Tax treaty negotiation: what affects the outcome for developing countries? (Part 1)

[T]ax administration and tax policy officials in Uganda are not sufficiently trained in the area of tax treaties and international taxation. As a result, Uganda has a weak tax treaty negotiation team that concludes treaties more intensively reflecting the position of the other contracting state.

A first attempt to answer a question that I’ll be coming back to many times. It’s based on an afternoon spent in the library reading  The Impact of the OECD and UN Model Conventions on Bilateral Tax Treaties, a very heavy tome that thus far I’ve only been able to cherry pick through.

The proposition

Considering the negotiation of tax treaties between developing and developing countries (or, more generally, between capital-importing and capital-exporting countries), I think there are several common sense hypotheses:

  1. Developing countries want to maximise a) the base that they can tax, for example by using a wider definition of the circumstances in which they are entitled to a tax foreign investor in their country, and b) the rate at which they can tax it.
  2. This means that, where they differ, developing countries will prefer the provisions of the UN model treaty, while developed countries will prefer the provisions of the OECD Model.
  3. The outcome of negotiations is likely to reflect the power balance between the two countries, which some empirical studies (1,2) have shown is a function of the relative size of the investment flows in and out of the two countries.

This book combines evidence from interviews conducted in a number of countries with detailed studies of treaty provisions and case law, to give us practical examples that help us test these hypotheses. The evidence base is are not always spelled out, so it’s not always clear on what basis the author has reached a conclusion such as that cited from the chapter written by a Ugandan tax lecturer and legal practitioner above. Still, it’s an interesting starting point.

As I read through, one case stuck out as particularly interesting: that of Colombia, a chapter written by a tax lawyer and Academic Secretary of the Colombian Tax Institute.

The Colombian case

Colombia stayed away from negotiating tax treaties until 2004, “under the philosophy of protecting the tax base,” until the Uribe government adopted an approach described by the author as “attracting investment at any price.”

In 2005 the Ministry of Trade thus issued a priority list of major trading partners for parallel negotiations of bilateral investment treaties (BITs) and treaties…Due to the urgency of negotiations, Colombian officials decided to implement the [OECD Model] as the only available tool for negotiating with OECD Member countries.

The OECD model went on to be used in all negotiations, rather than either the Community of Andean Nations (CAN) model or the United Nations model, in part because “it was a known fact that most OECD Member countries would not even consider the CAN model as a reference” and there was a “lack of acceptance of some of the UN provisions by major trading partners.” (Other issues also prevented the CAN model from gaining traction). A Colombian model that mixed and matched provisions from the OECD and UN models was also dropped “in view of the negative response from its trade partners.”

It would appear that a dramatic change in policy, and a desire to negotiate treaties at almost any cost, left Colombian negotiators with a weak hand, possibly weaker than if they had initiated their negotiations at a slower pace, in a more planned manner. The failure of CAN members to establish and hold firm on a regional model that might have had more strength than their national models is probably also a factor.

Perhaps because it has no model other than the OECD model, Colombia hasn’t been too successful at negotiating higher withholding tax rates, and its treaties tend to have lower rates than even the OECD model.

The exception is withholding taxes on royalties, where Colombian treaties follow the UN model, which permits source taxation. In fact, the book’s overview states that “the influence of the OECD Model royalties clause is more the exception than the rule,” and that “the UN royalties clause is instead the main point of reference.” But Colombia has also included technical services, technical assistance and consulting services in its definition of royalties, which neither model does . This has created “significant conflicts” with the UK and US, so much so that Colombia has been “unable to resume negotiations” with them. (This may be an interesting example of power politics in action, because elsewhere in the book it appears that both India and China have successfully negotiated some kind of provision on fees for technical services in treaties with the US and/or UK).

Not giving way on this expanded definition of royalties  is important to Colombia because its previous tax treaties have a ‘most favoured nation’ clause, which means that if Colombia gives in to the UK and US, it must effectively give the same concession to all of its existing treaty partners.

In fact, for the Colombian tax authorities, the benefit of levying a 10 per cent withholding on the gross payments for technical services, technical assistance and consulting services constitutes one of the most important – if not the most important – benefit obtained for the country in the difficult negotiations with capital-exporting countries. For this reason, losing the ability to apply the withholding would mean losing the balance achieved in all prior treaties.