A new approach to polarization modeling using Markov chains

Abstract

Abstract: In this study, we approach the problem of polarization modeling with Markov Chains (PMMC). We propose a probabilistic model that provides an interesting approach to knowing what the probability for a specific attitudinal distribution is to get to an i.e. social, political, or affective Polarization. It also quantifies how many steps are needed to reach Polarization for that distribution. In this way, we can know how risky an attitudinal distribution is for reaching polarization in the near future. To do so, we establish some premises over which our model fits reality. Furthermore, we compare this probability with the polarization measure proposed by Esteban and Ray and the fuzzy polarization measure by Guevara et al. In this way, PMMC provides the opportunity to study in deep what is the performance of these polarization measures in specific conditions. We find that our model presents evidence that in fact, some distributions will presumably show higher risk than others even when the entire population holds the same attitude. In this sense, according to our model, we find that moderate/indecisive attitudes present a higher risk for polarization than extreme attitudes and should not be considered the same scenario despite the fact that the entire population maintains the same attitude.