The original version was published at Financial Times Alphaville here. This version was updated on Sept 5th, 2019, responding to media comments. Scroll down to the bottom to see the new material.
Update: Taxwatch in the UK has written a useful follow-up article entitled “Is Tax Avoidance legal?,” with the same conslusion, drawing on overlapping but often different arguments.
Our provocative headline, for those who know this terrain, may seem to be plainly wrong. It is contradicted immediately by the Collins English dictionary – and probably by every other dictionary that contains this term.
And this is in contrast to tax evasion:
So tax avoidance is perfectly legal: it’s evasion, that’s illegal (and criminal)?
Not so fast.
This blog will show that much if not most of what routinely gets called corporate ‘tax avoidance’ involves activity that cannot be called ‘legal’. In fact, tax evasion and tax avoidance, and the questions of legality and illegality, can’t be separated from each other, whether definitionally, conceptually, or practically. Any efforts to draw bright lines between them will inevitably lead to confusion and bad policy.
This is important for many reasons, not least the fact that corporate lobbyists and their allies are currently working to get supposedly ‘legal’ corporate tax avoidance expunged from the global debates about, and measures of, illicit financial flows (IFFs). It is also relevant for our Corporate Tax Haven Index.
What gets called corporate “tax avoidance?”
In media reporting of scandal after corporate tax avoidance scandal, editors insist — often at the urging of their libel lawyers — on inserting lines such as:
‘the company has not broken the law,’ or, worse:
‘the company has done nothing wrong,’ or, perhaps even worse:
“what the company did was legitimate.”
They often include a quote from the multinational involved, which goes like this:
We abide by all tax laws in the places where we do business.
The point of this blog is to demonstrate that these statements are often, and perhaps usually, flat wrong.
To see how and why, let’s start with some striking statistics.
The first set of data comes from Germany’s Finance Ministry.
The coloured bars represent the billions of Euros in additional taxes collected from large corporations as a result of tax audits. The precise numbers are:
Year Large corps (€bn) All corps (€bn)
2013 13.4 17.1
2014 14 17.9
2015 12.9 16.7
2016 10.4 14.1
2017 13.8 17.5
Total 64.5 83.3
In other words, Germany collected 64 billion in extra tax revenue from large multinationals over five years, as a result of audits to get taxpayers to pay the right amount, and a total of 83 billion from all corporations. An accompanying graph in the same document (Graph 2) shows that about half of this, or €39.8 billion, was down to corporate income taxes (“Körperschaftsteuer,” which is the federal-level tax, plus “Gewerbesteuer,” which is local-level tax). This is serious money. And who knows how much more should have been, but didn’t, get caught by the auditors?
What’s happening here is that taxpayers tried to claim these billions for themselves but official audits found them to have transgressed the law in doing so. To label this stuff ‘legal’ is plainly wrong.
And this is true, of course, beyond Germany. Large Canadian corporations failed to pay around a third of taxes assessed, according to one analysis.
Data from the OECD, the club of rich countries that dominates international tax policy-making, show a consistent pattern.
50 percent of tax audits led to adjustments in 2015, suggesting that the original tax returns were unlawful. That was higher for large companies — and with the in-depth comprehensive audits, around 80 percent resulted in adjustments.
In other words, in half of the corporate tax audits, companies were found to have filed incorrect tax returns. In every country, the audits resulted in additional taxes being due, suggesting that companies were systematically trying to underpay: to get away with more than what was legally allowed. Unlawfully trying to obtain money that wasn’t theirs.
The tables contain further details. For the large taxpayers (Table 2 in the pdf,) 57 and 58 percent of audits led to tax adjustments in 2014 and 2015 respectively. And for transfer pricing audits, the figure was 58 percent in 2015. The audit adjustments added up to an average of 10.7 percent of total corporate income tax revenues (Table 5) though in some countries it was huge: in Brazil, for example, an average of 45 percent of corporate income (CIT) tax revenues stemmed from audit adjustments in 2014 and 2015, while in Italy the figures were 63 percent and 49 percent. Finally, when it comes to comprehensive audits – more in-depth and penetrating than routine audits – Table 4 shows that the rate in 2014 and 2015 was, respectively, 79 and 83 percent.
Lastly, Table 1 distinguishes between “audit assessments” — extra money the tax auditors think should be collected, and “audit collected,” the money actually recovered. On balance, it turns out, most audit assessments end up with the money being collected.
Overall, the patten is clear: multinationals routinely try to get away with (potentially unlawfully) dodging billions in tax they should by law be pay to contribute to schools and hospitals and tax courts and other public services.
It’s by now clear that this stuff can’t be called “legal.” But does that make it illegal? Or illicit? Are these illicit financial flows?
What is illegal, what is illicit?
The question of what is legal, and what isn’t, is a set of grey areas with many nuances and disagreements. Lawyers may have different opinions to accountants, who may have different opinions to tax authorities, or police forces. Different dictionaries offer different shades of meaning. The private enablers of tax avoidance actively exploit all these grey areas, and plenty more.
The legal profession is riven by different philosophies. A hardline approach might insist that something is only illegal if it has been judged so by the courts, beyond all appeal: innocent until proven guilty. But does that mean that if you commit murder and bury the body in a desert, nobody finds out, and you never go court, that you will die innocent? Few reasonable people would accept that. Putting this in a tax avoidance context, there is endless stuff that would, if discovered, be knocked down by the auditors or the tax courts, but it stays hidden, and they get away with it. Few reasonable people would accept this as ‘legal’ either.
A second layer of confusion is the fact that these cross-border corporate structures usually straddle several countries: what’s legal in Luxembourg may well not be legal in Louisiana. And what’s not legal for a multinational may be fully legal for the Big Four accounting firm that set up the structure for them.
Other confusing boundaries exist, with question marks over them: between ‘illegal’ and ‘unlawful’ (and ‘not ‘legal’;’) between ‘illegal’ and ‘criminal,’ and between ‘illegal’ and ‘illicit.’ And then there’s the age-old question of the difference between tax avoidance and tax evasion.
How can we navigate all this? Well, one helpful tool here is the concept of risk mining.
A senior official at a Big Four accountancy firm said they would sell tax schemes to clients even if they thought there was only a 25 percent chance they would survive a court challenge.
From a tax point of view the highest risk transactions are often those that are happening specifically for tax purposes e.g. a tax driven reorganisation.
extracts value from the public exchequer by seeking to withhold money, some of which should as a matter of law be paid into it.
Companies do, of course, often submit uncertain but innocent tax positions, often with complex unique transactions, or because their international activities are generally so complex that they can never be sure they’ve got the all tax bits 100 percent right. Filing an uncertain tax position is not in itself tax avoidance. It’s the deliberate structuring leading to their filing uncertain tax positions that is tax avoidance. Quentin’s diagram lays out the steps.
The diagram suggests there are three parts to this.
- “Effective avoidance.” Multinational sets up a structure to avoid tax, it’s legally watertight, and tax authorities don’t knock it down (whether they challenge it or not.)
- “Ineffective avoidance.” Multinational tries it on, gets audited, was found to have unlawfully claimed money that wasn’t due to it, and must pay up.
- Multinational tries it on with a scheme to claim money that isn’t due to it, and doesn’t get caught. They get away with this money, unlawfully.
The first category here can probably be labeled “legal” tax avoidance, while the second two can’t. As Quentin frames it, the behaviour crosses over into illegal tax evasion when the chances of a filing position succeeding a tax authority challenge are zero (though in practical terms the dividing line is entirely fuzzy.)
For low income countries, of course, the third category here will be especially large, because it’s generally much harder for them to challenge and claw money back from large corporations. They must typically send underpaid, overworked, sometimes under-skilled, often demoralised and intimidated staff up against large teams of highly skilled, highly paid and highly motivated corporate lawyers and accountants working for large multinationals, who have been known to offer personal inducements for favourable tax treatment.
If multinationals try to get away with so many billions in Germany, how large are the corresponding flows out of low-income countries in Africa and elsewhere? We will never know. Does this constitute the unlawful theft of public funds from poor countries by multinationals? Are these ill-gotten gains? Is this stuff “illegal?” Or “illicit?” Are these Illicit Financial Flows, or IFFs?
We have been banging the drum on IFFs (and on a closely related concept, capital flight) since our founding in 2003. Few people seemed interested back then: but the fight against IFFs now has international momentum. The United Nations’ Sustainable Development goals, for instance, include this recommendation, which the tax justice movement fought hard to obtain:
By 2030, significantly reduce illicit financial and arms flows, strengthen the recovery and return of stolen assets and combat all forms of organized crime.
The key question is, of course, how to define “illicit financial flows.” This sentence above looks extremely nefarious, and of course multinational enterprises don’t want their tax shenanigans publicly mixed up in this. No, they insist: they don’t do illicit financial flows – they do ‘tax avoidance.’ The illicit stuff is all the fault of poor, corrupt countries. TJN’s Director Alex Cobham illustrated this message in a recent presentation.
The more (apparently) cerebral argument that is being made here, is this:
conflating legal and illegal behaviour under a single definition involves a loss of clarity and a risk of confusion. . . . if the overall goal is to strengthen the rule of law, democratic accountability, and the effectiveness of states.
Are they right?
Well, let’s start with some more definitions. the Oxford English Dictionary (OED) defines illegal as “contrary to or forbidden by law, especially criminal law,” while illicit is defined as “forbidden by law, rules or custom”. So ‘illegal’ is a subset of ‘illicit.’
We’ll get to the ‘criminal’ bit in a minute. But for now, a swathe of corporate tax avoidance is, on these definitions, illegal, and illicit or potentially illicit. That swathe includes the third category above (and potentially the second category too, if or until the correct money ends up being paid.)
The first category is more interesting. It is certainly not illegal, but if it ends up with a low-income country being (legally) tricked and cheated out of a large large sum of money, that definition, and others, might consider this to be an illicit financial flow. For instance, the High Level Panel on Illicit Financial Flows from Africa, says that Illicit Financial Flows focus on money “illegally earned, transferred or used” but may include activities:
“that, while not strictly illegal in all cases, go against established rules and norms, including avoiding legal obligations to pay tax.”
Clearly, then, corporate tax avoidance belongs here from a definitional perspective: certainly in Category 3, and potentially in both Categories 2 and 1.
We won’t deny, however, that this is a nuanced and complex picture, involving large grey areas.
So it’s good that there’s a better, clearer way to cut through the nonsense: move away from legalistic and definitional questions, and out into the real world of economics and democracy: the practical, policy-related perspective.
Here, it’s obvious that the categories of avoidance and evasion, or the categories of legal and illegal, have far more similarities than differences:
- cross-border avoidance and evasion both rely on the international tax haven and offshore secrecy system, involving highly paid tax advisers and grey areas in the law; and
- economically and politically, the two are essentially the same thing. Both result in losses to government revenues, reducing the funds available for public services. Both increase the burden for ordinary taxpayers, corrupt the integrity of markets, draw powerful multinationals into lawbreaking, increase inequality, undermine democracy, and stoke public anger. They are anti-social.
This, for us, settles the debate. The enormous economic and political dimensions of this are the ones that matter: over and above the legal hair-splitting. The High-Level Panel’s flexible approach is the right one.
This image, from TJN, helps underline the point:
This isn’t quite the end of the story, however, for this image introduces the c-word. When does ‘tax avoidance’ constitute criminal behaviour?
Dictionary definitions usually say it’s tax evasion, not tax avoidance, that involves criminal behaviour. Once again, though, this distinction crumbles on examination.
Here’s one perspective, from Jolyon Maugham, a prominent UK tax barrister:
I have on my desk an Opinion – a piece of formal tax advice – from a prominent QC at the Tax Bar. In it, he expresses a view on the law that is so far removed from legal reality that I do not believe he can genuinely hold the view he says he has. At best he is incompetent. But at worst, he is criminally fraudulent: he is obtaining his fee by deception. And this is not the first such Opinion I have seen. Such pass my desk All The Time.
It’s not just lawyers who may be guilty of this. Consider the giant “Luxleaks” scandal, where Luxembourg authorities facilitated and even encouraged PwC and other accounting firms effectively to industrialise the processes of cooking up complex international tax schemes, to help multinational enterprises escape billions in tax elsewhere. This was reported around the globe as a “tax avoidance” scandal, along with the usual claims that it is “all perfectly legal.”
Not everyone fell for this story, though. The occasional clear-eyed investigator explained what was going on:
(Source: p95 here.)
Why “mass tax crimes?” Well, criminal activity means illegal activity which is dealt with under criminal codes, with criminal penalties. Most definitions of ‘criminal’ require criminal intent: where people make a deliberate, knowing attempt to deceive and (in these cases) take money that isn’t theirs to take.
Nailing down criminality, though, is usually tough. It’s hard to prove deliberate intent to deceive in court. The question often hinges on matters of interpretation, what constitutes economic ‘substance,’ what corporate bosses were actually thinking when they set up these structures, and so on. And tax havens often don’t criminalise stuff that most countries would criminalise: for example, plenty of tax-related activity that would be considered criminal activity in most countries does not get treated under the criminal code in Switzerland.
Luxleaks may have involved “mass tax crimes.”
When a company claims that all its decisions are taken in an office in Jersey, but it turns out that that the relevant directors make all their decisions in London then fly out to Jersey once every few months just to tick the ‘economic decision-making in Jersey’ box, then there’s a deliberate subterfuge going on. There’s the old expression, “everyone knows what is going on.” That kind of behaviour could attract criminal penalties – and if it does, we can call it criminal.
All this just underlines our earlier point: a lot of what gets called “tax avoidance” is not ‘legal’, a lot is illegal, and a fair bit may be criminal.
Conclusion and recommendations
Finally, let’s return to that argument made by those apparently seeking to absolve multinationals of the ‘illicit’ tag:
conflating legal and illegal behaviour under a single definition involves a loss of clarity and a risk of confusion.
If you’ve read this far in this blog, it will be abundantly clear that it’s the opposite proposition that is true. It’s the trying to separate the legal from the illegal and the illicit that involves a loss of clarity and inevitable confusion.
Several recommendations emerge from this.
- It makes no sense, conceptually, definitionally, or practically, to expunge or separate corporate tax avoidance from definitions of illicit financial flows.
- Governments need to get serious about beefing up their audit capacities: there is easy money here. Help is now on hand, from (among others) Tax Inspectors without Borders (we have recently written about other pitfalls to be avoided here). Our Financial Secrecy Index provides further pointers, as will our imminent Corporate Tax Haven Index: coming very soon.
- Governments must beef up enforcement, and apply criminal and other sanctions for multinationals that transgress.
- Sanctions must be extended to the ‘enablers’ – especially the accountants, lawyers and offshore bankers who aren’t just passive players in this game, but often actively encourage multinationals to set up dodgy tax schemes.
- Tax court decisions should be made public with only very narrow exceptions.
Update: some comments and our responses
There have been a few responses, some more sober than others, to our article. There are plenty the Financial Times Alphaville article, but the most coherent (mostly temperate) set come in an article about the original version of this blog, researched and written by Tax Analysts journalist William Hoke. It contains some helpful observations. First:
The ‘state of mind’ point is interesting. We disagree. Let’s say you’re driving a car along a contested border between two countries with different speed limits. If you drive at the higher speed, then under the lower-speed country’s laws you are driving at an illegal speed. You may have an honest belief you’re not breaking the law, but that doesn’t mean you aren’t. (Likewise, all those “tax is theft” zealots who believe the state is illegitimate would be exonerated of lawbreaking, just by virtue of their weird beliefs). Having an “honest belief” that my risky tax position is lawful does not exonerate me of lawbreaking if it isn’t.
The second point – that if the belief is honestly held then there is nothing ‘illegal’ – is tilting at a straw man – and this is the main misunderstanding of several people cited in the Tax Analysts article. (Many other comments – from Dan Mitchell of the Cato Institute, from Zana Ritchie of BDO in Australia, and indeed from the Tax Analysts headline-writer, with “Claim that Corporate Tax Avoidance is Illegal Sparks debate” – fall into the same basic misunderstanding.) In this blog we carefully never argued that this stuff is generally ‘illegal.’ We argued that you generally can’t call it ‘legal’. That is a subtle, but fundamental difference. The main purpose of this whole blog was to push back against the “it’s all legal” brigade, which tends to dominate media coverage of these corporate shenanigans.
Ritchie was also paraphrased as saying that our figures given for percentages of tax audits leading to assessments was ‘inaccurate’ at least from an Australian perspective, since these numbers assumed that audits were carried out at random, rather than in a targeted fasion: targeted audits naturally will find a much higher percentage of transgressors. That is an important point to note – but the article, as originally worded, still remains correct. What is more, it is uncertain how good tax administrations are at targeting the highest-risk schemes – and we also know that in Germany at least, large-taxpayer audits happen routinely, at random.
Stefano Grilli, an Italian tax lawyer, fumed about the “socialist-minded” author’s credentials (memo: does arguing against rigged markets and exposing the whitewashing of lawbreaking make one a “socialist?”) but also made an interesting, if odd, point, that in Italy, “tax criminal charges have become de facto blackmailing tools to settle tax audits” and that multinationals often prefer to settle quickly since the alternative could be 10 years’ tax litigation. That may well be so, but it doesn’t negate our arguments.
The article makes a fair criticism that the cross-country aggregate data isn’t weighted for currency values and size – though again, this doesn’t negate any of the conceptual points, or the data itself: it’s just presented in rather crude form. (This is a blog, not a peer-reviewed academic article: others can please feel free to the time to do the weightings.) And it underlines a point we’ve long made: in country after country, we need more and better data, and better-resourced tax authorities. (To read more about these thorny questions, click here.)
There is also this, in Hoke’s article:
There was also quite a bit of pushback against the idea that tax lawyers would push such “25%” aggressive tax strategies as those cited by Margaret Hodge. Tax professor Omri Marian, for instance, said such extreme pushing of the envelope was hardly known in the US, especially since an IRS crackdown in the 1990s, while admitting that “U.K. practice may be different.”
Well, OK, maybe. But take a look at this story about Walmart – or re-read Maugham’s comments, above.
And we’ll add another point here, related to this. Sometimes, and certainly in the UK, the tax authority acts as if it is in cahoots with the multinational, trying to weaken tax collection (“competitive” tax collection, they often call it). Take this:
The British government refused to assist a French investigation into suspected money laundering and tax fraud by the UK telecoms giant Lycamobile – citing the fact that the company is the “biggest corporate donor to the Conservative party” and gives money to a trust founded by Prince Charles.
It’s incredible, isn’t it – but it’s far from the first time this kind of thing has happened (that was a cross-border example: there are plenty of more domestic examples).
So, apart from a few details, this article survives this latest wave of attacks pretty well.