Where have all the tax avoiders gone?

A new holding company created in the Cayman Islands expressly to sell assets worth tens of millions of dollars. Those assets assigned an accounting value of zero, effectively invisible in the company’s balance sheet. An “arbitrary” US$10,000 cash injection explicitly designed to activate a tax law permitting a “total tax free gain of about US$60 million”, calibrated not to affect the “commercial effects of the transaction”.

Obligatory tax-blog picture of the Cayman Islands. Pretty, isn't it. (Creative Commons slack12/Flickr)

What label should we give to engineered offshore transactions designed to avoid paying tax? I used to think I knew the answer to that question, but as 2014 draws another turbulent year of tax controversies to a close, I’m not so sure.

An odd cross-current has become detectable in the “tax avoidance” debate this year. The professional and business side of the debate used to be dominated by a fairly standard position: boiler-plate defensiveness crossed with ‘Duke-of-Westminster’ fundamentalism about the legal and ethical neutrality of shrinking one’s tax bill to the legal minimum. Over the last twelve to eighteen months almost every vocal big business, business association and tax professional body has moved away from that previous default. Almost everyone now accepts that choices between filing positions have consequences – if not ethical then at least reputational - that taxpayers should take into account. But in the next breath there tends to follow a second sentiment: that one’s own behaviour has either already changed, or didn’t need to change in the first place; that the scale of corporate tax avoidance has been overblown; and that in any case the corporate income tax isn’t a very good idea to start with.

This March 2014 report from the Association of Chartered Certified Accountants is a case in point. It’s prefaced by ACCA’s articulate head of tax with the usual social responsibility credo: "This paper may well show that the global corporate income tax system is not broken, but questions remain about the evasion/avoidance/aggressive tax planning debate. Multinationals need to be clear about the value they bring for the benefit of their shareholders and wider society, and to communicate to all stakeholders their underlying commitment to the building of a sustainable business. Companies must see the management of tax obligations as part of that process of creating long-term value." 

And yet the sentiment of the report itself is quite different: it’s a jeremiad against the “fiscal illusion” of wrong-headed tax campaigners. It argues simultaneously that corporate tax revenues are not declining; if they are declining, it’s the financial crisis not tax avoidance to blame; and that if corporations *are* reducing tax liabilities by booking income in tax havens, then not to worry because “[t]ax havens add value by allowing multinationals to reduce their tax liabilities while increasing their investments in high-tax economies.”

Now isn’t the time to delve into whether this report’s numbers stack up (that’s another blog post). For now, just note the jarring cross-current in what is fast becoming a new policy boiler-plate, which seems to amount to saying: ‘it’s a terrible thing, tax avoidance; fortunately there’s none here, and if you look for it too hard you’re going to cause another recession’. The same rhetorical rip-tide is evident in large businesses’ submissions to the OECD’s much-feted ‘Base Erosion and Profit Shifting’ (BEPS) project, the locus of wealthy economies’ current efforts to fix corporate tax avoidance. There’s near-universal acclaim for the BEPS project, and near-universal opposition to any of its proposals that require significant change to widespread tax practices. This statement from BIAC – the primary OECD lobby group for multinational business - on proposed measures against tax treaty shopping, is typical: 

“BIAC supports the broad aims of the BEPS initiatives, to tackle abusive, tax avoidance by a minority of taxpayers…[T]his must be addressed in a balanced and efficient manner, allowing the clarity and certainty of Treaty benefits appropriate to the vast majority of taxpayers entering into genuine commercial transactions.”

So a measure against tax avoidance is well-designed if it has a limited impact on the majority. Of course, lobbying exists precisely to stop governments doing things that are inconvenient to most members of a constituency. But as a wider argument, this seems an odd way to define the problem. Whether the “vast majority of taxpayers” are in fact engaging only in genuine commercial transactions is an empirical question that’s not tested here. Instead, tax avoidance by definition becomes a tax behaviour that is extreme or unusual. It's judged according to its prevalence. If everyone’s at it, it can’t be tax avoidance. 

To give due credit, this isn't how big businesses, governments or tax-professional bodies have defined tax avoidance – at least outside sound-bites and lobbying documents. Elsewhere, in an explosion of charters, guidelines and statements of principle, they’ve generated definitions of unacceptable tax avoidance based, as you’d expect, on an activity’s intention, effect and (especially) its character.

Global tax planners Deloitte, for instance, published a set of ‘tax advisory principles’ earlier this year which defines acceptable tax planning based on the character of an activity - ‘contrived’ or ‘artificial’ - not its prevalence: 

“We do not promote artificial tax planning structures. We do assist our clients with planning their tax affairs in relation to their business and commercial transactions. We believe that our clients should only engage in tax planning that has commercial or economic substance. We do not support tax planning that appears highly contrived or artificial.” (Emphasis added).

In January 2013, during a grilling at the UK’s Public Accounts Committee, Deloitte’s Head of Tax Policy, Bill Dodwell, fleshed out his conception of the uncontroversial, bread-and-butter work of tax structuring that shouldn’t be regarded as avoidance. With apologies for prolixity, the exchange is worth quoting in full:

Q29 Chair: Have you got an estimate of how much tax people do not pay because they take advantage of the advice you give them?

….

Bill Dodwell: No, it is impossible to do that. There are lots and lots of work where there is not a sort of, "I’m starting. I’m paying £100 in tax. Let me give you this piece of advice, and you will only pay £85 in tax." Most of the work we do has no relationship to that. At the moment, for example, I am working on a de-merger of a private company. That is a big piece of work. Our fees are several hundred thousand pounds for that, but there is not a tax saving. It is simply splitting up the group because that is what the owners want us to do. There is lots of work of that sort.

Q30 Ian Swales: So why are they paying you all that money if there is no tax dimension? Why are you there, and why are you charging all those fees? Why would they pay them?

Bill Dodwell: Because they have made a business decision that they would like their group split up into various pieces. It is an international group. We are trying to fit into the specific statutory reliefs we have in the UK and elsewhere to provide that this de-merger-this split-will not cost tax. If they stayed as they were, they would not pay anything. The split is designed in such a way that it falls into the statutory reliefs.

Q31 Chair: You are designing it within the statute, but you are designing it in such a way as to minimise their tax. Avoiding tax has become a new way of making profits.

Bill Dodwell: No, I do not accept that at all. This de-merger is being done quite specifically for commercial reasons. The purpose of our law is quite clear. De-mergers are supposed to be done in such a manner that no one gets a tax advantage, but equally there are not supposed to be tax costs. Fitting into that is what we are trying to do.

So what does one of these Deloitte-advised de-mergers look like in practice? Does it avoid Deloitte’s own shibboleths of being “contrived”, “artificial”, or failing to reflect “commercial substance”? As it happens, just two months before Bill Dodwell gave this description, an interesting Cayman Islands’ court ruling outlined Deloitte UK’s tax advice to the Canadian oil company Talisman Energy on a de-merger involving selling its stake in Colombia’s Ocensa pipeline. Again, apologies for prolixity, but it’s only through the detail that we can understand what’s going on and why.

Talisman wants to parcel up its Colombian assets for sale. These include:

• A 12.15% stake in the Colombian company that owns the pipeline (Oleoducto Central S.A.), which Talisman ended up selling for around $590m;

• Rights to transport oil through the pipeline – assigned to Talisman under a 1995 Transportation Agreement signed between Oleoducto Central S.A. and a Talisman subsidiary registered in the Cayman Islands, Santiago Oil Company – rights it appears, in the end, to have retained so far.

Like every other multinational engaged in a de-merger, the Talisman Group doesn’t want to pay tax on the gains it makes from the sale of assets. The standard way for a multinational to do this is to ensure that it’s not selling the physical asset directly (gains on whose sale might be taxed in the country where the asset is located) but selling shares in an offshore company that owns the asset, and to ensure that that offshore company is itself owned by a company in a jurisdiction where it isn’t liable to capital gains tax (like a tax haven). As the judge in the Caymans’ court case explains: “under the current Colombian tax regime [and that of many other countries] the sale or exchange of the outstanding shares of a foreign company by a foreign seller to a foreign buyer is not a taxable event for the purposes of income or capital gains taxes. For this reason it is potentially advantageous to the Talisman Group to sell the shares of group companies rather than sell their underlying assets.”

In Talisman’s case, their stake in the pipeline company, and slices of their stake in the Transportation Agreement, are held by two Caymans’ subsidiaries: Santiago Pipelines Company (‘SP’) and New Santiago Pipelines Company (‘New SP’). SP and New SP are themselves owned by a third Caymans’ subsidiary, Talisman Equion (Cayman) Inc. So if the Talisman Group wants to flog off its stake in the Colombian pipeline and the Transportation Agreement tax-free, it simply has Talisman Equion (Cayman Inc) sell its shares in SP and New SP. The assets being sold are, in theory, shares in two Caymans’ companies, outside Colombian tax jurisdiction. And any gains from flogging off SP and New SP accrue to Talisman Equion, itself registered (and presumably resident) in the Cayman Islands, where there is no capital gains tax. 

Diagram by author based on ruling in Cause No. FSD 91 of 2012 in the Grand Court of the Cayman Islands.

There’s absolutely nothing unusual about this. Almost every multinational-owned mine, oil pipeline, factory and forest is bought and sold via a tax-haven holding company. The IMF’s Fiscal Affairs Division may not like the manoeuvre very much - they say the “non-taxation of indirect transfers…raise[s] serious equity and political concerns” because they allow tax-free gains of billions of dollars on the sale of assets in some of the world’s poorest countries - but then multinationals and tax havens don’t really have to give a toss what the IMF thinks (that’s for losers like Nicaragua…), especially not in a Staff Report. As the IMF’s Fiscal Affairs Division pointedly notes, this is a practice that wealthy OECD member states have left entirely out of the BEPS project. It’s plain vanilla M&A: if we’re measuring tax avoidance as extreme or unusual behaviour - according to its prevalence - then this isn’t it.

Talisman also seems to want to sell its Colombia-related assets off in separate parcels, and that means getting the different assets into different subsidiary companies, so that those subsidiaries can be sold off separately. As the judge puts it, “it is advantageous to reorganise assets intended for sale so that they are owned by special purpose vehicles which hold only one asset and no liabilities.”

And now there’s a second problem. If the Talisman subsidiary owning a given asset simply sold that asset to a newly-created Talisman subsidiary (a ‘special purpose vehicle’), then the gain made from that sale could potentially be taxable – these are, after all, separate companies liable for tax on their own profits and gains, even if they’re all owned by the same ultimate parent. The solution is a series of ‘spin-off agreements’. Instead of Subsidiary A selling the asset to newly-created Subsidiary B for cash, Subsidiary A gives the asset to Subsidiary B for nothing. In return Subsidiary B issues shares to their common parent company with a value equal to the asset being transferred, and Subsidiary A simultaneously reduces its share capital by the value of the asset being transferred. Thus an asset has moved from A to B, but there’s no taxable gain anywhere, and the value of the parent company’s assets stay the same. 

In this case, what are being ‘spun off’ are the portions of Talisman’s Transportation Agreement – the rights to transport oil through the pipeline – held by SP and New SP. They themselves appear to have acquired these rights through spin-off agreements, and are now spinning them off to two newly-incorporated Cayman Islands subsidiaries, Talisman Santiago (Cayman) Inc and Talisman SO (Cayman) Inc.

Now here’s the weird bit. According to Deloitte, the assets being spun off - SP and New SP’s shares in the Transportation Agreement – have an “estimated value” of $60,274,500. But their ‘book value’ is recorded on the firms’ balance sheets as…zero. Deloitte says that when “[Talisman subsidiary Santiago Oil] entered the [Transportation] agreement, no accounting entries were recorded and therefore the [Transportation Agreement] rights were held at zero in the accounts of SO.” When parts of this Agreement were spun out, they therefore also carried a book value of zero. 

I confess I don’t really understand how or why this would be done. The judge doesn’t either: as he pointedly says, 

“Why property having an economic value (presumably market value) of US$18,124,500 for SP and US$42,150,000 for New SP is not reflected as an asset in their respective financial statements led me to raise questions which have not been answered, or at least not convincingly….I find this [Deloitte’s] explanation unconvincing, but I suspect that this accounting treatment is actually a very important part of the whole reorganisation plan.”

(I asked Deloitte UK to explain this too, and to check the accuracy of my description of the whole transaction, but they politely declined, saying they couldn't comment on specific cases. I also asked Talisman for a repsonse to this post, but have yet to receive a response).

  

And here comes a third problem. The zero value of these assets threatens the tax-free status of the spin-offs: under Colombian law, to be tax-free the spun-off asset has to have a positive book value. The solution is simply to have SP and New SP give away $10,000 in cash to the new spin-off companies, alongside the ‘no-value’ transportation rights. Explaining this to the court, Deloitte explicitly say that this ten grand figure is arbitrary, and designed not to affect or be related to the underlying commercial reality of the transaction: 

“The amount of $10,000 was chosen as representing an asset which is not so small that it could be effectively ignored (e.g. $1 or perhaps $100), but is also not sufficiently large that it materially alters the commercial effects of the transaction”.

The new Cayman Islands special purpose vehicles will then each, in the words of the judge, own “a single asset (cash at bank of US$10,000 in one case and an inter-company receivable of US$10,000 in the other case) which will exactly match its share capital, but they will also own rights under the Transportation Agreement having a book value of zero but a market value of US$18,124,500 and US$42,150,000 respectively. The shares of these SPVs can then be sold at their market value, thereby potentially realising a total tax free gain of about US$60 million.” (emphasis added).

So - should we call this tax avoidance?  

- BIAC's bad-apples theory of tax avoidance would suggest not. Despite its abstruse complexity, I suspect this kind of transaction is entirely routine. Companies getting rid of assets tax-free through ‘spin-offs’ or ‘starbursts’ is, after all, corporate M&A bread and butter, in part precisely because of its tax advantages over a sell-off.

- And yet this example ticks most of the boxes of the business lobby's new definitions and standards of tax avoidance based on character, intention or effect. Recall again what Deloitte says in its (character-based) Tax Advisory Principles: “We believe that our clients should only engage in tax planning that has commercial or economic substance. We do not support tax planning that appears highly contrived or artificial.” Yet here is Deloitte advising an ‘artificial’ tax-motivated transaction, explicitly designed not to reflect the ‘commercial or economic substance’ of the transaction, which a court has ruled is designed to effect a tax-free transfer of assets: avoiding tax on these assets is, the judge rules, its “business purpose”. 

- Its intention is likewise clear: to entirely avoid a tax charge. 

- And its effect is to do just that. 

Finally, Bill Dodwell argues something else: he suggests that the key question is whether you think the multinational should pay tax on the gains from its demerger, however abstrusely it’s achieved. He argues that they shouldn’t, since if the demerger wasn’t undertaken at all, no tax would be due. In his view it’s not the proper object of capital gains tax, it’s a bear-trap to be legitimately stepped around. On my part, I find it difficult to make up my mind about whether this is a bear-trap to be legitimately stepped around; or a bear-trap the stepping around of which prepares the way for avoiding tax on the gain when, in due course, the transportation rights are sold.


***

If you’ve got this far through my rambling tax voyage around the Caribbean, well done. The take-home, I think, is this: 

1. The tests we – governments, campaigners, businesses – are now designing to identify tax avoidance in the abstract are important, because they’re being used to re-design the global corporate tax system. 

2. And it seems to me that the three approaches on display here – ‘bad apples only’, ‘artifice only’, and ‘only if tax were due without the transaction’ – don’t take us very far. A transaction can be enormously common, but hugely artificial. A transaction can be wholly artificial, but designed to avoid an unintended bear-trap rather than an intended tax liability. And the circularity of the argument that a tax-free transaction is legitimate because there would be no liability without the transaction seems entirely self-defeating as a test of acceptable or unacceptable behaviour. 

3. Instead, these tests seem to be designed primarily to make ‘tax avoidance’ disappear in any specific instance: the current rhetorical vogue of the large-business lobby that we began with.

I am beginning to think that principles and tests of acceptable and unacceptable tax behaviour don’t help much in feeling our way to a better tax system, or better tax behaviour. They can’t replace the hard graft of empirical tax policy analysis, shorn of ethical or political judgements. In a given instance, is tax law leading to prevailing taxpayer behaviour that produces desirable tax incidence, with all its economic consequences? Neither ‘tax principles’, nor the BEPS process, are trying to answer that question: they’re seeking a “fairer” tax system without any empirical assessment, in aggregate and in the real world, of what the tax system, any given piece of tax law, or any prevalent tax behaviour actually means for who is paying tax and how much. (Incidentally, some economists are worried that there’s something similar happening in economic policymaking in general: a decline since the 1970s in actual modelling; and the rise of a purely ‘narrative’, ‘literary’ style of IS-LM argumentation and policy-making).

This is not a call for less politics in tax. It’s a call, actually, for more politics in tax. ‘Desirable tax incidence’ is an inherently political question to be decided a priori by electorates and legislators. Personally I believe capital and its owners should contribute more than they currently do to government revenues. I want a tax system that delivers this. I don’t care whether it prohibits or permits more or less ‘artificial’ tax transactions. In 2015, let’s stop labelling in the abstract, and start modelling and redesigning tax laws that actually deliver the incidence and the economic outcomes we want. However much they may grouch about the labelling, I suspect this is what the owners of capital want least of all.