The Catalyst’s Election Model considers two sets of variables: state-wide demographic data and candidate strength. For demographic data, the model accounts for variables representing a state’s racial composition, level of educational attainment, and key economic indicators. Candidate strength, on the other hand, is measured using state polling, national polling, and state-wide partisan leans. To calculate probabilistic estimates for key Senate races, the model runs 200,000 simulations for every race and gives an expected margin of victory. 


The model was constructed by Zoraiz Zafar ’24.

The graphic was built by the ARC GIS lab at Colorado College. We specifically want to extend special thanks to Matt Cooney and Josh Birndorf.