Base erosion and profit-shifting: What do we know?

[From the Tax Justice Research Bulletin 1(4)]

The OECD was mandated by the G8 and G20 to revise its international tax rules in order to combat multinational corporate tax dodging – known in OECD jargon as “Base Erosion and Profit-Shifting (BEPS). The final deliverables due later this year, and because of the complexity of international tax, it has been broken down into 15 “Action Points.” (On Monday, TJN blogged Action Points 3 and 12, as discussed by the BEPS Monitoring Group, or BMG.)

The geeks’ Action Point of choice has always been number 11, which requires the collation of baseline data on how far profits are misaligned away from genuine economic activity: e.g. by being shoveled into tax havens where nothing really happens. Action Point 11 also involves continuous tracking of progress in this area.

The team working on BEPS 11 has now published a discussion draft, which sets out not only their thinking on questions of data and measurement (check back later with the BMG for our response to these) but also a broad survey of the literature on the scale and channels of BEPS.

Some of the main findings:

• There is broad consensus (regardless of time period, country or data source) that there has been widespread tax-motivated BEPS activity by multinationals — though the intensity varies.
• BEPS via manipulation of debt and interest payments is particularly important (and relatively well-studied: e.g. Huizinga, Laeven and Nicodeme (2008).)
• Transfer pricing studies of various kinds show consistently, again, that prices are manipulated for tax reasons; and that intangible assets are important here. We might also point to the more recent study by Banque de France researcher Vincent Vicard, finding that the deviation between intra-firm pricing and true arm’s length prices may result in a revenuetax base loss to France of $8bn in 2008, having grown over the decade – and this remains just a part of estimated total profit-shifting. [Phew!]
• Treaty shopping is estimated to reduce withholding tax rates ‘by more than five percentage points from nearly 8% to 3%.’
• Transparency: evidence on the role of disclosure and transparency – i.e. measures such as BEPS 11 itself – is limited but striking. Most obviously, Dyreng, Hoopes and Wilder (2014) found that ActionAid’s revelations about FTSE100 companies failing to meet a rule on disclosure of subsidiaries resulted in greater disclosure, a reduction in ‘tax haven’ subsidiaries and a reduction in avoidance [Kudos to Actionaid!].

• Developing country scale: only two studies provide a comparative assessment of the likely scale of BEPS in developing as opposed to higher-income countries. The first, by authors including long-standing critics of NGO estimates in the broad area, finds that profit-shifting (in this case, only that of German multinationals and via debt manipulations) is roughly double the size in developing countries. The IMF, meanwhile, recently presented evidence that the revenue losses of developing countries may be several times higher than elsewhere (as a share of corporate tax revenue).

Without wishing to prejudge the BEPS Monitoring Group’s analysis (oh go on then), it is clear even from this chapter of the OECD study alone that the authors have reached a point very similar to that which motivated Richard Murphy to elaborate the first TJN proposal for country-by-country reporting back in the day: namely, that existing data sources can only provide parts of the picture, and to develop a comprehensive understanding of BEPS will require consistent and comprehensive data on global multinational activity. [Well done Richard!]

To this we might add, in the spirit of Uncounted, that current data availability has “non-random weaknesses” which will systematically result in disproportionately lower tax revenues in lower-income countries. BEPS 11, if it delivers that comprehensive data, might just be the greatest contribution to fairer international tax that is compatible with the insistence on separate accounting.