121 4 psychological inconsistencies (Krosnick, 2002: 117). One study suggests that voters even project their own month of birth onto politicians (Castelli et al., 2009). Projection can take place in a positive (i.e. voters expect politicians to hold the same policy position) or negative (i.e. voters expect politicians to hold the opposite policy position) sense (Krosnick, 2002: 117). Among researchers of projection, there have been many debates on whether people project positively onto in-groups while projecting negatively onto out-groups (i.e. “differentiation hypothesis”), or whether people project positively to out-groups as well (i.e. “induction hypothesis”) (Robbins and Krueger, 2005: 34, 35). Research in the United States suggests that white voters usually stereotype black politicians as being left-leaning on economic and cultural issues, while black voters mostly project their own attitudes onto black politicians (Lerman and Sadin, 2016), which points towards the “differentiation hypothesis” (Robbins and Krueger, 2005: 35). However, a meta-analysis concluded that, on a whole, there is more support for the “induction hypothesis:” people project their views onto both in-groups and out-groups, more strongly to in-groups, but out-group projection is positive nonetheless (Robbins and Krueger, 2005: 40). Scholars of the “induction hypothesis” attribute positive projection to both in-groups and out-groups to various forms of perceived similarity (Davis, 2017; Machunsky et al., 2014; Stathi and Crisp, 2008). Being part of the in-group garners the highest positive projection effects, followed by being in contact with the out-group (contact theory). “Extended contact,” having friends who are friends with the particular out-group, garners positive projection effects as well. Even “imagined contact” induced by merely priming respondents to think about being in contact with an out-group increases out-group projection (Stathi and Crisp, 2008: 944) and reduces perceptions of out-group homogeneity (stereotyping) (Turner et al., 2007). Instead of manipulating imagined contact in a laboratory setting, a priori existing “perceived similarity” also explains projection to out-groups, even when the particular out-group is generally “disliked” (Davis, 2017). There is some debate over whether “perceived similarity” is a cause or a consequence of projection (Machunsky et al., 2014), which is why I will refrain from using causal language when discussing the relation between perceived similarity and projection. It remains unclear whether voters are more likely to stereotype politicians (e.g. this politician is Muslim, Muslims are against homosexuality, this is why I expect this Muslim politician to be against homosexuality) or project their own ideas onto them (e.g. I am in favor of homosexuality and perceive myself to be similar to Muslims, therefore I expect Muslim politicians, to think the same). Since the assumption that voters stereotype Muslim politicians as homophobic is so widely held by party gatekeepers, I pre-registered the following hypothesis: Voters use politician characteristics (stereotypes) more than their own policy positions (projection) to infer the policy positions of politicians.21 21 See pre-registration at Open Science Framework osf.io/jtdqw
RkJQdWJsaXNoZXIy MjY0ODMw