Over the last 50 years, the role of polling has taken on increasing prominence in the American electoral process. This book brings together some of the leading figures in political science to present their election forecasts, discuss their methodology and present their critiques of the forecasting enterprise. They consider whether ever more accurate models of predicting voting behaviour damage the political climate by making politicians increasingly enslaved by pollsters, and the effect on turnout when there is a broad consensus as to who the winner will be.