Disentangling positive and negative partisanship in affective polarization using a coevolving latent space network with attractors model

We develop a broadly applicable class of coevolving latent space network with attractors (CLSNA) models, where nodes represent individual social actors assumed to lie in an unknown latent space, edges represent the presence of a specified interaction between actors, and attractors are added in the latent level to capture the notion of attractive and repulsive forces. The models are estimated using Bayesian inference. We apply the CLSNA models to the question of affective polarization, which expects Republicans and Democrats to cohere, favor and interact with their own party and to distance, repel and interact less with the opposing party. Using two different longitudinal social networks from the social media platforms, Twitter and Reddit, we investigate the relative contributions of positive (attractive) and negative (repulsive) affect among political elites and the public, respectively. Our goal is to uncover and quantify polarization -- and disentangle the positive and negative forces within and between parties, in particular. Our analysis confirms the existence of affective polarization among both political elites and the public. While positive partisanship remains dominant across the full period of study for both Democratic elites and the public, a notable decrease in Republicans' strength of in-group affect since 2015 has led to the dominance of negative partisanship in their behavior.