Social Media and Donations During Disasters

Disaster response organizations (DROs) utilize social media to support fundraising efforts. They post appeals for donations and share updates on their relief efforts to engage potential donors. While prior research has shown that DROs' social media activity and public engagement can boost donations, it has typically considered these factors in isolation. However, DROs' donation appeals unfold in a dynamic environment where organizational messaging, user engagement, and public discourse may influence one another over time. Drawing on theories of prosocial behavior, we conceptualize a social media system for donations comprising DRO social media activity, public engagement, and public interest, and examine how these elements interact to influence donations over time. To test this framework, we partnered with the Canadian Red Cross and analyzed hourly Twitter activity and donation data from the 2016 Fort McMurray wildfire. Using a vector autoregressive (VAR) model, we show that DROs' tweets directly increase donations through greater impressions and clicks. Moreover, we identify three indirect pathways to donations: (1) diffusion through retweets; (2) increases in user-generated content about the disaster; and (3) heightened public interest. These indirect effects eventually surpass the direct effect such that after ten hours, they increase donations by about 35% versus 26% for the direct effect. We also find that DROs' tweets expressing positive empathy are more effective than those expressing negative empathy. Overall, our findings reveal the dynamics through which DROs' social media activity affects donations and offer guidance for designing social media strategies that help DROs raise funds for their operations.