it to Dublin, where a manager either accepted or rejected it, the test for a permanent establishment has not been met.
Transfer pricing
It is interesting to speculate on what the committee would have asked if it had been properly advised and asked the right questions; or even if ot had listened carefully to the answers they were given.
John Dixon, from Ernst & Young, would only speak in non-specific terms about companies with a similar UK/Irish structure, but he made the following, very provocative, comment:
“In the theoretical example I am painting, we have a UK service company whose role is to provide services to the Irish company. There are two tests that a tax authority would need to look at.
“One is: is the Irish company trading in the UK through a permanent establishment? Secondly, is the UK company being properly remunerated for the services that it provides?
“There is another leg to the issue that you need to consider, which is: if the service company is providing real services that might get close to a sale but not quite be actually at the point of sale, one would expect its remuneration to be higher than a pure, classic service company. There are two dimensions here: the selling and the services that are being provided.”
Which should have made Ian Swales MP, who had asked the question to which Dixon was responding, forget about the permanent establishment angle for a moment, and look at transfer pricing (he actually responded with a question about where the customer was situated, which he appeared to have derived from VAT principles…).
If the UK-based service company is providing, not just a lead, but a complete potential sale, which Dublin simply has to decide to accept or reject, you would expect the UK company to get a pretty good return.
Is £40m or so profit (based on the corporation tax paid) a fair rate of remuneration on deals worth about £12bn? I genuinely don’t know, although I note that, worldwide, Google seems to have made about a 20% profit margin in 2012. A lot depends on how much value you can put on the intellectual property, such as the search algorithms, used by the Irish company.
Tax tranparency
And so the debates go on, characterised mostly by heat rather than light. John Watson argued persuasively for action on tax avoidance in last week’s issue (“Fruit at the top of the tree), but much of this is insoluble at a national level.
The OECD plan on base erosion and profit shifting (BEPS) shows the way forward, but it is going to take some considerable time to come to fruition.
One proposal which seems to be implicit in the OECD action plan, in Article 13 on transfer pricing documentation, is country-by-country reporting. The plan says that new rules may require multinationals to provide “all relevant governments with needed information on the global allocation of income, economic activity and taxes paid among countries according to a common template”.
I have always been sceptical as to whether this is really helpful, since a great deal of time and effort will be spent in determining whether particular profits were (and should have been) correctly allocated between countries which have broadly similar rates of tax.
There is, however, a much better argument for multinationals disclosing the profits they make in very low-tax states, and particularly in secrecy jurisdictions, and if country-by-country reporting is coming anyway, it would be good PR for companies to start by releasing this information.
It may not be required as yet, but there is no reason why publicly owned companies should not be prepared to justify why they have profits of a given level in a jurisdiction where very little of their commercial activity is carried on.
Those who could do so adequately (and the measurement of such adequacy would not be easy) might well be appropriate recipients of a positive tax transparency score, if that were to be the final outcome of a revamped Fair Tax Mark.
