In October 2015, the G20 Finance Ministers endorsed the base erosion and profit shifting (BEPS) project, which has been acclaimed as the most fundamental transformation of the international tax rules in almost a century. The BEPS project contains 15 actions, of which Action 14 aims to strengthen the mechanisms of international tax dispute resolution, particularly the mutual agreement procedure (MAP). This article provides a theoretical reflection on a key initiative of Action 14—the new MAP Statistics Reporting Framework (vis-à-vis the 2007 MAP Statistics Reporting Framework). While scholarly attention to this initiative is scant, this article explores its theoretical and practical implications. Specifically, the logic underlying this framework is to structure the MAP process so that the contribution of an individual competent authority could be segregated and measured and that the agency problem characterizing the MAP process could be mitigated. Based on this logic, several prospects of improving the current mechanism of MAP have been envisaged. First of all, it is proposed that the MAP process could be further structured so that the performance of the competent authorities could be better measured. Second, a more structured MAP process lends itself to more comprehensive domestic judicial review. In a broader sense, this article could also be regarded as a response to a methodological gap in the past studies of international tax dispute resolution—the shortage of theoretical reflection in this area.