Social Media for Disaster Management: A Study of Hurricane Sandy

Abstract

The use of social media to support the disaster management cycle is increasing. Organizations use social media to inform users about operational activities such as aid distribution, and receive information from users about demand for help and in-kind donations. As such, this research investigates the social conversation between relief organizations and social media users at the disaster stages of preparedness, response and recovery. We use Facebook data from five benchmark organizations that responded to Hurricane Sandy, 2012. We create a novel data dictionary to classify actionable information (e.g. information related to aid distribution, donations and volunteering), and build an econometric model to examine the social conversations between organizations and users. By analyzing all the organizations' posts and users' comments, we find that informational support is most effective during disaster response. Nevertheless, there is a mismatch between the actionable information that organizations post and the one that users are interested in. While organizations focus on informing users about aid distribution, users are asking about how they as individuals can either donate or volunteer. This research provides insights on how organizations can mitigate the mismatch and improve their social conversation with users, which has operational implications on unsolicited in-kind donations and volunteer management. Our findings are relevant as disaster operations such as the Hurricane Harvey response in 2017 engage millions of social media users that can exchange valuable information with relief organizations.

Keywords: Disaster Management, Content Analysis, Humanitarian Operations, Optimal Matching Theory, Social Media