Measuring affective polarization, examining its potential as a sociopolitical stressor, and investigating the role of identity threat : a longitudinal investigation : a thesis submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Psychology at Massey University, Palmerston North, New Zealand

Abstract

Affective polarization—antipathy between members of sociopolitical groups—appears to be increasing in many countries around the world, prompting a surge of research aiming to mitigate its purportedly pernicious effects on society. It is construed largely as an expressive phenomenon driven by identity threats and exacerbated by extended, vitriolic political campaigns in the era of electronic media. It has recently been hypothesised to have an intrapersonal effect on stress, health, and social support. Progress in understanding affective polarization has been hindered by inconsistent measurement and by the fact that many studies on its causes and consequences use correlational research or survey vignette experiments. This PhD project sought to advance the field by (1) developing a psychometrically sound measure of affective polarization to address scale proliferation, (2) investigating social identity threat as a cause of affective polarization, and (3) testing whether affective polarization (expressed political prejudice) or social identity threat (anticipated political prejudice) exert harmful effects on wellbeing. To investigate (2) and (3), we used a longitudinal survey design and statistical models which are capable of disaggregating within- and between-person variance, to more accurately estimate causal effects between the constructs within individuals over time. In study 1, the Affective Polarization Scale was developed. Exploratory and confirmatory factor analyses supported a reliable three-factor structure. The scale demonstrated good construct validity, replicating associations with authoritarianism, need for closure, identity strength, and intellectual humility. Study 2 and 3 used a three-month longitudinal survey of American participants preceding the 2024 US presidential election. Random-intercept cross-lagged panel models revealed that social identity threat predicted subsequent increases in affective polarization. No effects of affective polarization or identity threat on social support, stress, or health were detected, though stress had a significant effect on affective polarization, and affective polarization was correlated with greater stress, lower social support, and poorer health. No trends in affective polarization, social identity threat, or stress were seen, which conflicts with past research on the effect of contentious elections. Overall, this thesis contributes to the study of affective polarization by presenting a new affective polarization scale. It then shows that affective polarization may be a defensive phenomenon; driven by anticipated political prejudice and by stress. This suggests that the onus for mitigating opponents' polarization may lie, at least in part, with each of us and what we signal to opponents. It also suggests that claims about harmful effects of identity threat and polarization on wellbeing may arise from a reliance on cross-sectional designs that fail to distinguish between within-person and between-person associations.