Computational and Analytical Modeling for Senate Decision Dynamics

For decades, congressional scholars have struggled to measure and describe the impact of influences on Senate decisions. While quantitative data on these political issues is scarce, this study introduces an interdisciplinary stochastic model to determine Senate decision outcomes based on assumptions about influence. We consider both interactions between Senators and the effect of external lobbying. The model is based on the classic Ising Model, and the system is defined as a 1D lattice with 100 sites, where the sites hold values for the senators’ political convictions. Peer influence is analogous to the spin-to-spin coupling constant in the Ising model, and the effect of professional lobbying replaces the interaction of the spins with an external magnetic field. Lobbying influence approximates the effects of social media campaigns, face-to-face meetings with Senators, and financial contributions. Peer-to-peer influence is modeled by using each Senator’s partisan score, such that the influence is a function of the numerical difference between these scores. We analyze the model using Python and matrix algebra. Given a specific political issue, the model predicts the outcome of the senate vote.