Defenses and the Burden of Proof in International Law
Forthcoming in: EXCEPTIONS IN INTERNATIONAL LAW (L. BARTELS & F. PADDEU, OUP, 2017)
28 Pages Posted: 4 Nov 2016 Last revised: 26 Jul 2017
Date Written: November 3, 2016
Modern treaties include an increasingly complex set of rules, carve-outs and exceptions, ensuring, for example, the policy space of states. Their impact may be tempered also by outside instruments or custom. International law disputes, in turn, are increasingly fact-intensive, addressing, for example, complex scientific, environmental or economic questions. This double evolution enhances the role, in international litigation, of defenses and the importance of who carries the burden of proof for them. Some international tribunals may spend more time than necessary on questions of burden of proof and different types of claims in defense (dispute settlement at the World Trade Organization (WTO) comes to mind); most, however, tend to under-estimate the issue and do not devote it enough attention (for example, the International Court of Justice).
The burden of proving a defense is said to be on the party invoking it. As trite as this proposition may sound, in international law, it hides a far more complex litigation reality. Distinctions must be made both in terms of (i) types of claims in defense, and (ii) types of burdens this may impose (or not) on the respondent.
This paper distinguishes six different “claims in defense”: (i) objections to jurisdiction, (ii) objections to admissibility, (iii) exemptions/alternative rules, (iv) absence of breach, (v) exceptions, and (vi) defenses under secondary rules.
For each of these six “claims in defense”, five types of “burdens” on the parties or tribunal are identified: (i) burden of raising a claim in defense, (ii) burden of production of evidence, (iii) burden of persuasion (or real burden of proof), (iv) standard or quantum of proof and (v) standard of review.
Although for some claims in defense some types of burden are on the defendant, this is certainly not the case for all.
Suggested Citation: Suggested Citation