Replication Data for: Citizen Forecasting 2020: A State-by-State Experiment

The research uses 'citizen forecasts' to predict the US Presidential Election. This approach asks citizens to forecast which presidential candidate will win in their state and the nation as a whole, and predicts the winning candidate to be the one which most citizens say will win. Previous studies have shown that 'citizen forecasts' predict better than any other approach in Great Britain (Murr et al. forthcoming) and the United States (Graefe 2014). But the timing of the data collection forced most of the studies using citizen forecasts to forecast elections ex post, that is after they occurred. The proposed research asked survey questions on Amazon.com's Mechanical Turk in mid and late July. The survey design and questions parallel come from previous American National Election Studies. The survey responses are used to predict which candidate will carry a state and which candidate will win the Presidency. This will provide a strong test for the accuracy of 'citizen forecasting' in the United States.