Cancelling tax treaties: a lesson from the 1970s

Mike Lewis at ActionAid had a blog yesterday about the problem posed by Zambia’s tax treaty with Ireland, following up on ActionAid Zambia’s call for the treaty to be renegotiated. But how easy would that be?

One of the main points about the Ireland-Zambia treaty is that when it was negotiated 40 years ago, Ireland’s economic position, and its tax rates, were very different. For example, it hadn’t yet introduced its 12.5% tax rate as part of moves to attract mobile capital. As Mike points out, the problem now is not just that the treaty is extraordinarily lopsided in Ireland’s favour, but also that it doesn’t include the anti-treaty shopping clauses that can be introduced to prevent treaties with tax planning ‘conduit’ countries like Ireland and the Netherlands from being exploited. So it would be sensible for Zambia to renegotiate this treaty. The trouble is, that’s not always easy to do. After all, in this case, what would be in it for Ireland?

Tax Justice Network recently noted with satisfaction that both Mongolia and Argentina had cancelled a number of treaties. It’s interesting to look into these moves in a bit more detail. Argentina cancelled its treaties with Switzerland, Chile and Spain last year, but it’s clear that in the latter case what it really wanted was a renegotiation, because a new Argentina-Spain treaty has already been signed.

In Mongolia’s case, the focus seems to have been on treaties with tax planning conduit countries, and according to E&Y [pdf], cancellation was a last ditch option, after the Netherlands had “agreed to amend the dividend clause of its treaty from 0% to 5%, but have refused various other amendment requests.” Luxembourg had apparently “agreed to start negotiations” before Mongolia changed its tune. E&Y’s explanation of the rationale is very interesting, and taken with the above points suggests that anti-treaty shopping provisions were exactly what Mongolia was trying to get:

The MoF has been conducting research into Mongolia‟s double tax treaties, comparing Mongolia‟s treaties with the treaties of countries in a similar economic position to Mongolia (ie Philippines and Ghana) and undertaking some quantitative analysis of the estimated lost revenue from perceived treaty shopping. The conclusion reached from their analysis was that Mongolia’s double tax treaties are too generous towards taxpayers and result in significant tax revenue being lost.

Recalcitrance by wealthier countries who got a good deal in the original treaty is not a new problem. Around the time that the Zambia-Ireland treaty was being negotiated, a footnote to a rather old academic paper shows that one of its neighbours had other ideas:

based on conversations between the writer and the Assistant Secretary to the Kenya Treasury and the Financial Adviser to the Kenya Treasury on Mar. 26 and 27, 1973, it appears that Kenya recently cancelled all its tax agreements because it was recognised that the agreements were resulting in substantial losses of tax revenues and developed countries were found to be dragging their feet on negotiations for new double taxation agreements (or modifications of old double taxation agreements) while the old agreements continued in force.

This tactic has clearly worked for some countries. But it’s a dangerous game to play. At a meeting last week in central London, one participant cited a tax treaty negotiator saying that cancelling a tax treaty is “like declaring war”. Business representatives at that meeting emphasised that the main reason that they like to invest in countries with tax treaties is because of the ‘stability’ they create. The abrupt cancellation of a treaty rather sends the opposite message. Ernst & Young’s report on the Mongolian cancellation [pdf] says, menacingly, that the government,

does not discuss the implications to foreign direct investment and international reputational issues associated with unilateral cancellation of treaties, which is an important consequence of this move.