Social Media as a Powerful Tool for Health Policy Evaluations: Using Twitter Data to Reveal How the Public Reacted to the Philadelphia Beverage Tax

This entry is part 5 of 11 in the series Nov 2023

We demonstrate how social media data can help researchers and policymakers understand how the public reacts to a health policy (anger, fear, trust), where and when these conversations occur, and who shapes them.

When researchers evaluate the impact and implementation of health policies, they often focus on its health, economic, and social outcomes. However, one question that is often particularly difficult to answer is how everyday people experience and react to the policy. Public reactions and conversations about health policies can be very volatile yet have enormous consequences on these policies’ impact and their long-term political and social sustainability. Social media data can be insightful in answering this question, and in our recently published analysis, we examined longitudinal changes in Twitter (now X) conversations about the Philadelphia Beverage Tax, which taxed beverage distributors $0.015 per ounce of sweetened beverages to raise revenue for public services.

Click to Read the Article

Specifically, we used the Academic Twitter Application Programming Interface (API) to longitudinally evaluate 45,891 tweets related to the Philadelphia Beverage Tax during its implementation (2016-2019). By analyzing tweet content, authors, engagement, and location in parallel to key events during the implementation of the tax, we asked the following questions: what events were associated with more conversations about the PBT? How did the sentiment of conversations about PBT change over time? Who were the most influential conversation-makers on the PBT? Finally, where in the US were conversations about the PBT occurring?

What were our key findings?

  • Although anger and fear sentiments were high in the months before the Philadelphia Beverage Tax implementation, they progressively decreased after its implementation, whereas trust sentiments displayed an inverse trend.
  • The passage of a health policy (particularly the months prior), litigation related to the policy, and first reports of its impact were key stages of heightened public conversation.
  • Media outlets and public personalities (journalists, commentators, philanthropists) were influential tweeters, although their tweets often displayed polarized sentiments (either highly negative or highly positive).

What are the implications of our findings?

  • Researchers and policy makers can leverage social media as a cost-effective and insightful way to get information on public experiences of a health policy.
    • However, it is important to complement these insights with other data sources (such as surveys and interviews) for a more comprehensive overview of public opinions.
  • Policy makers should carefully analyze the social media conversation landscape during policy implementation to better use these platforms to communicate information about health policies.
    • Given the potential of influential tweeters to shape the frequency and sentiments (both positive and negative) of conversations, it is crucial to acknowledge, and potentially leverage, their influence in social media.
  • Given Twitter (now X) has now ended the free access of its API for academics, researchers may need to explore other ways of leveraging social media data for health policy evaluations.
    • For example, Facebook and Instagram offer a similar (currently free) service to analyze publicly available posts on their platforms. However, access is currently limited to specific types of research questions (eg, COVID or election related).

To learn more about our findings and how we collected and analyzed data, be sure to check out our full report: “Leveraging Multiyear, Geospatial Social Media Data for Health Policy Evaluations: Lessons From the Philadelphia Beverage Tax” in the Nov 2023 issue of the Journal of Public Health Management and Practice.


Shahmir H. Ali is a postdoctoral fellow at the National University of Singapore who conducted this research while completing his PhD at the NYU School of Global Public Health. His current research focus on family-led and community-led approaches to equitable cardiovascular disease prevention in Asian, immigrant, and ethnic minority populations.

Dr. Angela Trude is an Assistant Professor in the Department of Nutrition and Food Studies at New York University. Her research encompasses two main areas: i) design, implementation, and evaluation of community-based interventions in low-resource settings; ii) food policy strategies to ensure equitable access to affordable, nutritious food.

Other Authors

Caitlin M. Lowery is a PhD Candidate at University of North Carolina at Chapel Hill, Gillings School of Global Public Health. Her research interests include policy approaches to prevent chronic disease, promote health and well-being, and reduce health disparities.

You Might Also Enjoy:

Series Navigation<< Building New York State’s Wastewater Surveillance NetworkEmerging Leaders in Public Health: Leaders Develop in Action >>