Nick Shaxson ■ Luxembourg’s Sweetheart Deals: Could the OECD Stop Them?

Prof. Sol Picciotto

Prof. Sol Picciotto

Guest blog by Prof. Sol Picciotto, a TJN Senior Adviser.

Luxembourg’s Sweetheart Deals: Could the OECD Stop Them?

In an earlier comment on the Luxembourg tax leaks, we pointed out that there is no world tax authority which can judge whether the sweetheart deals it has been offering multinationals are legal. The only bodies which could do so are the EU and the OECD – so how likely are they to take effective action?

Concern about these kinds of tax preferences were expressed way back in 1996 especially by the French and German governments. This led to the OECD report on Harmful Tax Practices of 1998, resulting in setting up a Forum to evaluate such `preferential tax regimes’. In parallel, the EU agreed a Code of Conduct, and set up an intergovernmental Group to identify and phase out “harmful” preferential regimes.

Neither of these bodies have managed to prevent Luxembourg from handing out these sweetheart deals to the over 340 multinationals, now revealed in the documents released by the ICIJ. The reason? Both bodies are toothless, as they lack any form of sanction to pressurise governments. The 1998 report did mention the possibility of “coordinated defensive measures”, but this was controversial, and was quietly dropped. The only legal constraint available has been the powers of the European Commission to challenge tax breaks which may amount to “state aids”. These were used against some national measures, for example to oblige Ireland to end its tax holidays for foreign companies, resulting in the Irish government opting for a general low corporate tax rate (now 12.5%).

Then the Commission became more timid, and made no more challenges until last year, when it targeted both Ireland and Luxembourg. Yet the cosy Irish arrangement with Apple Computer revealed in documents released on 30th September 2014 dates back to 1991. The secret Luxembourg deals documented by the ICIJ are more recent, but they go back to 2008. Why no action until 2013?

We now have a new effort to reform taxation of multinationals, the BEPS project launched by the OECD in September 2014, following a mandate from the G20 world leaders. What does the OECD propose now to deal with this?

The answer is: more of the same.

Dealing with Harmful Tax Practices is one of the 15 points of the BEPS Action Plan, and the OECD released its report on this on 16th September 2014. All it proposes is to revive the Forum process, by which government representatives working in secret evaluate each other’s tax breaks for multinationals, to decide if they are “harmful”. The tests include whether the deals are secretive, and whether they allow income to be sheltered without there being any “substantial economic presence” in the state. The OECD government representatives have spent a year trying to define both of these, and the report shows that they have failed.

One of the main kinds of tax break for multinationals is the “patent box”, which offers a low tax rate for patent royalties, to encourage multinationals to transfer ownership of patents to the country concerned. A French scheme of this type was found to have “harmful features” by the EU Code Group in 1999, and was modified. When the Netherlands introduced its own scheme in 2007, the Group decided it did not need evaluation, so other countries (Luxembourg, Belgium, Spain, Cyprus) followed suit.

The UK’s scheme introduced in 2013 was estimated by the Treasury to cost the UK taxpayer £1.1b per year, and we have commented on it before. This raised more hackles, and the EU Commission recommended that it should be considered harmful, mainly because it failed the “substantive activities” test. However, the Group could not agree, so the issue was passed on to the OECD. The report released in September revealed that a new test for “substantive activities” proposed by the OECD experts could not be agreed by the working party, due to opposition from … Luxembourg, the Netherlands, Spain and the UK. So zero progress there.

The report sounded more positive in relation to secret rulings, where it proposed a new procedure for transparency, described as “a framework for compulsory spontaneous information exchange on rulings” (p.36). But the actual description of the system (from p.38) shows that each state will decide for itself when it should notify, by a “spontaneous” exchange of information. The procedure is supposed to reduce the state’s discretion by requiring it to apply “filters” to decide when to supply the information — but it still just decides for itself. The Luxembourg rulings concern transfer pricing, and for these the Forum still has to decide the “practical guidance” (p.44). But it does seem that not all such rulings need to be notified, only those which meet the first three filters.

So the bottom line is that under the OECD approach states can’t even agree when to notify each other of secret rulings, and will only do so voluntarily. And they haven’t yet agreed a definition of “substantial economic activities”. Even if they do so, what would be the result? Those with such regimes, such as the UK, would tweak their rules. Others would devise their own variations – Switzerland and Ireland have already announced they plan to do so.

The process is worse than toothless – if the participants can’t agree that something should be condemned as harmful, it’s a green light encouraging adoption by all. It’s hardly surprising that researchers at the respected Institute for Fiscal Studies have warned of the dangers of a “race to the bottom” from patent boxes.

So the OECD’s report on Harmful Tax Practices is largely an admission of failure. It is due to be considered by the G20 world leaders at their meeting in Brisbane on 15-16 November, as one of the seven “deliverables” of the first year of the BEPS project. What, we wonder, will the G20 have to say about dealing with sweetheart tax deals such as those revealed by the Luxembourg leaks?

Related articles

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.