Is tax treaty arbitration really a bad thing for developing countries?

I’m at the United Nations tax committee annual session this week, where I’ve learnt that I have to be careful what I write here, after a couple of posts from this blog were included in an input document [pdf]. Erk!

I’ve been taking the opportunity to discuss with delegates the recent article [£] by the chief of the UN committee’s secretariat, Michael Lennard, on the inclusion of arbitration in tax treaties with developing countries. Lennard begins from the concern that I’ve raised here too, which is that the BEPS Action Plan implies the use of a multilateral treaty to introduce mandatory arbitration clauses into existing tax treaties.

There are a number of potential problems with arbitration from a developing country perspective, which Lennard outlines, drawing on the experience of investment treaty arbitration. Briefly:

  • the cost of a full blown arbitration could be prohibitive for developing countries, forcing them to capitulate on some occasions, or alternatively stacking the process in favour of more wealthy countries who can afford the most skilled and experienced legal representatives.
  • Because transfer pricing expertise is limited in developing countries, it’s likely that most arbitrators themselves will come from developed countries, in which case their neutrality – or at least their sensitivity to the realities of developing country tax administration – might be questioned.
  • There is a range of concerns about the lack of transparency in arbitration outcomes, which is a problem for scrutiny (I’ve shared before a paper by Alison Christians on this topic), but could also mean that countries and lawyers only have their own experience to learn from, further biasing the arbitration process against developing countries to whom arbitration would be new.

Lennard also talks about a range of ways that arbitration might work better for developing countries. I’m just going to focus on one aspect of this, which is the ‘simplified procedure’ that is used in the UN model’s optional clause. (It’s also known as ‘baseball arbitration’, because it’s used in pay negotiations in the US baseball league).

Under the simplified procedure, the arbitrator doesn’t have to produce her own solution that tries to synthesise the concerns of the two sides. Instead she just chooses between one side’s position or the other.

Delegates I spoke to here said that the analogy with investment arbitration doesn’t hold when the simplified procedure is applied:

  • The costs are much lower, because there’s no need for lengthy meetings, the process demands much less time from the arbitrator, and – so one committee member said to me – you don’t need a lawyer to prepare written submissions.
  • The arbitration tends to move countries’ positions closer together, because it’s an all or nothing outcome, and a more conservative position might be more likely to succeed.
  • The democratic scrutiny point notwithstanding, transparency isn’t such a concern because the arbitrator is only permitted to choose one side, not to explain her reasoning – so there is very little to learn from past experience.

These discussions have certainly changed my thinking a bit, but I’m not sure that it is as clear-cut as those who favour arbitration suggest. In a judgement-based process, as opposed to a purely rules-based one, the quality of submissions is sure to affect the outcome. And it’s inevitable that an imbalance between countries in resources, expertise and experience will translate into an imbalanced outcome. Isn’t it?