Forecasting US Presidential Elections: A Political Economy and Security Model

• In Event: Forecasting the 2024 US Presidential Election

Abstract

Some of the most prominent and accurate models for forecasting US presidential elections are political economy models. These models suggest that politics and economics evident in the president’s approval ratings and the state of the economy influences voter preference. The political economy and security model outlined below builds on the political economy models. This political economy and security model argues that American voters’ support for the president (his approval ratings), the state of the economy (indicated by consumer spending and the number of building permits), and the perceived state of security in the country (indicated by the homicide rate) and the fear of (immigration) determines the party in power in the White House. A logistic regression model will be built using data from 1952 to present to forecast the 2024 Presidential Election. A jack-knife resampling will also be conducted to evaluate model performance.

Author

• Christopher Andrew Dwight Charles, University of the West Indies, Mona