Title:Social Media Study of Public Opinions on Potential COVID-19 Vaccines: Informing Dissent, Disparities, and Dissemination

Abstract:The current development of the vaccines for SARS-CoV-2 is unprecedented. However, little is known about the fine-grained public opinions on the coming vaccines. Using more than 40,000 rigorously selected tweets (from over six million tweets collected using keywords) posted by over 20,000 distinct Twitter users, we adopt a human-guided machine learning framework to capture the public opinions on the potential vaccines for SARS-CoV-2 and classify them into three groups: pro-vaccine, vaccine-hesitant, and anti-vaccine. To investigate the scope and causes of public opinions, we compare the demographics, social capitals, income, religious statuses as well as the political affiliations of these three groups and find significant differences. A lower acceptance level for the potential COVID-19 vaccines is observed among the socioeconomically disadvantaged groups. We further aggregate the opinions at the state and country levels, and find that the percentage of the pro-vaccine group is lower in the Southeast part of the United States. Interestingly, the major changes in the percentages of different opinion groups roughly correspond to the major pandemic-related events. Next, by conducting counterfactual analyses, we find that the U.S. public is most concerned about the safety, effectiveness, and political issues with regards to the potential vaccines for COVID-19. In particular, we also investigate the public opinions about the newly emerged mRNA vaccines and find the opinions vary significantly across the religious beliefs: There are proportionally more religious people in the anti-vaccine group. To our best knowledge, this is the first large-scale social media-based study to analyze the public opinions on the potential COVID-19 vaccines that can inform more effective vaccine distribution policies and strategies.