Patterns of Investor-State Dispute Settlement Decisions

This paper examines the pattern of settlements, investor wins and state wins in investor-state dispute settlement cases, focusing on the impact of arbitration rules, the resources and experience of the litigants, the type of treaty breach under investigation, and the pattern of state intervention. Logistic regression results show that settlements are more likely when investors do not disclose their allegations publicly, when the host state has been involved in a greater number of past cases, and when the respondent state has a higher control-of-corruption rating. Panel awards to investors are more likely in cases of alleged direct expropriation and less likely for indirect expropriation. Panel decisions favoring the state increase with higher per capita income and rule-of-law ratings.