A Survey Experiment on Climate Change Beliefs, Opinions, and Actions

We examine how individuals update their views on climate change when exposed to competing advocacy or skeptical signals on social media. In a representative survey experiment of 1,633 adults in Italy, respondents were randomly assigned to one of three conditions: (1) exposure to climate-advocacy content, (2) exposure to climate-skeptical content, or (3) an active control featuring content on gender equality. We elicit beliefs about the reality of climate change and its human causes both before and after treatment, along with climate-related opinions and actions. We first focus on beliefs and interpret their dynamics by developing a structural updating model that nests Bayesian-like updating within a framework allowing for anchoring to a survey-wide reference point, underreaction to information, and treatment-specific experimenter-demand effects. We find that the advocacy treatment significantly increases beliefs about the reality of climate change but not about human causation, whereas the skeptical treatment shifts beliefs toward skepticism in both domains and with larger magnitude. Underreaction patterns imply opposite belief dynamics: with equally likely advocacy and skepticism signals, intermediate beliefs fluctuate closer to advocacy regarding climate-change reality but closer to skepticism regarding human causation. Opinions and actions are largely mediated by posterior beliefs, though we also detect belief-independent effects, including a backfire in support for a fossil-fuel tax under skepticism and an additional positive effect of advocacy on signing a climate-action petition.