The national identity divide : a mixed-methods, social representations approach to political identity and polarisation in the United States

Evidence indicates that US partisan and ideological identities shape political outcomes including the recent increase in affective political polarisation. As predicted by social identity theory, self-placement as a US liberal or conservative, or as a Democrat or Republican, is associated with favouritism toward the ideological in-group and negative attitudes and behaviours toward the outgroup. The theory also holds that the link between self-categorisation and behaviour is mediated by the content of that identity, by what an individual believes it means to be a member of that group (Huddy, 2001). This thesis employs a mixed methods sequential design to investigate the meanings of US political identities and the relationship of these meanings to affective political polarisation. It builds on social identity theory work by drawing on the social representations approach (Elcheroth et al., 2011), conceptualising social representations as the building blocks of social identity content (Breakwell, 1993b). Four empirical studies comprise this investigation. Study 1 is a qualitative analysis of lay representations of US liberal and conservative identities. A number of asymmetries were identified, including a difference in national attachment which contrasted a conservative group-centric, symbolic attachment (named national reverence) with a liberal individual-centric, instrumental attachment (named individual support). These organising principles of national reverence and individual support identified in Study 1 were then operationalised in Studies 2a and 2b through exploratory and confirmatory analyses. The principles were found to be significant and substantive descriptors of ideological identity and predictors of political outcomes including voting behaviour and affective political polarisation. In addition, national reverence and individual support were superior predictors of these variables as compared to right-wing authoritarianism, social dominance orientation, and nationalism. Study 3 examined national identity attachment on the political left through a qualitative analysis of Democratic candidate speeches during the 2019/2020 primary season. This work revealed a fractured representation of national attachment across the Democratic ideological spectrum with tensions between individual-centric and group-centric representations of the nation. Finally, Study 4 explored the causal role of the principles in affective political polarisation in the presence of issue positions and group demographic information. The study pointed to the substantive role of the principles in affective political polarisation, but noted that the contribution of the principles, issue positions, and demographic information varied by party. Overall, the studies suggest that the meanings of US national and ideological identities are highly intertwined and that these meanings are predictive of political outcomes, including affective political polarisation. This programme of investigation contributes new measures (national reverence and individual support) that are not only useful descriptors of US political identities, but also strong predictors of political outcomes. In addition, this thesis contributes theoretically and methodologically to the study of political identity and affective political polarisation. By framing social representations as elements of identity content, the political identities that drive outcomes are conceptualised as dynamic, contextual, and changeable. The implications of this framing, and of the findings related to the dominance of conservative representations of national identity, are discussed. Overall, this thesis underscores the need to study political outcomes using a theoretical framework that prioritises the issues of identity content and context-what political identities mean to people in particular socio-political environments.