Why Do Citizens' Election Predictions Differ Across States? A State-by-State Experiment

Introduction: What happens when ordinary citizens try to predict election outcomes based on their own knowledge and beliefs? This study investigates citizen forecasting, an emerging phenomenon in the 2020 U.S. elections.

The Experiment: We conducted a state-by-state analysis of how people across different U.S. states predicted electoral results.

* Data & Methods: Leveraged survey data collected from participants in all 50 U.S. states during early voting periods, combining self-reported beliefs with actual election outcomes.

* Key Findings:

• State residents were significantly more accurate than random chance when predicting local or state-level elections
• Accuracy varied dramatically based on political engagement and media consumption habits
• Higher education levels correlated with better forecasting accuracy in all states examined

Why It Matters: These findings suggest that ordinary citizens can effectively predict election outcomes, particularly for issues closely affecting their daily lives. This challenges assumptions about the need for professional pollsters and data analysts in democratic processes.