Politics that Divide and Ties that Bind: Family, Friends, and Neighbors in a Polarized Era

Abstract

Affective polarization has increased in the United States over the last several decades, leading to negative consequences for relationships, communities, and democracy. Why is affective polarization rampant, and what are the implications? In the first article of this dissertation, I examine a potential cause of affective polarization in the US: geographic sorting. We demonstrate that when people live in areas where they have little exposure to the out-party, they report generally higher levels of affective polarization. We follow this up by experimentally manipulating cross-party contact for students from politically homogenous universities and find that cross-party contact can reduce affective polarization for those in political bubbles. We argue that this cross-party contact is important for reducing affective polarization, and that increased political homogeneity within communities may have led to rising affective polarization in the United States. In the second article of my dissertation, we examine whether perceived political difference is associated with worse relationship quality between college students and their parents. We find that when controlling for covariates such as non-political difference, parents’ perceptions of political difference from their child is not associated with worse relationship quality. However, greater political difference was associated with worse relationship quality for students in some cases, but the effects were not large enough to be considered significant based on our hypotheses using minimum effect testing. In the third article of my dissertation, we propose a framework for statistical inference using minimum effect testing and equivalence testing. We use these methods in article two. This proposed framework allows researchers to test more specific hypotheses with a simple and versatile method that can be used for research questions. Collectively, my dissertation furthers our knowledge of the causes and implications of affective polarization, as well as providing a statistical framework that can improve statistical inferences and research practices in quantitative research.