Abstract
The objective was to predict the 2024 US Presidential Election. Establish models are parsimonious with two predictors explaining about 90% of the variance. The hot button issues of immigration and abortion together could not achieve similar variance because of data limitations. Other variables were added to achieve variance near 90%. The predictors are net differences in partisan trust in government; the consumer price index (CPI); annualized abortion growth rate; annualized number of immigrants as a proportion of the population; and the annualized prison population growth rate. The outcome is party in power candidate. A logistic regression model was built. An increase in CPI reduces the log odds of the party in power candidate remaining in power. An increase in all the other predictors increase the log odds of the party in power candidate winning which means Joe Biden would win, and now his replacement Kamala Harris will win.
Supplementary materials
Title
Variables, Source, Data Cleaning and Preparation
Description
This is the definition of the variables used, the source of the variables with weblink where the data can be found. Also how the data were cleaned and prepared
Actions
Title
The Dataset
Description
This is the data set with the outcome and predictor variables
Actions