The humanitarian localization agenda calls for power and resources to be transferred from international actors to local and national responders in the places affected by crisis. Despite a consensus that humanitarian responses should become more localized, debates persist on how best to achieve this practically. Moreover, there is little evidence to show how such a shift in power affects humanitarian outcomes.

This paper asks how localization is understood and implemented in the Rohingya refugee response in Cox’s Bazar in Bangladesh, and with what consequences for the Rohingya population there. It argues that the response scores high on many of the criteria against which localization is measured. For example, Bangladeshi NGOs and Bangladeshi staff of international agencies have significant power and resources. However, the response also exemplifies many of the critiques of the localization agenda, notably in terms of who constitutes a “local” actor and how well local actors can withstand governmental pressures and maintain their independence and neutrality. Moreover, localization in the Rohingya refugee response fails to achieve some of the most important underlying goals of the localization agenda and may even be counterproductive for the achievement of some of those goals. Indeed, many Rohingya individuals and community groups believe that the layers of Bangladeshi bureaucracy and NGOs between international actors and themselves have restricted their own access to power and resources, and reduced the quality and effectiveness of the humanitarian response. Finally, the paper draws out some more general conclusions about the shortcomings of the localization agenda and dominant conceptualizations of localization when applied in refugee crises.