Michael Lewis-BeckUniversity of Iowa
Charles TienHunter College, CUNY & The Graduate Center, CUNY
The Political Economy model represents one of the earliest political science forecasting models for presidential elections. It relies on a multiple regression equation with just a few predictor variables, drawing from leading theories of vote choice, measured several months before the election. Much has been said about the uniqueness of the 2020 election campaign, e.g., COVID-19, the extreme character of Trump’s rule, the dramatic economic collapse, widespread protests ignited by George Floyd’s killing, among other things. We do not deny these circumstances. However, we believe their effects are absorbed, then expressed, via the causal structure the model suggests. Here we apply the model to forecast the 2020 presidential election (popular and Electoral College votes), followed by forecasts for the House and Senate elections. We forecast a Democratic sweep.
pptx : 0.08mb