If there was ever a need to demonstrate the need to improve international tax dispute resolution mechanisms, it was as a speaker at the TP Minds conference earlier this month. In the course of our panel discussion on dispute resolution post BEPS, I asked the audience of over 100 tax directors and managers of multinational companies how long their mutual agreement procedure (MAP) cases were taking to resolve. The electronic (anonymous) poll showed that not one current dispute lasted less than five years. Given that most treaties permit a case to be presented for MAP up to three years of first notification of taxation not in accordance with the treaty, clearly international tax disputes are taking a very long time to resolve, even before the BEPS Actions bite.
BEPS Action 14 recognised that the interpretation and application of new rules resulting from the work product of the other BEPS actions, could introduce undesirable uncertainty. Action 14 therefore called for work to examine and address obstacles that prevent countries from resolving treaty-related disputes under the mutual agreement procedure (MAP).
The Minimum Standard
The Final Report on Action 14 proposed a minimum standard on dispute resolution, consisting of specific measures to remove obstacles to an effective and efficient MAP. The objectives of the minimum standard are that:
- treaty obligations related to the mutual agreement procedure must be fully implemented in good faith and that MAP cases resolved in a timely manner;
- administrative processes must promote the prevention and timely resolution of treaty-related disputes; and
- taxpayers who meet the requirements of paragraph 1 of Article 25(1) can access the MAP.
Each of these objectives is elaborated by detailed requirements that countries should commit to, and are backed up by 11 MAP best practices.
The only mandatory Action 14 obligation that is enshrined in the BEPS Multilateral Convention (Article 17) is the inclusion of the OECD Model Article 25(1)-(3) MAP provisions. Countries intending to become party to the Convention will be required to agree to this, or to other mechanisms by which those, or equivalent provisions, are included in their bilateral treaties.
International legal effect
How then are the minimum standards to be given effect, if they are not included as part of the treaty obligations that participating states become bound in international law to adopt? The OECD answer is that OECD and G20 countries, plus others that commit to the minimum standard, will undergo peer reviews by the FTA MAP Forum of their the legal framework including tax treaties, domestic law and regulations, as well as administrative guidance to evaluate implementation of the minimum standard by each state. Peer reviews will report on the strengths and shortcomings of each reviewed state and recommend improvements. It is intended that the first reports will be published by the end of 2017.
With over 100 countries showing interest in the BEPS Multilateral Convention, the process is likely to take time with uncertain outcomes, particularly where legislative reform is required.
Shortcomings in existing MAP have been well documented by the OECD and others. 27 July 2004, The 2004 OECD report “Improving the Process for Resolving International Tax Disputes”, which set out 31 reform proposals and the 2007 report “Improving the Resolution of Tax Treaty Disputes” which proposed arbitration and introduced the Manual on Effective Mutual Agreement Procedure, both identified difficulties. These include, time limits, limited taxpayer participation, double collection of tax pending resolution of a case, denial of access to MAP and an lack of uniform relationships between domestic remedies and MAP.
Recent judicial decisions demonstrate that taxpayers are not without legal rights in this area, even where OECD recommendations are not given explicit legal force in domestic law. In CGI Holding LLC v. Canada (National Revenue) 2016 FC 1086 (CanLII), the Canadian Federal Court held that the court had jurisdiction to consider the administrative actions of the tax authority within MAP process. The conduct of MAP fell to be determined in accordance with the normal standards required by administrative law. Any discretionary decision-making of the executive including MAP is to be reviewed on a standard of reasonableness. This standard is supported by the treaty language establishing MAP. A broad margin of appreciation is afforded to the competent authorities in the context of the MAP process. Thus the court will only intervene if it can be demonstrated that the decision falls outside the range of possible, acceptable outcomes defensible in respect of the facts and the law or if there is a breach of procedural fairness rights.
The central difficulty with MAP is that the competent authorities are mandated by Article 25 to “endeavour to resolve the case” rather than “to resolve the case”. The result is that there is no obligation to agree. The English High Court has recently ruled that it is possible to give sensible content to an undertaking to endeavour. In Astor Management AG & Anor v Atalaya Mining Plc & Ors  EWHC 425 (Comm) the court held that where there is a clearly defined object of the endeavours (in MAP it is taxation in accordance with the treaty) it is simply a question of fact whether the person required to do so has endeavoured to resolve the case.
Access to information
The burden of proving a failure to endeavour is on the taxpayer in a judicial review of the competent authority. Suitable evidence may not be easy to obtain given the limited taxpayer participation in MAP. However in SA Garlon v. Belgian State (No. 225.438), the Belgian Raad van State (Supreme Administrative Court) ruled that a taxpayer may have access to mutual agreement correspondence. In coming to this conclusion, the Court referred to the OECD 2007 Recommendations and Manual on Effective Mutual Agreement Procedure as “normative”. With OECD Action 14 work product agreed as a minimum standard, courts will have a firm framework to judge the endeavours of competent authorities in judicial review proceedings.
MAP by its nature ought to be a collaborative process. As with all genuine disputes, while there may be different positions, not everybody can be right and there are winners and losers. The best route out of the problems posed by MAP available today is still mandatory binding arbitration.