<p>The extant literature has suggested that threatening time periods and situations can enhance people’s political conservatism levels. This thesis provides a systematic examination of the impact of societal threat on political conservatism, and whether distinct types of threat (economic, natural, and social) differentially impact political conservatism. In particular, the present research examines two main competing hypotheses. The conservative shift hypothesis (Jost, Glaser, Kruglanski, & Sulloway, 2003b) postulates that people become more politically conservative during/after threatening periods. The cultural worldview enhancement hypothesis (Pyszczynski, Solomon, & Greenberg, 2003) postulates that people tend to cling more strongly to what they believe when their death is made salient, be it a more conservative or liberal political belief – we assume that societal threat could also make mortality salient, enhancing a held political worldview. Additionally, the present research also examines in more depth how societal threat impacts political conservatism. The dual-process mediation hypothesis suggests that the impact of societal threat on political conservatism happens via variables in the dual-process motivational (DPM; Duckitt, 2001) model, being differentially mediated by world beliefs (dangerous and competitive) and socio-political attitudes (right-wing authoritarianism and social dominance orientation). Finally and because the direct impact of threat on authoritarianism has been questioned (Feldman & Stenner, 1997), the activation of authoritarianism by social threat hypothesis suggests an interaction between social threat and previous levels of right-wing authoritarianism in predicting political conservatism and variables in the DPM model. Seven empirical studies were carried out to test these hypotheses using different methodological designs. The correlational findings of Study 1 showed that the DPM model was useful in predicting political conservatism. The mix-method findings of Study 2A showed that economic, natural, and social threats have distinct psychological meanings. Based on these initial findings, two mix-method studies were conducted to develop textual (Study 2B) and pictorial (Study 2C) experimental stimuli depicting economic, natural and social threat scenarios plus a control scenario. Three experiments were then conducted to test the complete set of hypotheses: Study 3A was completed online with textual stimuli, Study 3B was completed in a lab with textual stimuli, and Study 3C was completed in a lab with pictorial stimuli. Overall, the experimental findings provided more support for the conservative shift hypothesis with participants showing greater conservative political orientation after the threat manipulation (compared to the control condition) in Study 3C. However, this effect did not generalise to other measures of political conservatism (right-wing political orientation, conservative voting intention and preference for political discourses of conservative parties). Additionally, some distinctions between threat types were observed. The dual-process mediation hypothesis was partially supported in Studies 3A and 3C. The threat manipulation impacted political conservatism indirectly via an increase in competitive world beliefs and an increase in RWA in Study 3A, while it impacted political conservatism via an increase in dangerous world beliefs and increases in right-wing authoritarianism and/or social dominance orientation in Study 3C. No empirical support was observed for both the cultural worldview enhancement and the activation of authoritarianism by social threat hypotheses. Overall, the present research suggests that it is possible to experimentally modify people’s political conservatism using threat manipulation, that threat enhances (at least to some extent) political conservatism, and that this impact may be differentially mediated by variables in the DPM model. These conclusions are discussed in more depth along with limitations and future directions in the general discussion of the thesis.</p>