Self-Judging Security Exceptions in International Investment Agreements: Functions and Limits of Anti-Abuse Clauses

Recent international investment agreements (IIAs) often include self-judging security exceptions, but the legal impact of associated provisions designed to prevent their abuse remains unexplored. Using qualitative and quantitative methods, this paper examines the functions and limitations of such anti-abuse clauses. After clarifying similarities and differences between self-judging security exceptions in contemporary IIAs and GATT Article XXI, it analyzes the first arbitral applications of self-judging clauses in Seda v Colombia and Riverside Coffee v Nicaragua to understand how self-judging exceptions without anti-abuse clauses are interpreted and applied. It then presents a large-scale survey of anti-abuse provisions in IIAs based on an original article-level dataset, identifying their principal forms and distribution. The paper argues that these devices do not interact with the exception’s self-judging nature in a uniform manner, with some types of provisions potentially affecting self-judgment more directly than others. Treaty negotiators should not underestimate the practical impact of anti-abuse provisions and should adopt a treaty design that best reconciles State discretion and legal control.